Part II

 
CONTENT

[Federal Register: August 12, 2005 (Volume 70, Number 155)]

[Rules and Regulations]

[Page 47277-47707]

From the Federal Register Online via GPO Access [wais.access.gpo.gov]

[DOCID:fr12au05-18]

[[Page 47277]]

Part II

Department of Health and Human Services

Centers for Medicare & Medicaid Services

42 CFR Parts 405, 412, 413, 415, et al.

Medicare Program; Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2006 Rates; Final Rule

[[Page 47278]]

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Centers for Medicare & Medicaid Services

42 CFR Parts 405, 412, 413, 415, 419, 422, and 485

[CMS-1500-F]

RIN 0938-AN57

Medicare Program; Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2006 Rates

AGENCY: Centers for Medicare and Medicaid Services (CMS), HHS.

ACTION: Final rule.

SUMMARY: We are revising the Medicare hospital inpatient prospective payment systems (IPPS) for operating and capital-related costs to implement changes arising from our continuing experience with these systems. In addition, in the Addendum to this final rule, we describe the changes to the amounts and factors used to determine the rates for Medicare hospital inpatient services for operating costs and capital- related costs. We also are setting forth rate-of-increase limits as well as policy changes for hospitals and hospital units excluded from the IPPS that are paid in full or in part on a reasonable cost basis subject to these limits. These changes are applicable to discharges occurring on or after October 1, 2005, with one exception: The changes relating to submittal of hospital wage data by a campus or campuses of a multicampus hospital system (that is, the changes to Sec. 412.230(d)(2) of the regulations) are effective on August 12, 2005.

Among the policy changes that we are making are changes relating to: The classification of cases to the diagnosis-related groups (DRGs); the long-term care (LTC)-DRGs and relative weights; the wage data, including the occupational mix data, used to compute the wage index; rebasing and revision of the hospital market basket; applications for new technologies and medical services add-on payments; policies governing postacute care transfers, payments to hospitals for the direct and indirect costs of graduate medical education, submission of hospital quality data, payment adjustment for low-volume hospitals, changes in the requirements for provider-based facilities; and changes in the requirements for critical access hospitals (CAHs).

DATES: Effective Dates: The provisions of this final rule, except the provisions of Sec. 412.230(d)(2), are effective on October 1, 2005. The provisions of Sec. 412.230(d)(2) are effective on August 12, 2005. This rule is a major rule as defined in 5 U.S.C. 804(2). Pursuant to 5 U.S.C. 801(a)(1)(A), we are submitting a report to Congress on this rule on August 1, 2005.

FOR FURTHER INFORMATION CONTACT:

Marc Hartstein, (410) 786-4548, Operating Prospective Payment, Diagnosis-Related Groups (DRGs), Wage Index, New Medical Services and Technology Add-On Payments, Hospital Geographic Reclassifications, Postacute Care Transfers, and Disproportionate Share Hospital Issues.

Tzvi Hefter, (410) 786-4487, Capital Prospective Payment, Excluded Hospitals, Graduate Medical Education, Critical Access Hospitals, and Long-Term Care (LTC)-DRGs, and Provider-Based Facilities Issues.

Steve Heffler, (410) 786-1211, Hospital Market Basket Revision and Rebasing.

Siddhartha Mazumdar, (410) 786-6673, Rural Hospital Community Demonstration Project Issues.

Mary Collins, (410) 786-3189, Critical Access Hospitals (CAHs) Issues.

Debbra Hattery, (410) 786-1855, Quality Data for Annual Payment Update Issues.

Martha Kuespert, (410) 786-4605, Specialty Hospitals Definition Issues.

SUPPLEMENTARY INFORMATION:

Electronic Access

This Federal Register document is also available from the Federal Register online database through GPO Access, a service of the U.S. Government Printing Office. Free public access is available on a Wide Area Information Server (WAIS) through the Internet and via asynchronous dial-in. Internet users can access the database by using the World Wide Web; the Superintendent of Documents home page address is http://www.access.gpo.gov/nara_docs/, by using local WAIS client

software, or by telnet to swais.access.gpo.gov, then login as guest (no password required). Dial-in users should use communications software and modem to call (202) 512-1661; type swais, then login as guest (no password required).

Acronyms

AAOS American Association of Orthopedic Surgeons ACGME Accreditation Council on Graduate Medical Education AHIMA American Health Information Management Association AHA American Hospital Association AICD Automatic implantable cardioverter defibrillator AMI Acute myocardial infarction AOA American Osteopathic Association ASC Ambulatory Surgical Center ASP Average sales price AWP Average wholesale price BBA Balanced Budget Act of 1997, Pub. L. 105-33 BES Business Expenses Survey BIPA Medicare, Medicaid, and SCHIP [State Children's Health Insurance Program] Benefits Improvement and Protection Act of 2000, Pub. L. 106-554 BLS Bureau of Labor Statistics CAH Critical access hospital CBSAs Core-Based Statistical Areas CC Complication or comorbidity CIPI Capital Input Price Index CMS Centers for Medicare & Medicaid Services CMSA Consolidated Metropolitan Statistical Area COBRA Consolidated Omnibus Reconciliation Act of 1985, Pub. L. 99- 272 CoP Condition of Participation CPI Consumer Price Index CRNA Certified registered nurse anesthetist CRT Cardiac Resynchronization Therapy DRG Diagnosis-related group DSH Disproportionate share hospital ECI Employment Cost Index FDA Food and Drug Administration FIPS Federal Information Processing Standards FQHC Federally qualified health center FTE Full-time equivalent FY Federal fiscal year GAAP Generally accepted accounting principles GAF Geographic adjustment factor HIC Health Insurance Card HIS Health Information System GME Graduate medical education HCRIS Hospital Cost Report Information System HIPC Health Information Policy Council HIPAA Health Insurance Portability and Accountability Act of 1996, Pub. L. 104-191 HHA Home health agency HHS Department of Health and Human Services HPSA Health Professions Shortage Area HQA Hospital Quality Alliance ICD-9-CM International Classification of Diseases, Ninth Revision, Clinical Modification ICD-10-PCS International Classification of Diseases, Tenth Edition, Procedure Coding System ICU Intensive Care Unit IHS Indian Health Service IME Indirect medical education IPPS Acute care hospital inpatient prospective payment system IPF Inpatient psychiatric facility IRF Inpatient rehabilitation facility IRP Initial residency period JCAHO Joint Commission on Accreditation of Healthcare Organizations LAMCs Large area metropolitan counties LTC-DRG Long-term care diagnosis-related group LTCH Long-term care hospital MCE Medicare Code Editor MCO Managed care organization MDC Major diagnostic category

[[Page 47279]]

MDH Medicare-dependent small rural hospital MedPAC Medicare Payment Advisory Commission MedPAR Medicare Provider Analysis and Review File MEI Medicare Economic Index MGCRB Medicare Geographic Classification Review Board MMA Medicare Prescription Drug, Improvement, and Modernization Act of 2003, Pub. L. 108-173 MRHFP Medicare Rural Hospital Flexibility Program MSA Metropolitan Statistical Area NAICS North American Industrial Classification System NCD National coverage determination NCHS National Center for Health Statistics NCVHS National Committee on Vital and Health Statistics NECMA New England County Metropolitan Areas NICU Neonatal intensive care unit NQF National Quality Forum NTIS National Technical Information Service NVHRI National Voluntary Hospital Reporting Initiative OES Occupational Employment Statistics OIG Office of the Inspector General OMB Executive Office of Management and Budget O.R. Operating room OSCAR Online Survey Certification and Reporting (System) PRM Provider Reimbursement Manual PPI Producer Price Index PMS Performance Measurement System PMSAs Primary Metropolitan Statistical Areas PPS Prospective payment system PRA Per resident amount ProPAC Prospective Payment Assessment Commission PRRB Provider Reimbursement Review Board PS&R Provider Statistical and Reimbursement System QIA Quality Improvement Organizations RHC Rural health clinic RHQDAPU Reporting Hospital Quality Data for Annual Payment Update RNHCI Religious nonmedical health care institution RRC Rural referral center RUCAs Rural-Urban Commuting Area Codes SCH Sole community hospital SDP Single Drug Pricer SIC Standard Industrial Codes SNF Skilled nursing facility SOCs Standard occupational classifications SOM State Operations Manual SSA Social Security Administration SSI Supplemental Security Income TEFRA Tax Equity and Fiscal Responsibility Act of 1982, Pub. L. 97- 248 UHDDS Uniform Hospital Discharge Data Set

Table of Contents

I. Background

A. Summary

1. Acute Care Hospital Inpatient Prospective Payment System (IPPS)

2. Hospitals and Hospital Units Excluded from the IPPS

a. IRFs

b. LTCH

c. IPFs

3. Critical Access Hospitals (CAHs)

4. Payments for Graduate Medical Education (GME)

B. Summary of Provisions of the FY 2006 IPPS Proposed Rule

1. Changes to the DRG Reclassifications and Recalibrations of Relative Weights

2. Changes to the Hospital Wage Index

3. Revision and Rebasing of the Hospital Market Basket

4. Other Decisions and Changes to the PPS for Inpatient Operating and GME Costs

5. PPS for Capital-Related Costs

6. Changes for Hospitals and Hospital Units Excluded From the IPPS

7. Payment for Blood Clotting Factors for Inpatients With Hemophilia

8. Determining Prospective Payment Operating and Capital Rates and Rate-of-Increase Limits

9. Impact Analysis

10. Recommendation of Update Factor for Hospital Inpatient Operating Costs

11. Discussion of Medicare Payment Advisory Commission Recommendations

C. Public Comments Received in Response to the FY 2006 IPPS Proposed Rule II. Changes to DRG Classifications and Relative Weights

A. Background

B. DRG Reclassifications

1. General

2. Yearly Review for Making DRG Changes; Request for Public Comment

3. Pre-MDC: Intestinal Transplantation

4. MDC 1 (Diseases and Disorders of the Nervous System)

a. Strokes

b. Unruptured Cerebral Aneurysms

5. MDC 5 (Diseases and Disorders of the Circulatory System)

a. Severity Adjusted Cardiovascular Procedures

b. Automatic Implantable Cardioverter/Defibrillator

c. Coronary Artery Stents

d. Insertion of Left Atrial Appendage Device

e. External Heart Assist System Implant

f. Carotid Artery Stent

g. Extracorporeal Membrane Oxygenation (ECMO)

6. MDC 6 (Diseases and Disorders of the Digestive System): Artificial Anal Sphincter

7. MDC 8 (Diseases and Disorders of the Musculoskeletal System and Connective Tissue)

a. Hip and Knee Replacements

b. Kyphoplasty

c. Multiple Level Spinal Fusion

d. Charite(tm) Spinal Disc Replacement Device

8. MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)): Severe Sepsis

9. MDC 20 (Alcohol/Drug Use and Alcohol/Drug Induced Organic Mental Disorders): Drug-Induced Dementia

10. Medicare Code Editor (MCE) Changes

a. Newborn Age Edit

b. Newborn Diagnoses Edit

c. Diagnoses Allowed for ``Males Only'' Edit

d. Tobacco Use Disorder Edit

e. Noncovered Procedure Edit

11. Surgical Hierarchies

12. Refinement of Complications and Comorbidities (CC) List

a. Background

b. Comprehensive Review of the CC List

c. CC Exclusion List for FY 2006

13. Review of Procedure Codes in DRGs 468, 476, and 477

a. Moving Procedure Codes from DRG 468 or DRG 477 to MDCs

b. Reassignment of Procedures among DRGs 468, 476, and 477

c. Adding Diagnosis or Procedure Codes to MDCs

14. Changes to the ICD-9-CM Coding System

15. Other Issues

a. Acute Intermittent Porphyria

b. Prosthetic Cardiac Support Device (Code 37.41)

c. Coronary Intravascular Ultrasound (IVUS) (Procedure Code 00.24)

d. Islet Cell Transplantation

C. Recalibration of DRG Weights

D. LTC-DRG Reclassifications and Relative Weights for LTCHs for FY 2006

1. Background

2. Changes in the LTC-DRG Classifications

a. Background

b. Patient Classifications into DRGs

3. Development of the Proposed FY 2006 LTC-DRG Relative Weights

a. General Overview of Development of the LTC-DRG Relative Weights

b. Data

c. Hospital-Specific Relative Value Methodology

d. Low-Volume LTC-DRGs

4. Steps for Determining the FY 2006 LTC-DRG Relative Weights

5. Other Public Comments Relating to the LTCH PPS Payment Policies

E. Add-On Payments for New Services and Technologies

1. Background

2. FY 2006 Status of Technology Approved for FY 2005 Add-On Payments

3. Reevaluation of FY 2005 Applications That Were Not Approved

4. FY 2006 Applicants for New Technology Add-On Payments III. Changes to the Hospital Wage Index

A. Background

B. Core-Based Statistical Areas for the Hospital Wage Index

C. Occupational Mix Adjustment to FY 2006 Index

1. Development of Data for the Occupational Mix Adjustment

2. Calculation of the Occupational Mix Adjustment Factor and the Occupational Mix Adjusted Wage Index

D. Worksheet S-3 Wage Data for the FY 2006 Wage Index Update

E. Verification of Worksheet S-3 Wage Data

F. Computation of the FY 2006 Unadjusted Wage Index

G. Computation of the FY 2006 Blended Wage Index

H. Revisions to the Wage Index Based on Hospital Redesignation

[[Page 47280]]

1. General

2. Effects of Reclassification

3. Application of Hold Harmless Protection for Certain Urban Hospitals Redesignated as Rural

4. FY 2006 MGCRB Reclassifications

5. FY 2006 Redesignations under Section 1886(d)(8)(B) of the Act

6. Reclassifications under Section 508 of Pub. L. 108-173

I. FY 2006 Wage Index Adjustment Based on Commuting Patterns of Hospital Employees

J. Requests for Wage Index Data Corrections IV. Rebasing and Revision of the Hospital Market Baskets

A. Background

B. Rebasing and Revising the Hospital Market Basket

1. Development of Cost Categories and Weights

2. PPS--Selection of Price Proxies

3. Labor-Related Share

C. Separate Market Basket for Hospitals and Hospital Units Excluded from the IPPS

1. Hospitals Paid Based on Their Reasonable Costs

2. Excluded Hospitals Paid Under Blend Methodology

3. Development of Cost Categories and Weights for the 2002-Based Excluded Hospital Market Basket

D. Frequency of Updates of Weights in IPPS Hospital Market Basket

E. Capital Input Price Index Section V. Other Decisions and Changes to the IPPS for Operating Costs and GME Costs

A. Postacute Care Transfer Payment Policy

1. Background

2. Changes to DRGs Subject to the Postacute Care Transfer Policy

B. Reporting of Hospital Quality Data for Annual Hospital Payment Update

1. Background

2. Requirements for Hospital Reporting of Quality Data

C. Sole Community Hospitals and Medicare Dependent Hospitals

1. Background

2. Budget Neutrality Adjustment to Hospital Payments Based on Hospital-Specific Rate

3. Technical Change

D. Rural Referral Centers

1. Case-Mix Index

2. Discharges

3. Technical Change

E. Payment Adjustment for Low-Volume Hospitals

F. Indirect Medical Education (IME) Adjustment

1. Background

2. IME Adjustment for IPPS-Excluded Hospitals Converting to IPPS Hospitals

3. Section 1886(d)(3)(E) Teaching Hospitals That Withdraw Rural Reclassification

G. Payment to Disproportionate Share Hospitals (DSHs)

1. Background

2. Implementation of Section 951 of Pub. L. 108-173

3. Calculation of the Medicare Fraction

4. Calculation of the Medicaid Fraction

H. Geographic Reclassifications

1. Background

2. Multicampus Hospitals

3. Urban Group Hospital Reclassifications

4. Clarification of Goldsmith Modification Criterion for Urban Hospitals Seeking Reclassification as Rural

I. Payment for Direct Graduate Medical Education

1. Background

2. Direct GME Initial Residency Period

a. Background

b. Direct GME Initial Residency Period Limitation: Simultaneous Match

3. New Teaching Hospitals' Participation in Medicare GME Affiliated Groups

4. GME FTE Cap Adjustments for Rural Hospitals

5. Technical Changes: Cross-References

J. Provider-Based Status of Facilities under Medicare

1. Background

2. Limits on Scope of Provider-Based Regulations--Facilities for Which Provider-Based Determinations Will Not Be Made

3. Location Requirement for Off-Campus Facilities: Application to Certain Neonatal Intensive Care Units

4. Technical and Clarifying Changes

K. Rural Community Hospital Demonstration Program

L. Definition of a Hospital in Connection with Specialty Hospitals VI. PPS for Capital-Related Costs VII. Changes for Hospitals and Hospital Units Excluded From the IPPS

A. Payments to Excluded Hospitals and Hospital Units

1. Payments to Existing Excluded Hospitals and Hospital Units

2. Updated Caps for New Excluded Hospitals and Units

3. Implementation of a PPS for IRFs

4. Implementation of a PPS for LTCHs

5. Implementation of a PPS for IPFs

6. Report of Adjustment (Exception) Payments

B. Critical Access Hospitals (CAHs)

1. Background

2. Policy Change Relating to Continued Participation by CAHs in Lugar Counties

3. Policy Change Relating to Designation of CAHs as Necessary Providers

a. Determination of the Relocation Status of a CAH

b. Relocation of a CAH Using a Waiver To Meet the CoP for Distance VIII. Payment for Blood Clotting Factor Administered to Hemophilia Inpatients IX. MedPAC Recommendations

A. Medicare Payment Policy

1. Update Factor

2. Quality Incentive Payment Policy

3. Refinement of DRGs Based on Severity of Illness

4. APR-DRGs

5. DRG Relative Weights

6. High-Cost Outliers

B. Other MedPAC Recommendations X. Other Required Information

A. Requests for Data From the Public

B. Collection of Information Requirements

Regulation Text

Addendum--Schedule of Standardized Amounts Effective with Discharges Occurring On or After October 1, 2005 and Update Factors and Rate-of- Increase Percentages Effective With Cost Reporting Periods Beginning On or After October 1, 2005

I. Summary and Background II. Changes to Prospective Payment Rates for Hospital Inpatient Operating Costs for FY 2006

A. Calculation of the Adjusted Standardized Amount

1. Standardization of Base-Year Costs or Target Amounts

2. Computing the Average Standardized Amount

3. Updating the Average Standardized Amount

4. Other Adjustments to the Average Standardized Amount

a. Recalibration of DRG Weights and Updated Wage Index--Budget Neutrality Adjustment

b. Reclassified Hospitals--Budget Neutrality Adjustment

c. Outliers

d. Rural Community Hospital Demonstration Program Adjustment (Section 410A of Pub. L. 108-173)

5. FY 2006 Standardized Amount

B. Adjustments for Area Wage Levels and Cost-of-Living

1. Adjustment for Area Wage Levels

2. Adjustment for Cost-of-Living in Alaska and Hawaii

C. DRG Relative Weights

D. Calculation of Prospective Payment Rates for FY 2006

1. Federal Rate

2. Hospital-Specific Rate (Applicable Only to SCHs and MDHs)

a. Calculation of Hospital-Specific Rate

b. Updating the FY 1982, FY 1987, and FY 1996 Hospital-Specific Rates for FY 2006

3. General Formula for Calculation of Prospective Payment Rates for Hospitals Located in Puerto Rico Beginning On or After October 1, 2005 and Before October 1, 2006

a. Puerto Rico Rate

b. National Rate III. Changes to Payment Rates for Acute Care Hospital Inpatient Capital-Related Costs for FY 2006

A. Determination of Federal Hospital Inpatient Capital-Related Prospective Payment Rate Update

1. Capital Standard Federal Rate Update

a. Description of the Update Framework

b. Comparison of CMS and MedPAC Update Recommendation

2. Outlier Payment Adjustment Factor

3. Budget Neutrality Adjustment Factor for Changes in DRG Classifications and Weights and the Geographic Adjustment Factor

4. Exceptions Payment Adjustment Factor

5. Capital Standard Federal Rate for FY 2006

6. Special Capital Rate for Puerto Rico Hospitals

B. Calculation of Inpatient Capital-Related Prospective Payments for FY 2006

C. Capital Input Price Index

1. Background

[[Page 47281]]

2. Forecast of the CIPI for FY 2006 IV. Changes to Payment Rates for Excluded Hospitals and Hospital Units: Rate-of-Increase Percentages

A. Payments to Existing Excluded Hospitals and Units

B. Updated Caps for New Excluded Hospitals and Units V. Payment for Blood Clotting Factor Administered to Hemophilia Inpatients

Tables

Table 1A--National Adjusted Operating Standardized Amounts, Labor/ Nonlabor (69.7 Percent Labor Share/30.3 Percent Nonlabor Share If Wage Index Is Greater Than 1) Table 1B--National Adjusted Operating Standardized Amounts, Labor/ Nonlabor (62 Percent Labor Share/38 Percent Nonlabor Share If Wage Index Is Less Than or Equal to 1) Table 1C--Adjusted Operating Standardized Amounts for Puerto Rico, Labor/Nonlabor Table 1D--Capital Standard Federal Payment Rate Table 2--Hospital Case-Mix Indexes for Discharges Occurring in Federal Fiscal Year 2004; Hospital Wage Indexes for Federal Fiscal Year 2006; Hospital Average Hourly Wage for Federal Fiscal Years 2004 (2000 Wage Data), 2005 (2001 Wage Data), and 2006 (2002 Wage Data); Wage Indexes and 3-Year Average of Hospital Average Hourly Wages Table 3A--FY 2006 and 3-Year Average Hourly Wage for Urban Areas by CBSA Table 3B--FY 2006 and 3-Year Average Hourly Wage for Rural Areas by CBSA Table 4A--Wage Index and Capital Geographic Adjustment Factor (GAF) for Urban Areas by CBSA Table 4B--Wage Index and Capital Geographic Adjustment Factor (GAF) for Rural Areas by CBSA Table 4C--Wage Index and Capital Geographic Adjustment Factor (GAF) for Hospitals That Are Reclassified by CBSA Table 4F--Puerto Rico Wage Index and Capital Geographic Adjustment Factor (GAF) by CBSA Table 4J--Out-Migration Wage Adjustment--FY 2006 Table 5--List of Diagnosis-Related Groups (DRGs), Relative Weighting Factors, and Geometric and Arithmetic Mean Length of Stay (LOS) Table 6A--New Diagnosis Codes Table 6B--New Procedure Codes Table 6C--Invalid Diagnosis Codes Table 6D--Invalid Procedure Codes Table 6E--Revised Diagnosis Code Titles Table 6F--Revised Procedure Code Titles Table 6G--Additions to the CC Exclusions List Table 6H--Deletions from the CC Exclusions List Table 7A--Medicare Prospective Payment System Selected Percentile Lengths of Stay [FY 2004 MedPAR Update March 2005 GROUPER V22.0] Table 7B--Medicare Prospective Payment System Selected Percentile Lengths of Stay: [FY 2004 MedPAR Update March 2005 GROUPER V23.0] Table 8A--Statewide Average Operating Cost-to-Charge Ratios-July 2005 Table 8B--Statewide Average Capital Cost-to-Charge Ratios-July 2005 Table 9A--Hospital Reclassifications and Redesignations by Individual Hospital and CBSA--FY 2006 Table 9B--Hospital Reclassifications and Redesignation by Individual Hospital Under Section 508 of Pub. L. 108-173--FY 2006 Table 9C--Hospitals Redesignated as Rural under Section 1886(d)(8)(E) of the Act--FY 2006 Table 10--Geometric Mean Plus the Lesser of .75 of the National Adjusted Operating Standardized Payment Amount (Increased to Reflect the Difference Between Costs and Charges) or .75 of One Standard Deviation of Mean Charges by Diagnosis-Related Groups (DRGs)--July 2005 Table 11--FY 2006 LTC-DRGs, Relative Weights, Geometric Average Length of Stay, and 5/6ths of the Geometric Average Length of Stay Appendix A--Regulatory Impact Analysis Appendix B--Recommendation of Update Factors for Operating Cost Rates of Payment for Inpatient Hospital Services

I. Background

A. Summary

1. Acute Care Hospital Inpatient Prospective Payment System (IPPS)

Section 1886(d) of the Social Security Act (the Act) sets forth a system of payment for the operating costs of acute care hospital inpatient stays under Medicare Part A (Hospital Insurance) based on prospectively set rates. Section 1886(g) of the Act requires the Secretary to pay for the capital-related costs of hospital inpatient stays under a prospective payment system (PPS). Under these PPSs, Medicare payment for hospital inpatient operating and capital-related costs is made at predetermined, specific rates for each hospital discharge. Discharges are classified according to a list of diagnosis- related groups (DRGs).

The base payment rate is comprised of a standardized amount that is divided into a labor-related share and a nonlabor-related share. The labor-related share is adjusted by the wage index applicable to the area where the hospital is located; and if the hospital is located in Alaska or Hawaii, the nonlabor-related share is adjusted by a cost-of- living adjustment factor. This base payment rate is multiplied by the DRG relative weight.

If the hospital treats a high percentage of low-income patients, it receives a percentage add-on payment applied to the DRG-adjusted base payment rate. This add-on payment, known as the disproportionate share hospital (DSH) adjustment, provides for a percentage increase in Medicare payments to hospitals that qualify under either of two statutory formulas designed to identify hospitals that serve a disproportionate share of low-income patient. For qualifying hospitals, the amount of this adjustment may vary based on the outcome of the statutory calculations.

If the hospital is an approved teaching hospital, it receives a percentage add-on payment for each case paid under the IPPS (known as the indirect medical education (IME) adjustment). This percentage varies, depending on the ratio of residents to beds.

Additional payments may be made for cases that involve new technologies or medical services that have been approved for special add-on payments. To qualify, a new technology or medical service must demonstrate that it is a substantial clinical improvement over technologies or services otherwise available, and that, absent an add- on payment, it would be inadequately paid under the regular DRG payment.

The costs incurred by the hospital for a case are evaluated to determine whether the hospital is eligible for an additional payment as an outlier case. This additional payment is designed to protect the hospital from large financial losses due to unusually expensive cases. Any outlier payment due is added to the DRG-adjusted base payment rate, plus any DSH, IME, and new technology or medical service add-on adjustments.

Although payments to most hospitals under the IPPS are made on the basis of the standardized amounts, some categories of hospitals are paid the higher of a hospital-specific rate based on their costs in a base year (the higher of FY 1982, FY 1987, or FY 1996) or the IPPS rate based on the standardized amount. For example, sole community hospitals (SCHs) are the sole source of care in their areas, and Medicare- dependent, small rural hospitals (MDHs) are a major source of care for Medicare beneficiaries in their areas. Both of these categories of hospitals are afforded this special payment protection in order to maintain access to services for beneficiaries. (An MDH receives only 50 percent of the difference between the IPPS rate and its hospital- specific rates if the hospital-specific rate is higher than the IPPS rate. In addition, an MDH does not have the option of using FY 1996 as the base year for its hospital-specific rate.)

Section 1886(g) of the Act requires the Secretary to pay for the capital-related costs of inpatient hospital services ``in accordance with a prospective payment system established by the Secretary.'' The basic methodology for determining

[[Page 47282]]

capital prospective payments is set forth in our regulations at 42 CFR 412.308 and 412.312. Under the capital PPS, payments are adjusted by the same DRG for the case as they are under the operating IPPS. Similar adjustments are also made for IME and DSH as under the operating IPPS. In addition, hospitals may receive an outlier payment for those cases that have unusually high costs.

The existing regulations governing payments to hospitals under the IPPS are located in 42 CFR Part 412, Subparts A through M. 2. Hospitals and Hospital Units Excluded From the IPPS

Under section 1886(d)(1)(B) of the Act, as amended, certain specialty hospitals and hospital units are excluded from the IPPS. These hospitals and units are: Psychiatric hospitals and units; rehabilitation hospitals and units; long-term care hospitals (LTCHs); children's hospitals; and cancer hospitals. Various sections of the Balanced Budget Act of 1997 (Pub. L. 105-33), the Medicare, Medicaid and SCHIP [State Children's Health Insurance Program] Balanced Budget Refinement Act of 1999 (Pub. L. 106-113), and the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000 (Pub. L. 106- 554) provide for the implementation of PPSs for rehabilitation hospitals and units (referred to as inpatient rehabilitation facilities (IRFs)), psychiatric hospitals and units (referred to as inpatient psychiatric facilities (IPFs)), and LTCHs, as discussed below. Children's hospitals and cancer hospitals continue to be paid under reasonable cost-based reimbursement.

The existing regulations governing payments to excluded hospitals and hospital units are located in 42 CFR parts 412 and 413. a. IRFs

Under section 1886(j) of the Act, as amended, rehabilitation hospitals and units (IRFs) have been transitioned from payment based on a blend of reasonable cost reimbursement subject to a hospital-specific annual limit under section 1886(b) of the Act and the adjusted facility Federal prospective payment rate for cost reporting periods beginning on or after January 1, 2002 through September 30, 2002, to payment at 100 percent of the Federal rate effective for cost reporting periods beginning on or after October 1, 2002 (66 FR 41316, August 7, 2001; 67 FR 49982, August 1, 2002; 68 FR 45674, August 1, 2003, and 69 FR 45721, July 30, 2004). The existing regulations governing payments under the IRF PPS are located in 42 CFR part 412, subpart P. b. LTCHs

Under the authority of sections 123(a) and (c) of Pub. L. 106-113 and section 307(b)(1) of Pub. L. 106-554, LTCHs are being transitioned from being paid for inpatient hospital services based on a blend of reasonable cost-based reimbursement under section 1886(b) of the Act to 100 percent of the Federal rate during a 5-year period, beginning with cost reporting periods that start on or after October 1, 2002. For cost reporting periods beginning on or after October 1, 2006, LTCHs will be paid 100 percent of the Federal rate (LTCH PPS final rule (70 FR 24168)). LTCHs not meeting the definition in Sec. 412.23(e)(4) of the regulations may elect to be paid based on 100 percent of the Federal rate instead of a blended payment in any year during the 5-year transition period. LTCHs meeting the definition in Sec. 412.23(e)(4) will be paid based on 100 percent of the standard Federal rate. The existing regulations governing payment under the LTCH PPS are located in 42 CFR part 412, subpart O. c. IPFs

Under the authority of sections 124(a) and (c) of Pub. L. 106-113, inpatient psychiatric facilities (IPFs) (formerly psychiatric hospitals and psychiatric units of acute care hospitals) are paid under the new IPF PPS. Under the IPF PPS, some IPFs are transitioning from being paid for inpatient hospital services based on a blend of reasonable cost- based payment and a Federal per diem payment rate, effective for cost reporting periods beginning on or after January 1, 2005 (November 15, 2004 IPF PPS final rule (69 FR 66921)). For cost reporting periods beginning on or after January 1, 2008, IPFs will be paid 100 percent of the Federal per diem payment amount. The existing regulations governing payment under the IPF PPS are located in 42 CFR 412, subpart N. 3. Critical Access Hospitals (CAHs)

Under sections 1814, 1820, and 1834(g) of the Act, payments are made to critical access hospitals (CAHs) (that is, rural hospitals or facilities that meet certain statutory requirements) for inpatient and outpatient services based on 101 percent of reasonable cost. Reasonable cost is determined under the provisions of section 1861(v)(1)(A) of the Act and existing regulations under 42 CFR parts 413 and 415. 4. Payments for Graduate Medical Education (GME)

Under section 1886(a)(4) of the Act, costs of approved educational activities are excluded from the operating costs of inpatient hospital services. Hospitals with approved graduate medical education (GME) programs are paid for the direct costs of GME in accordance with section 1886(h) of the Act; the amount of payment for direct GME costs for a cost reporting period is based on the hospital's number of residents in that period and the hospital's costs per resident in a base year. The existing regulations governing payments to the various types of hospitals are located in 42 CFR part 413.

B. Summary of the Provisions of the FY 2006 IPPS Proposed Rule

In the FY 2006 IPPS proposed rule (70 FR 23306), we set forth proposed changes to the Medicare IPPS for operating costs and for capital-related costs in FY 2006. We also set forth proposed changes relating to payments for GME costs, payments to certain hospitals and units that continue to be excluded from the IPPS and paid on a reasonable cost basis, payments for DSHs, and requirements and payments for CAHs. The changes were proposed to be effective for discharges occurring on or after October 1, 2005, unless otherwise noted.

The following is a summary of the major changes that we proposed and the issues we addressed in the FY 2006 IPPS proposed rule. 1. Changes to the DRG Reclassifications and Recalibrations of Relative Weights

As required by section 1886(d)(4)(C) of the Act, we proposed annual adjustments to the DRG classifications and relative weights. Based on analyses of Medicare claims data, we proposed to establish a number of new DRGs and make changes to the designation of diagnosis and procedure codes under other existing DRGs.

We also presented analysis of FY 2006 applicants for add-on payments for high-cost new medical services and technologies (including public input, as directed by Pub. L. 108-173, obtained in a town hall meeting).

We proposed the annual update of the long-term care diagnosis- related group (LTC-DRG) classifications and relative weights for use under the LTCH PPS for FY 2006. 2. Changes to the Hospital Wage Index

We proposed revisions to the wage index and the annual update of the wage data. Specific issues addressed included the following:

[[Page 47283]]

The FY 2006 wage index update, using wage data from cost reporting periods that began during FY 2002.

The occupational mix adjustment to the wage index that we began to apply effective October 1, 2004.

The revisions to the wage index based on hospital redesignations and reclassifications.

The adjustment to the wage index for FY 2006 based on commuting patterns of hospital employees who reside in a county and work in a different area with a higher wage index.

The timetable for reviewing and verifying the wage data that were in effect for the FY 2006 wage index. 3. Revision and Rebasing of the Hospital Market Baskets

We proposed rebasing and revising the hospital operating and capital market baskets to be used in developing the FY 2006 update factor for the operating prospective payment rates and the excluded hospital market basket to be used in developing the FY 2006 update factor for the excluded hospital rate-of-increase limits. We also set forth the data sources used to determine the proposed revised market basket relative weights and choice of price proxies. 4. Other Decisions and Changes to the PPS for Inpatient Operating and GME Costs

In the proposed rule, we discussed a number of provisions of the regulations in 42 CFR parts 412 and 413 and set forth proposed changes concerning the following:

Solicitation of public comments on two options for possible expansion of the current postacute care transfer policy.

The reporting of hospital quality data as a condition for receiving the full annual payment update increase.

Changes in the application of the budget neutrality adjustment to MDHs and SCHs for computing the hospital-specific rate.

Updated national and regional case-mix values and discharges for purposes of determining rural referral center status.

The payment adjustment for low-volume hospitals.

The IME adjustment for TEFRA hospitals that are converting to IPPS hospitals, and IME FTE resident caps for urban hospitals that are granted rural reclassification and then withdraw that rural classification.

Changes to implement section 951 of Pub. L. 108-173 relating to the provision of patient stay days/SSI data maintained by CMS to hospitals for the purpose of determining their DSH percentage.

Changes relating to hospitals' geographic classifications, including multicampus hospitals and urban group hospital reclassifications.

Changes and clarifications relating to GME, including GME initial residency period limitation, new teaching hospitals' participation in Medicare GME affiliated groups, and the GME FTE cap adjustment for rural hospitals;

Solicitation of public comments on possible changes in requirements for provider-based entities relating to the location requirements for certain neonatal intensive care units as off-campus facilities;

Discussion of the second year of implementation of the Rural Community Hospital Demonstration Program; and

Clarification of the definition of a hospital as it relates to ``specialty hospitals'' participating in the Medicare program. 5. PPS for Capital-Related Costs

In the proposed rule, we did not propose any policy changes to the capital-related prospective payment system. For the readers' benefit, we discussed the payment policy requirements for capital-related costs and capital payments to hospitals. 6. Changes for Hospitals and Hospital Units Excluded from the IPPS

In the proposed rule, we discussed the proposed revisions and clarifications concerning excluded hospitals and hospital units, proposed policy changes relating to continued participation by CAHs located in counties redesignated under section 1886(d)(8)(B) of the Act (Lugar counties), and proposed policy changes relating to designation of CAHs as necessary providers. 7. Changes in Payment for Blood Clotting Factor

In the proposed rule, we discussed the proposed change in payment for blood clotting factor administered to inpatients with hemophilia for FY 2006. 8. Determining Prospective Payment Operating and Capital Rates and Rate-of-Increase Limits

In the Addendum to the proposed rule, we set forth proposed changes to the amounts and factors for determining the FY 2006 prospective payment rates for operating costs and capital-related costs. We also established the proposed threshold amounts for outlier cases. In addition, we addressed the proposed update factors for determining the rate-of-increase limits for cost reporting periods beginning in FY 2006 for hospitals and hospital units excluded from the PPS. 9. Impact Analysis

In Appendix A of the proposed rule, we set forth an analysis of the impact that the proposed changes would have on affected hospitals. 10. Recommendation of Update Factor for Hospital Inpatient Operating Costs

In Appendix B of the proposed rule, as required by sections 1886(e)(4) and (e)(5) of the Act, we provided our recommendations of the appropriate percentage changes for FY 2006 for the following:

A single average standardized amount for all areas for hospital inpatient services paid under the IPPS for operating costs (and hospital-specific rates applicable to SCHs and MDHs).

Target rate-of-increase limits to the allowable operating costs of hospital inpatient services furnished by hospitals and hospital units excluded from the IPPS. 11. Discussion of Medicare Payment Advisory Commission Recommendations

Under section 1805(b) of the Act, the Medicare Payment Advisory Commission (MedPAC) is required to submit a report to Congress, no later than March 1 of each year, in which MedPAC reviews and makes recommendations on Medicare payment policies. MedPAC's March 2005 recommendation concerning hospital inpatient payment policies addressed only the update factor for inpatient hospital operating costs and capital-related costs under the IPPS and for hospitals and distinct part hospital units excluded from the IPPS. This recommendation is addressed in Appendix B of the proposed rule. MedPAC issued a second Report to Congress: Physician-Owned Specialty Hospitals, March 2005, which addressed other issues relating to Medicare payments to hospitals for inpatient services. The recommendations on these issues from this second report were addressed in section IX. of the preamble of the proposed rule. For further information relating specifically to the MedPAC March 2005 reports or to obtain a copy of the reports, contact MedPAC at (202) 220-3700 or visit MedPAC's Web site at: http://www.medpac.gov .

[[Page 47284]]

C. Public Comments Received in Response to the FY 2006 IPPS Proposed Rule

We received over 2,000 timely items of correspondence containing multiple comments on the FY 2006 IPPS proposed rule. Summaries of the public comments and our responses to those comments are set forth below under the appropriate heading.

II. Changes to DRG Classifications and Relative Weights

A. Background

Section 1886(d) of the Act specifies that the Secretary shall establish a classification system (referred to as DRGs) for inpatient discharges and adjust payments under the IPPS based on appropriate weighting factors assigned to each DRG. Therefore, under the IPPS, we pay for inpatient hospital services on a rate per discharge basis that varies according to the DRG to which a beneficiary's stay is assigned. The formula used to calculate payment for a specific case multiplies an individual hospital's payment rate per case by the weight of the DRG to which the case is assigned. Each DRG weight represents the average resources required to care for cases in that particular DRG, relative to the average resources used to treat cases in all DRGs.

Congress recognized that it would be necessary to recalculate the DRG relative weights periodically to account for changes in resource consumption. Accordingly, section 1886(d)(4)(C) of the Act requires that the Secretary adjust the DRG classifications and relative weights at least annually. These adjustments are made to reflect changes in treatment patterns, technology, and any other factors that may change the relative use of hospital resources. The changes to the DRG classification system and the recalibration of the DRG weights for discharges occurring on or after October 1, 2005, are discussed below. 1. General

Cases are classified into DRGs for payment under the IPPS based on the principal diagnosis, up to eight additional diagnoses, and up to six procedures performed during the stay. In a small number of DRGs, classification is also based on the age, sex, and discharge status of the patient. The diagnosis and procedure information is reported by the hospital using codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).

The process of forming the DRGs was begun by dividing all possible principal diagnoses into mutually exclusive principal diagnosis areas referred to as Major Diagnostic Categories (MDCs). The MDCs were formed by physician panels as the first step toward ensuring that the DRGs would be clinically coherent. The diagnoses in each MDC correspond to a single organ system or etiology and, in general, are associated with a particular medical specialty. Thus, in order to maintain the requirement of clinical coherence, no final DRG could contain patients in different MDCs. Most MDCs are based on a particular organ system of the body. For example, MDC 6 is Diseases and Disorders of the Digestive System. This approach is used because clinical care is generally organized in accordance with the organ system affected. However, some MDCs are not constructed on this basis because they involve multiple organ systems (for example, MDC 22 (Burns)). For FY 2005, cases are assigned to one of 520 DRGs in 25 MDCs. (We note that, in the FY 2006 proposed rule (70 FR 23313), we inadvertently stated that there were 519 DRGs.) The table below lists the 25 MDCs.

Major Diagnostic Categories (MDCs)

1 Diseases and Disorders of the Nervous System. 2 Diseases and Disorders of the Eye. 3 Diseases and Disorders of the Ear, Nose, Mouth, and Throat. 4 Diseases and Disorders of the Respiratory System. 5 Diseases and Disorders of the Circulatory System. 6 Diseases and Disorders of the Digestive System. 7 Diseases and Disorders of the Hepatobiliary System and Pancreas. 8 Diseases and Disorders of the Musculoskeletal System and Connective Tissue. 9 Diseases and Disorders of the Skin, Subcutaneous Tissue, and Breast. 10 Endocrine, Nutritional and Metabolic Diseases and Disorders. 11 Diseases and Disorders of the Kidney and Urinary Tract. 12 Diseases and Disorders of the Male Reproductive System. 13 Diseases and Disorders of the Female Reproductive System. 14 Pregnancy, Childbirth, and the Puerperium. 15 Newborns and Other Neonates with Conditions Originating in the Perinatal Period. 16 Diseases and Disorders of the Blood and Blood Forming Organs and Immunological Disorders. 17 Myeloproliferative Diseases and Disorders and Poorly Differentiated Neoplasms. 18 Infectious and Parasitic Diseases (Systemic or Unspecified Sites). 19 Mental Diseases and Disorders. 20 Alcohol/Drug Use and Alcohol/Drug Induced Organic Mental Disorders. 21 Injuries, Poisonings, and Toxic Effects of Drugs. 22 Burns. 23 Factors Influencing Health Status and Other Contacts with Health Services. 24 Multiple Significant Trauma. 25 Human Immunodeficiency Virus Infections.

In general, cases are assigned to an MDC based on the patient's principal diagnosis before assignment to a DRG. However, for FY 2005, there are nine DRGs to which cases are directly assigned on the basis of ICD-9-CM procedure codes. These DRGs are for heart transplant or implant of heart assist systems, liver and/or intestinal transplants, bone marrow, lung, simultaneous pancreas/kidney, and pancreas transplants and for tracheostomies. Cases are assigned to these DRGs before they are classified to an MDC. The table below lists the current nine pre-MDCs.

[[Page 47285]]

Pre-Major Diagnostic Categories (Pre-MDCs)

DRG 103.............................. Heart Transplant or Implant of Heart Assist System DRG 480.............................. Liver Transplant and/or Intestinal Transplant DRG 481.............................. Bone Marrow Transplant DRG 482.............................. Tracheostomy for Face, Mouth, and Neck Diagnoses DRG 495.............................. Lung Transplant DRG 512.............................. Simultaneous Pancreas/Kidney Transplant DRG 513.............................. Pancreas Transplant DRG 541.............................. Tracheostomy with Mechanical Ventilation 96+ Hours or Principal Diagnosis Except for Face, Mouth, and Neck Diagnosis with Major Operating Room Procedures DRG 542.............................. Tracheostomy with Mechanical Ventilation 96+ Hours or Principal Diagnosis Except for Face, Mouth, and Neck Diagnosis Without Major Operating Room Procedures

Once the MDCs were defined, each MDC was evaluated to identify those additional patient characteristics that would have a consistent effect on the consumption of hospital resources. Since the presence of a surgical procedure that required the use of the operating room would have a significant effect on the type of hospital resources used by a patient, most MDCs were initially divided into surgical DRGs and medical DRGs. Surgical DRGs are based on a hierarchy that orders operating room (O.R.) procedures or groups of O.R. procedures by resource intensity. Medical DRGs generally are differentiated on the basis of diagnosis and age (less than or greater than 17 years of age). Some surgical and medical DRGs are further differentiated based on the presence or absence of a complication or a comorbidity (CC).

Generally, nonsurgical procedures and minor surgical procedures that are not usually performed in an operating room are not treated as O.R. procedures. However, there are a few non-O.R. procedures that do affect DRG assignment for certain principal diagnoses, for example, extracorporeal shock wave lithotripsy for patients with a principal diagnosis of urinary stones.

Once the medical and surgical classes for an MDC were formed, each class of patients was evaluated to determine if complications, comorbidities, or the patient's age would consistently affect the consumption of hospital resources. Physician panels classified each diagnosis code based on whether the diagnosis, when present as a secondary condition, would be considered a substantial complication or comorbidity. A substantial complication or comorbidity was defined as a condition which, because of its presence with a specific principal diagnosis, would cause an increase in the length of stay by at least one day in at least 75 percent of the patients. Each medical and surgical class within an MDC was tested to determine if the presence of any substantial comorbidities or complications would consistently affect the consumption of hospital resources.

A patient's diagnosis, procedure, discharge status, and demographic information is fed into the Medicare claims processing systems and subjected to a series of automated screens called the Medicare Code Editor (MCE). The MCE screens are designed to identify cases that require further review before classification into a DRG.

After patient information is screened through the MCE and any further development of the claim is conducted, the cases are classified into the appropriate DRG by the Medicare GROUPER software program. The GROUPER program was developed as a means of classifying each case into a DRG on the basis of the diagnosis and procedure codes and, for a limited number of DRGs, demographic information (that is, sex, age, and discharge status).

After cases are screened through the MCE and assigned to a DRG by the GROUPER, the PRICER software calculates a base DRG payment. The PRICER calculates the payments for each case covered by the IPPS based on the DRG relative weight and additional factors associated with each hospital, such as IME and DSH adjustments. These additional factors increase the payment amount to hospitals above the base DRG payment.

The records for all Medicare hospital inpatient discharges are maintained in the Medicare Provider Analysis and Review (MedPAR) file. The data in this file are used to evaluate possible DRG classification changes and to recalibrate the DRG weights. However, in the July 30, 1999 IPPS final rule (64 FR 41500), we discussed a process for considering non-MedPAR data in the recalibration process. In order for us to consider using particular non-MedPAR data, we must have sufficient time to evaluate and test the data. The time necessary to do so depends upon the nature and quality of the non-MedPAR data submitted. Generally, however, a significant sample of the non-MedPAR data should be submitted by mid-October for consideration in conjunction with the next year's proposed rule. This allows us time to test the data and make a preliminary assessment as to the feasibility of using the data. Subsequently, a complete database should be submitted by early December for consideration in conjunction with the next year's proposed rule.

In the FY 2006 IPPS proposed rule (70 FR 23312), we proposed numerous changes to the DRG classification system for FY 2006 and to the methodology used to recalibrate the DRG weights. The changes we proposed to the DRG classification system, the public comments we received concerning the proposed changes, the final DRG changes, and the methodology used to recalibrate the DRG weights are set forth below. The changes we are implementing in this final rule will be reflected in the FY 2006 GROUPER, version 23.0, and are effective for discharges occurring on or after October 1, 2005. Unless otherwise noted in this final rule, our DRG analysis is based on data from the September 2004 update of the FY 2004 MedPAR file, which contains hospital bills received through September 30, 2004 for discharges in FY 2004. 2. Yearly Review for Making DRG Changes; Request for Public Comment

Many of the changes to the DRG classifications are the result of specific issues brought to our attention by interested parties. We encourage individuals with concerns about DRG classifications to bring those concerns to our attention in a timely manner so they can be carefully considered for possible inclusion in the next proposed rule and, if included, may be subjected to public review and comment. Therefore, similar to the timetable for interested parties to submit non-MedPAR data for consideration in the DRG recalibration process, concerns about DRG classification issues should be brought to our attention no later than early

[[Page 47286]]

December in order to be considered and possibly included in the next annual proposed rule updating the IPPS.

The actual process of forming the DRGs was, and continues to be, highly iterative, involving a combination of statistical results from test data combined with clinical judgment. In deciding whether to create a separate DRG, we consider whether the resource consumption and clinical characteristics of the patients with a given set of conditions are significantly different than the remaining patients in the DRG. We evaluate patient care costs using average charges and lengths of stay as proxies for costs and rely on the judgment of our medical officers to decide whether patients are distinct or clinically similar to other patients in the DRG. In evaluating resource costs, we consider both the absolute and percentage differences in average charges between the cases we are selecting for review and the remainder of cases in the DRG. We also consider variation in charges within these groups; that is, whether observed average differences are consistent across patients or attributable to cases that are extreme in terms of charges or length of stay, or both. Further, we also consider the number of patients who will have a given set of characteristics and generally prefer not to create a new DRG unless it will include a substantial number of cases. As we explain in more detail in section IX. of this preamble, MedPAC has made a number of recommendations regarding the DRG system.

To date, we have not used specific statistical standards as part of our guidelines for determining when DRG changes are warranted. However, we could potentially establish objective guidelines that are used in the DRG development process. For instance, such standards could include a minimum percentage or absolute difference in average charges or length of stay and number of cases in order for us to create a DRG or change the DRG assignment of a particular code or service. As part of our review and analysis of MedPAC's recommendations, we will consider whether to establish such guidelines for making DRG reclassification decisions. We welcome public comments on this issue. 3. Pre-MDC: Intestinal Transplantation

In the FY 2005 IPPS final rule (69 FR 48976), we moved intestinal transplantation cases that were assigned to ICD-9-CM procedure code 46.97 (Transplant of intestine) out of DRG 148 (Major Small and Large Bowel Procedures with CC) and DRG 149 (Major Small and Large Bowel Procedures Without CC) and into DRG 480 (Liver Transplant). We also changed the title for DRG 480 to ``Liver Transplant and/or Intestinal Transplant.'' We moved these cases out of DRGs 148 and 149 because our analysis demonstrated that the average charges for intestinal transplants are significantly higher than the average charges for other cases in these DRGs. We stated at that time that we would continue to monitor these cases.

Based on our review of the FY 2004 MedPAR data, we found 959 cases assigned to DRG 480 with overall average charges of approximately $165,622. There were only three cases involving an intestinal transplant alone and one case in which both an intestinal transplant and a liver transplant were performed. The average charges for the intestinal transplant cases ($138,922) were comparable to the average charges for the liver transplant cases ($165,314), while the remaining combination of an intestinal transplant and a liver transplant case had much higher charges ($539,841), and would be paid as an outlier case. Therefore, we did not propose any DRG modification for intestinal transplantation cases for FY 2006.

We note that an institution that performs intestinal transplantation, in correspondence to us written following the publication of the FY 2005 IPPS final rule, agreed with our decision to move cases assigned to code 46.97 to DRG 480.

Comment: Several commenters, including an institute that performs intestinal transplantation, supported our decision to reassign intestinal transplantation cases to DRG 480. One commenter commended CMS for its progress, but urged us to continue to evaluate a separate DRG for intestinal transplantation. While payment has improved, the commenter stated that it is still inadequate, and insufficient reimbursement could ultimately hinder beneficiary access to care.

Response: As indicated in the FY 2006 IPPS proposed rule (70 FR 23315), we found only three cases in the Medicare data that included an intestinal transplant. We found that the average charges were less for intestinal transplant cases ($138,922) than liver transplant cases ($165,314). Thus, even though we have a very low number of cases to make these comparisons, the data do not suggest that intestinal transplants are underpaid in DRG 480. We remain committed to assigning procedures to the most appropriate DRG based on clinical coherence and utilization of resources using the most recently available data. As we stated in the FY 2005 IPPS final rule (69 FR 48977), when we receive sufficient additional Medicare data on intestinal transplantation cases, we will again consider the DRG assignment for intestinal transplants.

Comment: One commenter concurred with the decision to assign intestinal transplant cases to DRG 480 but recommended that CMS create separate DRGs for liver-intestinal and liver-kidney transplants. The commenter requested that CMS report average charges for these cases in the final rule. The commenter noted that DRGs have been created for double organ transplants such as DRG 512 (Simultaneous Pancreas/Kidney Transplant).

Response: While the focus of our review in the proposed rule was limited to whether we should reassign intestinal transplants to DRG 480, we reviewed all cases in this DRG. Based on our review of the FY 2004 MedPAR data, the following table illustrates our findings:

Number of Average length Average DRG

cases

of stay

charges

DRG 480.........................................................

959

16.65

$165,622 Liver Transplantation...........................................

876

16.5

165,314 Intestinal Transplantation......................................

3

26.0

138,922 Liver-Intestinal Transplantation................................

1

72.0

539,841 Liver-Kidney Transplantation....................................

79

21.3

237,759

As we stated in the proposed rule (70 FR 23315), while the average charges and length of stay were much higher for the one liver- intestinal transplantation case, for which we had data, than the other cases in DRG 480, the case would likely be paid as an outlier. One case is insufficient to create a new DRG. Similarly, we are reluctant to create a

[[Page 47287]]

new DRG for such a small number of liver-kidney transplant cases, even though average charges and length of stay are higher for liver-kidney transplants than other cases in DRG 480. As discussed, in section IX.A. of this final rule, we plan in the next year to undertake a comprehensive review of the existing Medicare DRG system and expect to make changes to the DRGs to better reflect the severity of illness. As we study this issue, we will further analyze hospital costs for patients needing multiple organ transplants. At this time, we are not making any further modifications to the DRGs for multiple transplants in FY 2006. 4. MDC 1 (Diseases and Disorders of the Nervous System) a. Strokes

In 1996, the Food and Drug Administration (FDA) approved the use of tissue plasminogen activator (tPA), one type of thrombolytic agent that dissolves blood clots. In 1998, the ICD-9-CM Coordination and Maintenance Committee created code 99.10 (Injection or infusion of thrombolytic agent) in order to be able to uniquely identify the administration of thrombolytic agents. Studies have shown that tPA can be effective in reducing the amount of damage the brain sustains during an ischemic stroke, which is caused by blood clots that block blood flow to the brain. tPA is approved for patients who have blood clots in the brain, but not for patients who have a bleeding or hemorrhagic stroke. Thrombolytic therapy has been shown to be most effective when used within the first 3 hours after the onset of a stroke, and it is contraindicated in hemorrhagic stroke. The presence or absence of code 99.10 does not currently influence DRG assignment. Since code 99.10 became effective, CMS has been monitoring the DRGs and cases in which this code can be found, particularly with respect to cardiac and stroke DRGs.

Last year, CMS met with representatives from several hospital stroke centers who recommended modification of the existing stroke DRGs 14 (Intracranial Hemorrhage or Cerebral Infarction) and 15 (Nonspecific CVA and Precerebral Occlusion Without Infarction) by using the administration of tPA as a proxy to identify patients who have severe strokes. The representatives stated that using tPA as a proxy would help to identify patients who have strokes that are more severely and costly and would recognize the higher charges that these cases generate because of their higher hospital resource utilization. At that time, the presenters provided evidence that strokes where tPA was used were both more severe and more resource intensive. Specifically, they showed that patients who were given tPA for strokes had higher stroke severity scores at presentation, and that they were more expensive to care for because of increased intensive care unit monitoring requirements, increased diagnostic imaging costs, and increased laboratory and pharmacy costs. They also demonstrated that these patients had markedly better clinical outcomes. The stroke representatives made two suggestions concerning the stroke DRGs.

The first proposal suggested modifying DRG 14 by renaming it ``Ischemic Stroke Treatment with a Reperfusion Agent'', and including only those cases containing code 99.10. The remainder of stroke cases where the patient was not treated with a reperfusion agent would be included in DRG 15, renamed ``Hemorrhagic Stroke or Ischemic Stroke without a Reperfusion Agent''. Hemorrhagic stroke cases now found in DRG 14 that are not treated with a reperfusion agent would migrate to DRG 15.

The second suggestion was to leave DRGs 14 and 15 as they currently exist, and create a new DRG, with a recommended title ``Ischemic Stroke Treatment with a Reperfusion Agent''. This suggested DRG would include only cases where patients with strokes caused by arterial occlusion (or clot(s)) are also treated with tPA thrombolytic therapy.

We have examined the MedPAR data for the cases in DRGs 14 and 15. We divided the cases based on the presence of a principal diagnosis of hemorrhage or occlusive ischemia and the presence of procedure code 99.10. The following table displays the results:

Average length Average DRG

Count

of stay

charges

14--All Cases...................................................

221,879

5.67

$18,997 14--Cases with intracranial hemorrhage..........................

41,506

5.40

19,193 14--Cases with intracranial hemorrhage with code 99.10..........

61

7.4

37,045 14--Cases with intracranial hemorrhage without code 99.10.......

41,445

5.3

19,167 14--Cases without intracranial hemorrhage.......................

180,373

5.74

18,952 14--Cases without intracranial hemorrhage with code 99.10.......

2,085

7.20

35,128 14--Cases without intracranial hemorrhage without code 99.10....

178,288

5.72

18,763 15--All cases...................................................

71,335

4.53

14,382 15--Cases with intracranial hemorrhage..........................

0

0

0 15--Cases without intracranial hemorrhage.......................

71,335

4.53

14,382 15--Cases without intracranial hemorrhage with code 99.10.......

302

5.10

24,876 15--Cases without intracranial hemorrhage without code 99.10....

71,033

4.53

14,337

The above table shows that the average standardized charges for cases treated with a reperfusion agent are more than $16,000 and $10,000 higher than all other cases in DRGs 14 and 15, respectively. While these data suggest that patients treated with a reperfusion agent are more expensive than all other stroke patients, this conclusion is based on a small number of cases. In the FY 2006 IPPS proposed rule, we did not propose a change to the stroke DRGs because of the small number of reperfusion cases reported. However, we stated that we believe it is possible that more patients are being treated with a reperfusion agent than indicated by our data because the presence of code 99.10 does not affect DRG assignment and may be underreported.

In the FY 2006 IPPS proposed rule, we invited public comment on the changes to DRGs 14 and 15 suggested by the hospital representatives. In addition, we solicited public comment on the number of patients currently being treated with a reperfusion agent as well as the potential costs of these patients relative to others with strokes that are also included in DRGs 14 and 15.

Comment: Forty commenters supported the creation of a new DRG to recognize the group of patients who presented with stroke and who also received thrombolytic therapy. The commenters cited the following reasons for supporting this proposal: Increased costs of caring for these patients, specifically in intensive care unit, more

[[Page 47288]]

diagnostic imaging studies, and laboratory and pharmacy resources. In addition, the commenters noted that the proposal is also supported by evidence that patients receiving thrombolytic therapy have strokes of increased severity. The commenters also stated that the proposal demonstrates the need for hospitals to have an incentive to establish the infrastructure necessary to provide stroke patients with aggressive evaluation and management services, such as thrombolytic therapy, which have become the standard of care.

Response: We appreciate the commenters' responses in reply to our solicitation for public comment on the changes to DRGs 14 and 15 as suggested in the proposed rule. The level of detail provide in the responses helped us to formulate a change to the medical stroke DRGs. We agree with the commenters that there is an increased cost in caring for these patients including increased use of the intensive care unit, more diagnostic imaging studies, and laboratory and pharmacy resources. We also agree that--(1) the data indicate that patients receiving thrombolytic therapy have increased severity; and (2) reperfusion therapy is a good means to segregate these patients into a separate DRG.

Comment: One commenter encouraged CMS to modify DRGs 14 and 15 using one of two options. The first option would be to create DRG ``A'' where hemorrhagic and ischemic strokes were combined, but only supportive care was given, while DRG ``B'' would contain those hemorrhagic and ischemic stroke cases in which reperfusion or hemostatic agents were administered.

Alternatively, the commenter suggested that DRGs 14 and 15 could be modified by creating four new DRGs. DRG ``A'' would contain cases of hemorrhagic stroke and supportive care, DRG ``B'' would contain cases of hemorrhagic stroke treated with hemostatic agents, DRG ``C'' would contain cases of ischemic stroke and supportive care, and DRG ``D'' would contain cases of ischemic stroke treated with reperfusion agents.

Response: This commenter is suggesting that the DRG system recognize treatment of hemorrhage strokes with hemostatic agents as well as ischemic strokes with reperfusion agents. While we anticipate great industry strides in the treatment of stroke, currently no approved hemostatic agent is on the market. According to the manufacturer(s) of hemostatic agents, it is unlikely that these agents will be available for use during FY 2006. Therefore, we do not have any Medicare charge information that supports creating separate DRGs for hemorrhagic stroke patients treated with hemostatic agents as we do for ischemic stroke patients treated with thrombolytic therapy. When hemostatic agents are available on the market, we will reevaluate this issue.

Comment: One commenter believed that the two potential changes to the stroke DRGs as set forth in the proposed rule are too limited as written. The commenter believed that the descriptor, ``reperfusion agent'', is not broad enough to encompass other promising pharmacotherapies for stroke that are in late stages of clinical development. The commenter pointed out that these therapies include treatment for both ischemic stroke and hemorrhagic stroke. The commenter further noted that it is unlikely that any of the potential therapies will be approved for use during FY 2006. The commenter recommended that CMS broaden the title for the proposed new DRG to include a wider range of any newly approved therapies.

Response: While we look forward to improved therapies for treating patients with strokes, we are unable to create DRGs that recognize as yet unapproved treatment modalities. When the FDA has approved additional pharmaceuticals for the treatment of either ischemic or hemorrhagic stroke, we will evaluate the data and make DRG changes as appropriate. We point out that the DRG titles cannot possibly acknowledge all the codes located therein. The important part of the DRG is the structure of the logic; that is, what codes are assigned to the DRG.

Comment: One commenter recommended that CMS commit to creating a surgical DRG for ischemic stroke patients who are treated with surgical interventions. The commenter included several scenarios of possible diagnosis and procedure coding combinations that CMS could use to identify stroke cases and increase the scope of our analysis.

Response: Our goal was not to review all stroke cases within the MedPAR database, but to identify those cases in medical DRG 14, and possibly DRG 15, that might have included the administration of tPA as identified by procedure code 99.10. DRGs that identify a precise surgical procedure already exist; all of the combinations of procedure codes suggested by the commenter already appropriately group to DRGs within MDC 1.

Comment: One commenter stated that because code 99.10 was not reimbursable [did not have an impact on DRG assignment], hospital coders often did not use it. Some hospitals in which reperfusion therapy was commonplace never used this code.

Response: We would like to take this opportunity to reiterate that all cases should be accurately and completely coded, irrespective of the DRG implications of a specific code or codes. By coding accurately and completely, we will have more information on patient care costs for different services and treatments that better enable us to research further changes to the DRG system.

Comment: One commenter noted that, because only a single type of reperfusion agent is presently approved for stroke treatment, the proposed change would create a DRG that is, de facto, product specific. In addition, the commenter stated that the DRG change on which CMS requested comment would improve access to therapy for only a small fraction of all stroke patients. The commenter added that implementation of a narrowly-defined change [by creating a specific stroke-plus-tPA DRG] may necessitate further changes to the stroke DRGs in the near future to ensure patient access to emerging drug therapies once approved.

Response: While we did not propose a specific change to the stroke DRGs in the proposed notice, we have decided to modify the DRGs to distinguish those cases in which tPA is used as a treatment modality based on the strong support for this change voiced by commenters. When we reviewed the data represented in the above table, we noted that the average standardized charges for all cases in DRG 14 were $18,997, but that the subset of 2,085 cases in which tPA was used had average standardized charges of $35,128. We noted that the cases in DRG 14 without hemorrhage that did not report the use of tPA had average standardized charges of $18,763, which was comparable with the figures for all cases in the DRG. Given that these cases are easily identifiable through the use of procedure code 99.10, and that the average standardized charges are $16,131 higher for the cases using tPA, we decided to carve these cases out of the existing DRGs 14 and 15, and represent them in a new DRG. We are changing the structure of stroke DRGs not to award higher payment for a specific drug but to recognize the need for better overall care for this group of patients. Even though a tPA is indicated only for a small proportion of stroke patients (only those experiencing ischemic strokes treated within 3 hours of the onset of symptoms), our data suggest that there are enough patients to support the DRG change. While our goal

[[Page 47289]]

is to make payment relate more closely to resource use, we also note that use of a tPA in a carefully selected patient population will lead to better outcomes and overall care and may lessen the need for postacute care. With regard to the potential need to modify stroke DRGs in the future, we note that we perform an update to the DRGs and modify DRGs every year. We reiterate that should additional types of therapy be approved, we will evaluate them, and after judicious study, will make appropriate DRG title and/or logic changes as required.

In this final rule, after consideration of public comments received and based on our analysis of MedPAR data that supports the creation of a DRG that identifies embolic stroke combined with tPA treatment, we are creating new DRG 559 (Acute Ischemic Stroke with Use of Thrombolytic Agent). From a data consistency standpoint, we believe that adding a new DRG identifying these cases will be less disruptive to our stakeholders than creating three new DRGs, two of which would mimic existing DRGs 14 and 15. The GROUPER logic for DRGs 14 and 15 will not be affected by this change; that is, the GROUPER content of DRGs 14 and 15 will be the same in FY 2006 as it was in FY 2005. The structure of the new DRG 559 includes the following codes:

Principal Diagnosis

433.01, Occlusion and stenosis of basilar artery, with cerebral infarction

433.11, Occlusion and stenosis of carotid artery, with cerebral infarction

433.21, Occlusion and stenosis of vertebral artery, with cerebral infarction

433.31, Occlusion and stenosis of multiple and bilateral arteries, with cerebral infarction

433.81, Occlusion and stenosis of other specified precerebral artery, with cerebral infarction

433.91, Occlusion and stenosis of unspecified precerebral artery, with cerebral infarction

434.01, Cerebral thrombosis, with cerebral infarction

434.11, Cerebral embolism, with cerebral infarction

434.91, Cerebral artery occlusion, unspecified, with cerebral infarction

and

Nonoperating Room Procedure

99.10, Injection or infusion of thrombolytic agent

We will continue to monitor stroke DRGs in the future. As noted above, should treatment modalities change, we will be open to making changes to the DRG structure that will recognize improvements in treatment and technology. b. Unruptured Cerebral Aneurysms

In the FY 2004 IPPS final rule (68 FR 45353), we created DRG 528 (Intracranial Vascular Procedures With a Principal Diagnosis of Hemorrhage) in MDC 1. We received a comment at that time that suggested we create another DRG for intracranial vascular procedures for unruptured cerebral aneurysms. For the FY 2004 IPPS final rule (68 FR 45353) and the FY 2005 IPPS final rule (69 FR 48957), we evaluated the data for cases in the MedPAR file involving unruptured cerebral aneurysms assigned to DRG 1 (Craniotomy Age >17 With CC) and DRG 2 (Craniotomy Age >17 Without CC) and concluded that the average charges were consistent with those for other cases found in DRGs 1 and 2. Therefore, we did not propose a change to the DRG assignment for unruptured cerebral aneurysms.

We have reviewed data for unruptured cerebral aneurysms cases in DRGs 1 and 2. In our analysis of these FY 2004 MedPAR data, we found 1,136 unruptured cerebral aneurysm cases assigned to DRG 1 and 964 unruptured cerebral aneurysm cases assigned to DRG 2. Although the average charges for the unruptured cerebral aneurysm cases in DRG 1 ($53,455) and DRG 2 ($34,028) were slightly higher than the average charges for all cases in DRG 1 ($51,466) and DRG 2 ($30,346), we do not believe these differences are significant enough to warrant a change in these two DRGs at this time. Therefore, we did not propose a change in the structure of these DRGs relating to unruptured cerebral aneurysm cases for FY 2006.

Comment: Several commenters agreed that the minimal differences in charges for unruptured cerebral aneurysms cases compared to all cases assigned to DRGs 1 and 2 do not justify a change in the DRG assignment for these cases. One commenter stated that unruptured cerebral aneurysm cases should be reclassified into a new DRG. The commenter stated that a new DRG is warranted to understand the true weight of these procedures and to establish reimbursement that recognizes the cost of medical devices used to treat unruptured cerebral aneurysms.

Response: Our analysis is based on the most recent charge information available reflecting the overall resources used to treat unruptured cerebral aneurysms in Medicare patients. We concur with the commenters that there are minimal differences in the charges for the unruptured cerebral aneurysm cases compared to all cases assigned to DRGs 1 and 2 and that the results of the data do not justify creation of a new DRG. We believe that unruptured cerebral aneurysms are appropriately assigned to DRGs 1 and 2. Therefore, we are not making any modifications to the DRG assignment for unruptured cerebral aneurysms. 5. MDC 5 (Diseases and Disorders of the Circulatory System) a. Severity Adjusted Cardiovascular Procedures

In response to the FY 2006 IPPS proposed rule, one commenter noted that section 507(c) of Pub. L. 108-173 required MedPAC to conduct a study to determine how the DRG system should be updated to better reflect the cost of delivering care in a hospital setting. The commenter noted that MedPAC reported that the ``cardiac surgery DRGs had high relative profitability ratios.'' While the commenter noted that it may take time to conduct and complete a thorough evaluation of the MedPAC payment recommendations for all DRGs, the commenter strongly encouraged CMS to revise the cardiac DRGs through patient severity refinements as part of the final rule to be effective for FY 2006. In section IX.A. of the preamble to this final rule, we are responding in detail to this comment by making significant revisions to a number of cardiovascular DRGs that currently contain patients with a wide range of severity and resource consumption in order to reflect more accurately the resources required to care for different kinds of cardiovascular patients. Accordingly, in response to the issues raised by the commenter and as an interim step until we can complete a comprehensive review of MedPAC's recommendations, we are deleting current DRGs 107, 109, 111, 116, 478, 516, 517, 526, and 527, and creating new DRGs 547 through 558 in their place.

We received several comments on the FY 2006 IPPS proposed rule that recommended that we split additional cardiovascular DRGs based on the presence or absence of heart failure, acute myocardial infarction, and shock. As indicated in section IX.A. of this final rule, we conducted a focused review of a number of different cardiovascular DRGs and are making revisions to them based on a newly designated list of ``major cardiovascular conditions.''

We believe these new DRGs will help to address a number of the concerns raised by these commenters. We intend to monitor these DRGs carefully in upcoming years and welcome input regarding the success of these DRGs in

[[Page 47290]]

reflecting patient severity and resource use. b. Automatic Implantable Cardioverter/Defibrillator

As part of our annual review of DRGs, for FY 2006, we performed a review of cases in the FY 2004 MedPAR file involving the implantation of a defibrillator in the following DRGs:

DRG 515 (Cardiac Defibrillator Implant Without Cardiac Catheterization) DRG 535 (Cardiac Defibrillator Implant With Cardiac Catheterization With Acute Myocardial Infarction, Heart Failure, or Shock) DRG 536 (Cardiac Defibrillator Implant With Cardiac Catheterization Without Acute Myocardial Infarction, Heart Failure, or Shock)

While conducting our review, we noted that there had been considerable comments from hospital coders on code 37.26 (Cardiac electrophysiologic stimulation and recording studies (EPS)), which is included in these DRGs. These comments from hospital coders were directed to both CMS and the American Hospital Association. The procedure codes for these three DRGs describe the procedures that are considered to be a cardiac catheterization. Code 37.26 is classified as a cardiac catheterization within these DRGs. Therefore, the submission of code 37.26 affects the DRG assignment for defibrillator cases and leads to the assignment of DRGs 535 or 536. When a cardiac catheterization is performed, the case is assigned to DRGs 535 or 536, depending on whether or not the patient also had an acute myocardial infarction, heart failure, or shock. The following chart shows the number of cases in each DRG, along with their average length of stay and average charges, found in the data:

Number of Average length Average DRG

cases

of stay

charges

515.............................................................

25,236

4.32 $83,659.76 535.............................................................

12,118

8.27 113,175.43 536.............................................................

18,305

5.39 94,453.62

We have received a number of questions from hospital coders regarding the correct use of code 37.26. There is considerable confusion about whether or not code 37.26 should be reported when the procedure is performed as part of the defibrillator implantation. Currently, the ICD-9-CM instructs the coder not to report code 37.26 when a defibrillator is inserted. There is an inclusion term under the defibrillator code 37.94 (Implantation or replacement of automatic cardioverter/defibrillator, total system [AICD]) which states that EPS is included in code 37.94. We discussed modifying this instruction at the October 7-8, 2004 meeting of the ICD-9-CM Coordination and Maintenance Committee. We received a number of comments opposing a modification to the use of code 37.26 that would also allow it to be reported with an AICD insertion. A report of this meeting can be found on the Web site: http://www.cms.hhs.gov/paymentsystem/icd9.

We performed an analysis of cases within DRGs 535 and 536 with cardiac catheterization and with and without code 37.26 and with code 37.26 only reported without cardiac catheterization and found the following:

Number of Average length- Average DRG

cases

of-stay

charges

535--Cardiac Catheterization Without Code 37.26.................

5,060

10.63 $127,130.79 535--With Code 37.26 Only Without Cardiac Catheterization.......

5,264

5.61 98,900.13 535--With Cardiac Catheterization and Code 37.26................

1,794

9.44 115,701.09 536--Cardiac Catheterization Without Code 37.26.................

4,799

8.11 110,493.86 536--With Code 37.26 Only Without Cardiac Catheterization.......

10,829

3.85 85,390.88 536--With Cardiac Catheterization and Code 37.26................

2,677

6.76 102,359.21

The data show that when code 37.26 is the only procedure reported from the list of cardiac catheterizations, the average charges and the average length of stay are considerably lower. For example, the average standardized charges for a defibrillator implant with only an EPS are $85,390.88 in DRG 536, while the average standardized charges for DRG 536 with a cardiac catheterization, but not an EPS, are $110,493.86. The average standardized charges for all cases in DRG 536 are $94,453.62. The data show similar findings for DRG 535, with lower lengths of stay and average charges when the only code reported from the cardiac catheterization list is an EPS. When we also consider the acknowledged coding problems in the use of code 37.26, we believe it is inappropriate to base a defibrillator DRG assignment on the EPS code. Cases identified with this code capture patients who require less resource use than patients who have a cardiac catheterization.

Data reflected in the chart above show that the average standardized charges for DRG 515 were $83,659.76. These average charges are closer to those in DRG 536 with code 37.26 and without any other cardiac catheterization code reported. While the cases in DRG 535 with code 37.26 and without a cardiac catheterization have higher average charges than the average charges for cases in DRG 515, these cases have much lower average charges than the average charges for overall cases in DRG 535. For these reasons, we proposed to remove code 37.26 from the list of cardiac catheterizations for DRGs 535 and 536. If a defibrillator is implanted and an EPS is performed with no other type of cardiac catheterization, the case would be assigned to DRG 515.

CMS issued a National Coverage Determination for implantable cardioverter defibrillators, effective January 27, 2005, that expands coverage and requires, in certain cases, that patient data be reported when the defibrillator is implanted for the clinical indication of primary prevention of sudden cardiac death. The submission of data on patients receiving an implantable cardioverter defibrillator for primary prevention to a data collection system is needed for the determination that the implantable cardioverter defibrillator is reasonable and necessary and for quality improvement. These

[[Page 47291]]

data will be made available in some form to providers and practitioners to inform their decisions, monitor performance quality, and benchmark and identify best practices. We made a temporary registry available for use when the policy became effective and used the Quality Net Exchange for data submission because Medicare-participating hospitals already use the Exchange to report data.

We intend to transition from the temporary registry using the Quality Net Exchange to a more sophisticated follow-on registry that will have the ability to collect longitudinal data. Some providers have suggested that CMS increase reimbursement for implantable cardioverter defibrillators to compensate the provider for reporting data. ICD data reporting includes elements of patient demographics, clinical characteristics and indications, medications, provider information, and complications. Since these data elements are commonly found in patient medical records, it is CMS' expectation that these data are readily available to the individuals abstracting and reporting data. Therefore, we believe that increased reimbursement is not needed at this time.

Comment: One commenter stated that there has been considerable confusion surrounding the use of code 37.26. The commenter indicated that coders are unclear whether code 37.26 should be reported when an electrophysiologic study (EPS) is performed as part of a defibrillator implantation or only when defibrillator device checks are performed. The commenter pointed out that the continuing efforts of the Editorial Advisory Board for Coding Clinic to clarify the use of this code have led to changes in coding advice published in Coding Clinic for ICD-9-CM by the American Hospital Association. However, the commenter stated, while the change in coding advice was intended to clarify use of code 37.26, coders continue to have questions about it. The commenter supported our proposal to remove code 37.26 from the list of cardiac catheterizations for DRGs 535 and 536 and agreed with CMS' plans to continue working to clarify use of this code or modify the code through the ICD-9-CM Coordination and Maintenance Committee. The commenter suggested that, once the coding issues are resolved and consistent data are collected, CMS should reexamine the DRG assignment(s) for code 37.26.

Other commenters opposed our proposal to remove code 37.26 from DRGs 535 and 536. These commenters stated that code 37.26 is used to capture a variety of disparate procedures with varying purposes, sites of service, and intensity, and that the resultant data are not representative of any one of these. Other commenters stated that the code contains three separate procedures of varying intensity: Electrophysiology study, intraoperative device interrogation, and noninvasive programmed stimulation. Several commenters believed that the payment change would have a severe financial impact on their hospitals. They believed it is inappropriate to make the change without the data to justify the change. Several commenters stated that the change would have a significant impact on the use of CRT-D implants because the devices are more costly. The commenters suggested that, before considering a revision to DRGs 535 and 536 for code 37.26, CMS should resolve the coding confusion. The commenters asked that the code be discussed at the September 29, 2005 ICD-9-CM Coordination and Maintenance Committee meeting and suggested that separate codes be created for the different procedures currently captured by code 37.26. According to the commenters, the new codes that are created could go into effect on October 1, 2006. The commenters suggested that, once data are available, CMS should consider a revision to DRGs 535 and 536 for EPS procedures.

Response: We agree with the commenter that there is considerable confusion regarding the use of code 37.26. It is possible that code 37.26 is being used for a variety of electrophysiologic procedures such as EPS, noninvasive programmed electrical stimulation, and programmed electrical stimulation. However, as indicated in the proposed rule and above in this final rule, our data show that the cases coded with 37.26 that were not separately coded with a cardiac catheterization had average charges of $98,900.13 in DRG 535 and $85,390.88 in DRG 536 compared to $127,130.79 and $110,493.86, respectively, for all other cases in these DRGs. For this reason, we believe it is appropriate to include code 37.26 in DRG 515 and no longer assign it to DRGs 535 and 536 that are for patients who receive a cardiac catheterization.

As we discussed earlier in this section of the preamble, Medicare significantly expanded coverage of implantable defibrillators on January 27, 2005 (Pub. No. 100-3, section 20.4) to patients who have a prior history of heart disease but are not in acute heart failure. These prophylactic defibrillator implants are expected to significantly increase the number of patients in DRGs 515, 535, and 536. It is our experience that most of these patients will not be receiving a cardiac catheterization and will be less resource-intensive than the acute heart failure patients receiving an implantable defibrillator. We note that the Bernstein Research Call publication of April 27, 2005 stated that this DRG change could ``dampen the elective implantation of de- novo CRT-D or dual chamber devices into relatively stable patients.'' The article further states that CMS ``realizes that the new prophylactic ICD [implantable cardioverter defibrillators] eligibility requirements do not require an EP test, and that EP tests per se do not consume sufficient resources to justify the reimbursement differentials seen between DRGs 515 versus 535 and 536.'' We believe it is particularly important to make the change to DRGs 515, 535, and 536 at this time, given the expansion of Medicare coverage of implantable defibrillators and the evidence that suggests that patients who receive an EP test, but not a cardiac catheterization, are less expensive than other patients receiving these devices.

We will address code 37.26 at our September 29-30, 2005 meeting of the ICD-9-CM Coordination and Maintenance Committee meeting. The public is encouraged to participate in this meeting and offer suggestions for code modifications. Information on this meeting can be found at: http://www.cms.hhs.gov/paymentsystems/icd9 .

Comment: Five commenters stated that CMS' data show that the average charges for cases with code 37.26 are significantly higher than those in DRG 515. The commenters suggested that the volume of cases is significant enough to create a new DRG for cases with cardiac defibrillator implant without cardiac catheterization, but with code 37.26.

Response: Given the extensive comments concerning coding problems with code 37.26, we do not believe it is appropriate to create a new DRG that would specifically capture defibrillator implants with this code. Therefore, we are not creating the suggested new DRG at this time. As stated earlier, we will continue to work with the coding and health care community to modify code 37.26 so that it will lead to more consistent reporting. Once we have better data, we will evaluate additional DRG modifications.

After consideration of the public comments received on the proposed rule, in this final rule, we are implementing the modification of DRGs 535 and 536 as proposed for FY 2006. We are removing code 37.26 from the list of cardiac catheterizations for DRGs

[[Page 47292]]

535 and 536 and adding the code in DRG 515. c. Coronary Artery Stents

In the FY 2005 IPPS final rule (69 FR 48971 through 48974), we addressed two comments from industry representatives about the DRG assignments for coronary artery stents. These commenters had expressed concern about whether the reimbursement for stents is adequate, especially for insertion of multiple stents. They also expressed concern about whether the current DRG structure represents the most clinically coherent classification of stent cases. In the FY 2006 proposed rule (70 FR 23318 through 23319), we included the following discussion regarding the commenter's concerns:

The current DRG structure incorporates stent cases into the following two pairs of DRGs, depending on whether bare metal or drug-eluting stents are used and whether acute myocardial infarction (AMI) is present:

DRG 516 (Percutaneous Cardiovascular Procedures with AMI)

DRG 517 (Percutaneous Cardiovascular Procedures with Nondrug-Eluting Stent without AMI)

DRG 526 (Percutaneous Cardiovascular Procedures with Drug-Eluting Stent with AMI)

DRG 527 (Percutaneous Cardiovascular Procedures with Drug-Eluting Stent without AMI)

The commenters presented two recommendations for refinement and restructuring of the current coronary stent DRGs. One of the recommendations involved restructuring these DRGs to create two additional stent DRGs that are closely patterned after the existing pairs, and would reflect insertion of multiple stents with and without AMI. The commenters recommended incorporating either stenting code 36.06 (Insertion of nondrug-eluting coronary artery stent(s)) or code 36.07 (Insertion of drug-eluting coronary artery stent(s)) when they are reported along with code 36.05 (Multiple vessel percutaneous transluminal coronary angioplasty [PTCA] or coronary atherectomy performed during the same operation, with or without mention of thrombolytic agent). The commenter's first concern was that hospitals may be steering patients toward coronary artery bypass graft surgery in place of stenting in order to avoid significant financial losses due to what it considered the inadequate reimbursement for inserting multiple stents.

In our response to comments in the FY 2005 IPPS final rule, we indicated that it was premature to act on this recommendation because the current coding structure for coronary artery stents cannot distinguish cases in which multiple stents are inserted from those in which only a single stent is inserted. Current codes are able to identify performance of PTCA in more than one vessel by use of code 36.05. However, while this code indicates that PTCA was performed in more than one vessel, its use does not reflect the exact number of procedures performed or the exact number of vessels treated. Similarly, when codes 36.06 and 36.07 are used, they document the insertion of at least one stent. However, these stenting codes do not identify how many stents were inserted in a procedure, nor distinguish insertion of a single stent from insertion of multiple stents. Even the use of one of the stenting codes in conjunction with multiple-PTCA code 36.05 does not distinguish insertion of a single stent from multiple stents. The use of code 36.05 in conjunction with code 36.06 or code 36.07 indicates only performance of PTCA in more than one vessel, along with insertion of at least one stent. The precise numbers of PTCA- treated vessels, the number of vessels into which stents were inserted, and the total number of stents inserted in all treated vessels cannot be determined. Therefore, the capabilities of the current coding structure do not permit the distinction between single and multiple vessel stenting that would be required under the recommended restructuring of the coronary stent DRGs.

We agree that the DRG classification of cases involving coronary stents must be clinically coherent and provide for adequate reimbursement, including those cases requiring multiple stents. For this reason, we created four new ICD-9-CM codes identifying multiple stent insertion (codes 00.45, 00.46, 00.47, and 00.48) and four new codes identifying multiple vessel treatment (codes 00.40, 00.41, 00.42, and 00.43) at the October 7, 2004 ICD-9-CM Coordination and Maintenance Committee Meeting. These eight new codes can be found in Table 6B of this proposed rule. We have worked closely with the coronary stent industry and the clinical community to identify the most logical code structure to identify new codes for both multiple vessel and multiple stent use. Effective October 1, 2005, code 36.05 will be deleted and the eight new codes will be used in its place. Coders are encouraged to use as many codes as necessary to describe each case, using one code to describe the angioplasty or atherectomy, and one code each for the number of vessels treated and the number of stents inserted. Coders are encouraged to record codes accurately, as these data will potentially be the basis for future DRG restructuring. While we agree that use of multiple vessel and stent codes will provide useful information in the future on hospital costs associated with percutaneous coronary procedures, we believe it remains premature to proceed with a restructuring of the current coronary stent DRGs on the basis of the number of vessels treated or the number of stents inserted, or both, in the absence of data reflecting use of this new coding structure. The commenter's second recommendation was that we distinguish ``complex'' from ``noncomplex'' cases in the stent DRGs by expanding the higher weighted DRGs (516 and 526) to include conditions other than AMI. The commenter recommended recognizing certain comorbid and complicating conditions, including hypertensive renal failure, congestive heart failure, diabetes, arteriosclerotic cardiovascular disease, cerebrovascular disease, and certain procedures such as multiple vessel angioplasty or atherectomy (as evidenced by the presence of procedure code 36.05), as indicators of complex cases for this purpose. Specifically, the commenters recommended replacing the current structure with the following four DRGs:

Recommended restructured DRG 516 (Complex percutaneous cardiovascular procedures with non-drug-eluting stents).

Recommended restructured DRG 517 (Noncomplex percutaneous cardiovascular procedures with non-drug-eluting stents).

Recommended restructured DRG 526 (Complex percutaneous cardiovascular procedures with drug-eluting stents).

Recommended restructured DRG 527 (Noncomplex percutaneous cardiovascular procedures with drug-eluting stents).

The commenter argued that this structure would provide an improvement in both clinical and resource coherence over the current structure that classifies cases according to the type of stent inserted and the presence or absence of AMI alone, without considering other complicating conditions. The commenter also presented an analysis, based on previous MedPAR data, that evaluated charges and lengths of stay for cases with expected high resource use and reclassified cases into its recommended new structure of paired ``complex'' and ``noncomplex'' DRGs. The commenter's analysis showed some evidence of clinical and resource coherence in the recommended DRG structure. However, we did not adopt the proposal in the FY 2005 IPPS final rule. First, the data presented by the commenter still represented preliminary experience under a relatively new DRG structure. Second, the analysis did not reveal significant gains in resource coherence compared to existing DRGs for stenting cases. Therefore, we were reluctant to adopt this approach because of comments and concern about whether the overall level of payment in the coronary stent DRGs was adequate. However, we indicated that this issue deserved further study and consideration, and that we would conduct an analysis of this recommendation and other approaches to restructuring these DRGs with updated data in the FY 2006 proposed rule.''

In response to those comments, we analyzed the MedPAR data to determine the impact of certain secondary diagnoses or complicating conditions on the four stent DRGs. Specifically, we examined the data in DRGs 516, 517, 526, and 527, based on the presence of coronary stents (codes 36.06 and 36.07) and the following additional diagnoses:

Congestive heart failure (represented by codes 398.91 (Rheumatic heart failure (congestive)), 402.01 (Hypertensive heart disease, malignant, with heart failure), 402.11, (Hypertensive heart disease, benign, with heart failure), 402.91 (Hypertensive heart disease, unspecified, with heart failure), 404.01 (Hypertensive heart and renal disease, malignant, with heart failure), 404.03 (Hypertensive heart and renal disease, malignant, with heart

[[Page 47293]]

failure and renal failure), 404.11 (Hypertensive heart and renal disease, benign, with heart failure), 404.13 (Hypertensive heart and renal disease, benign, with heart failure and renal failure), 404.91 (Hypertensive heart and renal disease, unspecified, with heart failure), 404.93 (Hypertensive heart and renal disease, unspecified, with heart failure and renal failure), 428.0 (Congestive heart failure, unspecified), and 428.1 (Left heart failure)).

Arteriosclerotic cardiovascular disease (represented by code 429.2 (Cardiovascular disease, unspecified)).

Cerebrovascular disease (represented by codes 430 (Subarachnoid hemorrhage), 431 (Intracerebral hemorrhage), 432.0 (Nontraumatic extradural hemorrhage), 432.1, Subdural hemorrhage, 432.9, (Unspecified intracranial hemorrhage), 433.01 (Occlusion and stenosis of basilar artery, with cerebral infarction), 433.11 (Occlusion and stenosis of carotid artery, with cerebral infarction), 433.21 (Occlusion and stenosis of vertebral artery, with cerebral infarction), 433.31 (Occlusion and stenosis of multiple and bilateral precerebral arteries, with cerebral infarction), 433.81 (Occlusion and stenosis of other specified precerebral artery, with cerebral infarction), 434.01 (Cerebral thrombosis with cerebral infarction), 434.11 (Cerebral embolism with cerebral infarction), 434.91 (Cerebral artery occlusion with cerebral infarction, unspecified), 436 (Acute, but ill-defined, cerebrovascular disease)).

Secondary diagnosis of acute myocardial infarction (represented by codes 410.01 (Acute myocardial infarction of anterolateral wall, initial episode of care), 410.11 (Acute myocardial infarction of other anterior wall, initial episode of care), 410.21 (Acute myocardial infarction of inferolateral wall, initial episode of care), 410.31 (Acute myocardial infarction of inferoposterior wall, initial episode of care), 410.41 (Acute myocardial infarction of other inferior wall, initial episode of care), 410.51 (Acute myocardial infarction of other lateral wall, initial episode of care), 410.61 (True posterior wall infarction, initial episode of care), 410.71 (Subendocardial infarction, initial episode of care), 410.81 (Acute myocardial infarction of other specified sites, initial episode of care), 410.91 (Acute myocardial infarction of unspecified site, initial episode of care)).

Renal failure (represented by codes 403.01 (Hypertensive renal disease, malignant, with renal failure), 403.11 (Hypertensive renal disease, benign, with renal failure), 403.91 (Hypertensive renal disease, unspecified, with renal failure), 585 (Chronic renal failure), V42.0 (Organ or tissue replaced by transplant, kidney), V45.1 (Renal dialysis status), V56.0 (Extracorporeal dialysis), V56.1 (Fitting and adjustment of extracorporeal dialysis catheter), V56.2 (Fitting and adjustment of peritoneal dialysis catheter)). Any renal failure with congestive heart failure will be captured in the 404.xx codes listed above.

We reviewed the cases in the four coronary stent DRGs and found that most of the additional or ``complicated'' cases did, in fact, have higher average charges in most instances. However, these results could potentially be duplicated for many DRGs, or sets of DRGs, within the PPS structure. That is, cases with selected complicating factors will tend to have higher average lengths of stay and average charges than cases without those complicating factors. Because cases with the selected complicating factors necessarily contain sicker patients, longer lengths of stay and higher average charges are to be expected. For example, cases in which patients with a cardiac condition also have renal failure are quite likely to consume higher resources than patients only with a cardiac condition. The presence of code 403.11 (Hypertensive renal disease, malignant, with renal failure) may distinguish cases with higher average charges, but the same argument could be raised for many other procedures across other MDCs.

Generally, we have taken into account the higher costs of cases with complications by maintaining a general list of comorbidities and complications (the CC) list), and, where appropriate, distinguishing pairs of DRGs by ``with and without CCs.'' (This system also specifies exclusions from each pair, to account for cases where a condition on the CC list is an expected and normal constituent of the diagnoses reflected in the paired DRGs.)

Thus, we proposed to restructure the coronary stent DRGs on the basis of the standard CC list to differentiate cases that require greater resources. We believed this list to be more inclusive of true comorbid or complicating conditions than selection of specific secondary diagnosis codes. Therefore, we anticipated that restructuring these DRGs on this basis would result in a logical arrangement of cases with regard to both clinical coherence and resource consumption. We compared the existing CC list with the list of the codes recommended by the commenter as secondary diagnoses. All of the recommended codes already appear on the CC list except for codes 429.2, 432.9, V56.1, and V56.2. Code 429.2 represents a very vague diagnosis (arteriosclerotic cardiovascular disease (ASCVD)). Code 432.9 represents a nonspecific principal diagnosis that is rejected by the MCE when reported as the principal diagnosis. Codes V56.1 and V56.2 describe conditions relating to dialysis for renal failure. Therefore, we believe that our proposal to utilize the existing CC list encompassed most of the cases on the recommended list, as well as other cases with additional CCs requiring additional resources. We examined the MedPAR data for the cases in the coronary stent DRGs, distinguishing cases that include CCs and those that do not. The following table displays the results:

Number of Average length- Average DRG

cases

of-stay

charges

DRG 516--All Cases..............................................

37,325

4.79

$40,278 DRG 516 Cases With CC...........................................

25,806

5.5

43,691 DRG 516 Cases Without CC........................................

11,519

3.0

32,631 DRG 517--All Cases..............................................

64,022

2.58

32,145 DRG 517 Cases With CC...........................................

50,960

2.8

33,178 DRG 517 Cases Without CC........................................

13,062

1.5

28,113 DRG 526--All Cases..............................................

51,431

4.36

45,924 DRG 526 Cases With CC...........................................

32,904

5.2

49,751 DRG 526 Cases Without CC........................................

18,527

2.8

39,126 DRG 527--All Cases..............................................

176,956

2.23

36,087 DRG 527 Cases With CC...........................................

137,641

2.4

37,142 DRG 527 Cases Without CC........................................

39,315

1.3

32,392

[[Page 47294]]

The data show a clear differentiation in average charges between the cases in DRG 516 and 526 ``with CC'' and those ``without CC.'' Therefore, the data suggested that a ``with and without CC'' split in DRG 516 and 526 was warranted. At the same time, the data did not show such a clear differentiation, in either average charges or lengths of stay, among the cases in DRGs 517 and 527.

As a result of this analysis, in the proposed rule, we had originally proposed to delete DRGs 516 and 526, and to substitute four new DRGs in their place. These new DRGs were to have been patterned after existing DRGs 516 and 526, except that they would be split based on the presence or absence of a secondary diagnosis on the existing CC list. Specifically, we intended to create DRG 547 (Percutaneous Cardiovascular Procedure with AMI with CC), DRG 548 (Percutaneous Cardiovascular Procedure with AMI without CC), DRG 549 (Percutaneous Cardiovascular Procedure with Drug-Eluting Stent with AMI with CC), and DRG 550 (Percutaneous Cardiovascular Procedure with Drug-Eluting Stent with AMI without CC). As we noted above, the MedPAR data did not support restructuring DRGs 517 and 527 based on the presence or absence of a CC. Therefore, we proposed to retain these two DRGs in their current forms. We believed this revised structure would result in a more inclusive and comprehensive array of cases within MDC 5 without selectively recognizing certain secondary diagnoses as ``complex.''

We received a number of comments on the proposed restructuring of DRGs 516, 517, 526, and 527 in the FY 2006 IPPS proposed rule.

Comment: All of the commenters approved of the proposed restructuring of these DRGs, especially with regard to dividing DRGs 516 and 526 on the basis of the presence or absence of complicating secondary diagnoses.

Response: We appreciate the comments submitted in support of this proposal.

Comment: One commenter noted that the average patient receives 1.5 stents, and expressed the desire for CMS to begin ``appropriate reimbursement'' in FY 2006, consistent with the additional expense involved when multiple stents are inserted. One commenter remained concerned that the DRG weights significantly underestimate the true costs of performing drug-eluting stent procedures, especially for multiple vessel, multiple stent procedures, and expressed concern that the proposed relative weights could result in financial losses for hospitals, with the result that access to stent procedures is discouraged.

Response: We created new ICD-9-CM procedure codes effective for discharges on or after October 1, 2005, to capture both the number of stents inserted and the number of vessels treated. Absent accurate charge data, we cannot predict the correct relative weight for a DRG containing more than one stent. We reiterate that we will continue to monitor the MedPAR data, and will make future evidence-based changes to the DRG structure and logic as warranted.

Comment: Several commenters supported the maintenance of separate reimbursement structures for drug-eluting stents and recommended that we continue to separate drug-eluting and bare metal stents in different DRGs until such time as the bare metal stents represent an insignificant proportion of the total coronary stent discharges.

Response: We recognize that the resources surrounding bare metal stents and drug-eluting stents differ appreciably and will continue to keep these cases separate from each other until such time as it is appropriate, according to the evidence provided in our MedPAR data, that these cases can be combined.

Comment: Several commenters supported CMS' proposal to create eight new procedure codes; four codes describing the number of vessels treated and four codes describing the number of stents inserted. In addition, two commenters suggested that CMS should issue a separate communication reiterating the correct use of these codes.

Response: We take this opportunity to clear up a misconception. The codes published in Tables 6A through 6F are not proposed codes. They are final codes, and as such, are not subject to comment. Absent any typographical errors or late changes to the codes, they may be considered available for use on October 1 of the following fiscal year. This year, because of the changes made by the March 31, 2005 and April 1, 2005 ICD-9-CM Coordination and Maintenance Committee, the codes in the proposed rule were not as complete as those codes published in this final rule. The codes contained in Tables 6A through 6F of this final rule include all new codes for FY 2006, which will go into effect on October 1, 2005.

CMS partners with the American Hospital Association with regard to correct coding advice published in the Coding Clinic for ICD-9-CM. AHA's fourth edition of the year always includes the new codes for the upcoming year and includes examples on their proper use. In addition, CMS' MedLearn site at http://www.cms.hhs.gov/medlearn/icd9code.asp#top

contains coding information.

Comment: One commenter recommended that the use of the eight new codes describing number of vessels and number of stents be used on both coronary and peripheral vessels.

Response: The note that will appear at the top of the 00.4 (Adjunct Vascular System Procedures) section of Tabular section of the ICD-9-CM Procedure Coding Book will read as follows: ``These codes can apply to both coronary and peripheral vessels. These codes are to be used in conjunction with other therapeutic procedure codes to provide additional information on the number of vessels upon which a procedure was performed or the number of stents inserted, or both. As appropriate, hospitals should code both the number of vessels operated on (00.40 through 00.43) and the number of stents inserted (00.45 through 00.48).

Comment: One commenter stated that by the time CMS gets data on the eight new codes, it will be FY 2008, and hospitals will have had inadequate reimbursement for multiple stents until then. The commenter suggested that CMS incorporated additional payment for multiple stents and multiple vessels treated into the FY 2007 weights.

Response: We will follow the use of these codes, but may not be prepared to make any DRG changes based on their use with only one year's worth of data.

Comment: One commenter stated that DRGs should not be restructured for multiple stent insertion without adequate data to support our decisionmaking process.

Response: We agree and intend to closely follow the use of these eight new codes in the MedPAR data.

Comment: One commenter was not convinced that the proposed new structure of DRGs 516 and 526, with and without comorbidities and complications should be the permanent solution for all coronary stent DRGs. This commenter agreed that the new structure of these DRGs should not preclude subsequent restructuring of the stent DRGs.

Response: We agree that restructuring DRGs 516 and 526 in the proposed manner might not be a permanent solution for classifying all stent DRGs. However, we have now decided not to adopt the proposed restructuring of DRGs 516 and 526 that was described in the proposed rule. We have now determined that it is appropriate to restructure nine DRGs in MDC 5, including DRGs 516, 517, 526, and 527,

[[Page 47295]]

on the basis of the presence or absence of a major cardiovascular condition. We are making this change in the DRG structure in response to public comments concerning our response to MedPAC's recommendations to better recognize severity in the DRG system. The full text of the changes we are making to the cardiovascular DRG, including the coronary artery stent DRGs, can be found in section IX.A. of this final rule.

Comment: One commenter requested that CMS adopt an ICD-9-CM code that was discussed at the October 7, 2004 ICD-9-CM Coordination and Maintenance Committee. That code, had it been adopted, would have been 00.44 (Procedure on bifurcated vessels) in the new series of codes describing the number of vessels treated. The commenter stated that the creation of this code is critical to understanding the contemporary approaches to treatment of coronary artery disease. The commenter further stated that treatment of stenosis [of a blood vessel] at a bifurcation represents 25 to 30 percent of percutaneous coronary interventions and recommended that coders use one code for number of vessels, one code for number of stents, and an additional code to note that a bifurcated vessel was treated. According to the commenter, a new code for the treatment of a bifurcated vessel is necessary because the existing codes that describe the number of vessels treated (codes 00.40 through 00.43) will only be used by coders for the counting of uninterrupted, straight vessels.

Response: We did not choose to create a new code for procedure on a bifurcated vessel for two reasons. First, we do not believe that level of granularity is needed in order to accurately code stent insertion for bifurcated vessels. We believe that the codes for multiple stents and vessels will provide the necessary information about resource use for the procedure. Second, we are concerned that coders will not have sufficient information documented in the medical record to identify procedures on bifurcated vessels as opposed to a specific number of procedures on a specific number of vessels. Because procedures on bifurcated vessels are so prevalent (25 to 30 percent, according to the commenter), they should be considered technical variants rather than distinct entities to be coded separately. We solicited input from the industry when creating the new coronary stent codes, and we believe that the new codes as they exist adequately capture resource utilization. We also note that this level of detail is not present in the Current Procedural Terminology (CPT) coding structure, which is the basis upon which physicians are paid.

Accordingly, in this final rule, for FY 2006, we are deleting DRGs 516, 517, 526, and 527 for percutaneous placement of both drug-eluting and nondrug-eluting stents. We are creating four new DRGs in their places. Rather than divide these DRG pairs based on whether the patient had an acute myocardial (AMI), we are splitting each pair of DRGs based on the presence or absence of a major cardiovascular condition. Although, as discussed in the proposed rule, in the past we have expressed concerns regarding selectively recognizing secondary diagnoses or complicating conditions, particularly conditions from other MDCs, in making DRG assignments, we believe these concerns are not relevant to the new cardiovascular DRGs. While we are adopting an approach for distinguishing patients with complex conditions, with a few exceptions, our approach uses complex cardiovascular conditions (or diagnoses within the MDC) to decide whether a patient should be assigned to the higher weighted DRG. In those cases where we have used a diagnosis from another MDC in assigning a patient to the MCV DRG, the condition is generally a closely related vascular condition that is linked to the patient's cardiovascular illness. We believe that this revised structure identifies subgroups of significantly more severe patients who use greater hospital resources more accurately than was possible under the previous DRGs. The new DRG titles are:

DRG 555 (Percutaneous Cardiovascular Procedure With Major Cardiovascular Diagnosis (formerly DRG 516)

DRG 556 (Percutaneous Cardiovascular Procedure With Non- Drug-Eluting Stent Without Major Cardiovascular Diagnosis (formerly DRG 517)

DRG 557 (Percutaneous Cardiovascular Procedure With Drug- Eluting Stent With Major Cardiovascular Diagnosis (formerly DRG 526)

DRG 558 (Percutaneous Cardiovascular Procedure With Drug- Eluting Stent Without Major Cardiovascular Diagnosis (formerly DRG 527)

We refer the reader to section IX.A. of the preamble to this final rule for a full presentation of the changes to the DRGs for coronary artery stents for FY 2006.

Although we are adopting some restructuring of the coronary stent DRGs for FY 2006, it is important to note that this change does not preclude proposals in subsequent years to further restructure the coronary stent DRGs based on the number of vessels treated. We will continue to monitor and analyze clinical and resource trends in this area. For example, we have found indications in the current data that treatment may be moving toward use of drug-eluting stents, and away from use of bare metal stents. Specifically, cases in DRGs 516 and 517, which utilize bare metal stents, comprise only 44.4 percent, or less than half, of the cases in the four coronary stent DRGs in the MedPAR data we analyzed. As use of drug-eluting stents becomes the standard of treatment, we may consider over time whether to dispense with the distinction between these stents and the older bare metal stent technology in the structure of the coronary stent DRGs. In addition, we will continue to consider whether the structure of these DRGs ought to reflect differences in the number of vessels treated or the number of stents inserted, or both. As we discussed above, a new coding structure capable of identifying multiple vessel treatment and the insertion of multiple stents will go into effect on October 1, 2005. It remains premature to restructure the coronary stent DRGs on the basis of the number of vessels treated or the number of stents inserted, or both, until data reflecting the use of these new codes become available. After we have pertinent data in our historical MedPAR database, we will analyze those data in order to determine whether a restructuring of the DRGs based on multiple vessel treatment or insertion of multiple stents, or both, is warranted.

We refer the reader to Table 6B of this final rule for the descriptions of four new ICD-9-CM codes identifying multiple stent insertion (codes 00.45, 00.46, 00.47, and 00.48) and four new codes identifying multiple vessel treatment (codes 00.40, 00.41, 00.42, and 00.43). Coders are encouraged to use as many codes as necessary to describe each case, using new code 00.66 (Percutaneous transluminal coronary angioplasty [PTCA] or coronary atherectomy) and one code each for the number of vessels treated and the number of stents inserted. Coders are encouraged to record codes accurately, irrespective of whether the code has an impact on the DRG assignment, as these data will potentially be the basis for future DRG restructuring. d. Insertion of Left Atrial Appendage Device

Atrial fibrillation is a common heart rhythm disorder that can lead to a cardiovascular blood clot formation

[[Page 47296]]

leading to increased risk of stroke. According to product literature, nearly all strokes are from embolic clots arising in the left atrial appendage of the heart: an appendage for which there is no useful function. Standard therapy uses anticoagulation drugs. However, these drugs may be contraindicated in certain patients and may cause complications such as bleeding. The underlying concept behind the left atrial appendage device is to block off the left atrial appendage, so that the blood clots formed therein cannot travel to other sites in the vascular system. The device is implanted using a percutaneous catheter procedure under fluoroscopy through the femoral vein. Implantation is performed in a hospital catheterization laboratory using standard transseptal technique, with the patient generally under local anesthesia. The procedure takes approximately 1 hour, and most patients stay overnight in the hospital.

In the FY 2005 IPPS final rule (69 FR 48978, August 11, 2004), we discussed the DRG assignment of new ICD-9-CM procedure code 37.90 (Insertion of left atrial appendage device) for clinical trials, effective for discharges occurring on or after October 1, 2004, to DRG 518 (Percutaneous Cardiovascular Procedure without Coronary Artery Stent or Acute Myocardial Infarction). In that final rule, we addressed the DRG assignment of procedure code 37.90 in response to a comment from a manufacturer who suggested that placement of the code in DRG 108 (Other Cardiothoracic Procedures) was more representative of the complexity of the procedure than placement in DRG 518. The manufacturer indicated that the suggested placement of procedure code 37.90 in DRG 108 was justified because another percutaneous procedure, described by ICD-9-CM procedure code 35.52 (Repair of atrial septal defect with prosthesis, closed technique), was assigned to DRG 108. As we indicated in the FY 2005 final rule (69 FR 48978), this comment prompted us to examine data in the FY 2003 MedPAR file for cases of code 35.52 assigned to DRG 108 and DRG 518 in comparison to all cases assigned to DRG 108. We found the following:

Number of Average length Average DRG

cases

of stay

charges

DRG 108 With Code 35.52 Reported................................

523

2.69

$29,231 DRG 108--All cases..............................................

5,293

10.1

76,274 DRG 518--All cases..............................................

39,553

4.3

31,955

Therefore, we concluded that procedure code 35.52 showed a decided similarity to the cases found in DRG 518, not DRG 108. At that time, we determined that we would analyze the cases for both clinical coherence and charge data as part of the IPPS FY 2006 process of identifying the most appropriate DRG assignment for procedure code 35.52.

We examined data from the FY 2004 MedPAR file and found results for cases assigned to DRG 108 and DRG 518 that are similar to last year's findings as indicated in the chart below:

Number of Average length- Average DRG

cases

of-stay

charges

DRG 108 With Code 35.52 Reported................................

872

2.42

$29,579 DRG 108--All cases..............................................

8,264

9.81

81,323 DRG 518--All cases..............................................

38,624

3.49

27,591

From this comparison, we found that when an atrial septal defect is percutaneously repaired, and procedure code 35.52 is the only code reported in DRG 108, there is a significant discrepancy in both the average charges and the average length of stay between the cases with procedure code 35.52 reported in DRG 108 and the total cases in DRG 108. The total cases in DRG 108 have average charges of $51,744 greater than the 872 cases in DRG 108 reporting procedure code 35.52 as the only procedure. The total cases in DRG 108 also have an average length of stay of 7.39 days greater than the average length of stay for cases in DRG 108 with procedure code 35.52 reported. In comparison, the total cases in DRG 518 have average charges of only $1,988 lower than the cases in DRG 108 with only procedure code 35.52 reported. In addition, the length of stay in total cases in DRG 518 is more closely related to cases in DRG 108 with only procedure code 35.52 reported. Based on this analysis, we proposed to move procedure code 35.52 out of DRG 108 and place it in DRG 518.

Comment: One commenter agreed that the left atrial appendage device procedure code should be moved out of DRG 108 and into DRG 518 based on significantly lower average charges and length of stay as compared to the majority of cases within the current classification.

Response: Even though this comment did not exactly reflect our proposal regarding the left atrial appendage device, we are interpreting the commenter's statement to mean that it agreed that code 35.52 should be removed from DRG 108.

Comment: One commenter addressed the proposed removal of code 35.52 from DRG 108. The commenter acknowledged that the resource intensity for patients undergoing percutaneous atrial septal defect repair is less than that of open repair, but did not believe that the costs are akin to procedures presently assigned to DRG 518 because of the cost of the closure device and additional testing, such as electrocardiography. The commenter recommended that CMS not move code 35.52 out of DRG 108 until better data can be gathered and a more appropriate reimbursement calculation can be developed.

Response: This year, CMS undertook an extensive review of MDC 5 after issuance of the FY 2006 IPPS proposed rule in response to MedPAC's recommendations regarding restructuring the Medicare DRG system to improve payment accuracy under the IPPS. A discussion of the results of that review and our subsequent decision in response to a comment on the proposed rule to make changes to nine cardiovascular DRGs, can be found in section IX.A. of this preamble. During that review, we evaluated each surgical DRG within MDC 5. In addition, within each DRG, we evaluated each procedure code to determine the number of cases,

[[Page 47297]]

the average length of stay, and the average standardized charges. In DRG 108, the results were the same as in the table shown above in this section, and published in the FY 2006 IPPS proposed rule. Code 35.52 had an average length of stay of approximately one fourth of the rest of the cases in that DRG, and had average charges that were greater than $51,700 less than the remainder of the cases in DRG 108. In addition, code 35.52 represents a closed technique approach, unlike the other cases in DRG 108. We believe this is compelling evidence that this procedure is not most appropriately assigned to DRG 108. Therefore, we are finalizing our proposal to move code 35.52 out of DRG 108 and into DRG 518 with cases that resemble it in average length of stay, average charges, and clinical coherence. We believe that this move will result in a more coherent group of cases in DRG 518 that reflect all percutaneous procedures.

Comment: Three commenters did not believe that the left atrial appendage device, represented by new code 37.90, should be placed in DRG 518. They believed that DRG 518 does not cover the costs for the procedure and device, and suggested placement in another DRG that would include similar procedures and a better reimbursement. Two commenters suggested that a more appropriate DRG would be either DRG 108 or DRG 111 (Major Cardiovascular Procedures Without CC).

Response: Based on our data review and discussion above, we do not believe that placement of code 37.90 is appropriate in DRG 108. Code 37.90 is a percutaneously placed device utilizing local anesthesia, and with an expected length of stay of one day.

We reviewed cases in the MedPAR file assigned to both DRG 110 (Major Cardiovascular Procedures With CC) and DRG 111. The results of the review show that both open and percutaneous procedures are grouped in these paired DRGs. A comparison of the MedPAR data in DRGs 110, 111, and 518 is shown in the following table:

Average DRG

Number of Average length standardized cases

of stay

charges

DRG 110.........................................................

53,527

\1\ 8.4

$66,475 DRG 111.........................................................

9,438

\1\ 3.43

26,941 DRG 518--All cases..............................................

38,624

3.49

27,591 DRG 518 with code 37.90.........................................

0

0

0

\1\ Days.

As shown in the table, code 37.90 in DRG 518 has not been reported in the database yet. It is a new code; therefore, it has no payment history. We note that the cases in DRG 518 closely match those in DRG 111 in terms of both average length of stay and average charges. However, we also note that DRGs 110 and 111 are paired DRGs with significantly different average charges and lengths of stay. Even with a CC, we believe it is unlikely that an endovascular placement of a left atrial appendage device will approximate the costs of cases to be assigned to DRG 110. Therefore, in our view, there is the potential for significant overpayment if we were to assign the left atrial appendage device to DRG pairs 110 and 111. We continue to believe that placement of the left atrial appendage device in DRG 518 is appropriate absent any evidence that would convince us otherwise. Therefore, we are not making any changes in our proposal in this final rule. We will continue to monitor its data in our annual review of DRGs and the IPPS.

As we proposed, in this final rule we are moving procedure code 35.52 out of DRG 108 and placing it in DRG 518. We believe that this move will result in a more coherent group of cases in DRG 518 that reflect all percutaneous procedures. e. External Heart Assist System Implant

In the August 1, 2002 final rule (67 FR 49989), we attempted to clinically and financially align ventricular assist device (VAD) procedures by creating DRG 525 (Heart Assist System Implant). We also noted that cases in which a heart transplant also occurred during the same hospitalization episode would continue to be assigned to DRG 103 (Heart Transplant).

After further data review during the subsequent 2 years, we decided to realign the DRGs containing VAD codes for FY 2005. In the August 11, 2004 final rule (69 FR 48927), we announced changes to DRG 103, DRG 104 (Cardiac Valve and Other Major Cardiothoracic Procedure with Cardiac Catheterization), DRG 105 (Cardiac Valve and Other Major Cardiothoracic Procedures Without Cardiac Catheterization), and DRG 525.

In summary, these changes included--

Moving code 37.66 (Insertion of implantable heart assist system) out of DRG 525 and into DRG 103.

Renaming DRG 525 as ``Other Heart Assist System Implant.''

Moving code 37.62 (Insertion of non-implantable heart assist system) out of DRGs 104 and 105 and back into DRG 525.

DRG 525 currently consists of any principal diagnosis in MDC 5, plus the following surgical procedure codes:

37.52, Implantation of total replacement heart system*

37.53, Replacement or repair of thoracic unit of total replacement heart system*

37.54, Replacement or repair of other implantable component of total replacement heart system*

37.62, Insertion of non-implantable heart assist system

37.63, Repair of heart assist system

37.65, Implant of external heart assist system

*These codes represent noncovered services for Medicare beneficiaries. However, it is our longstanding practice to assign every code in the ICD-9-CM classification to a DRG. Therefore, they have been assigned to DRG 525.

Since that decision, we have been encouraged by a manufacturer to reevaluate DRG 525 for FY 2006. The manufacturer requested that we again review the data surrounding cases reporting code 37.65, and suggested moving these cases into DRG 103. The manufacturer pointed out the following: Code 37.65 describes the implantation of an external heart assist system and is currently approved by the FDA as a bridge- to-recovery device. From the standpoint of clinical status, the patients in DRG 103 and the patients receiving an external heart assist system are similar because their native hearts cannot support circulation, and absent a heart transplant, a mechanical pump is needed for patient survival. The surgical procedures for implantation of both an internal VAD and an external VAD are very similar. However, the external heart assist system (code 37.65) is a less expensive device than the implantable heart assist system (code 37.66).

[[Page 47298]]

Further, the Medicare charge data show that patients in DRG 525 receiving the external heart assist system had an average length of stay that was more than 28 days less than all patients in DRG 103.

The manufacturer suggested that the payment differential between DRGs 103 and 525 provides an incentive to choose the higher paying device, and asserted that only a subset of patients receiving an implantable heart assist system are best served by this device (code 37.66). The manufacturer also suggested that the initial use of the least expensive therapeutically appropriate device yields both the best clinical outcomes and the lowest total system costs.

We note that, under the DRG system, our intent is to create payments that are reflective of the average resources required to treat a particular case. Our goal is that physicians and hospitals should make treatment decisions based on the clinical needs of the patient and not financial incentives.

When we reviewed the FY 2004 MedPAR data, we were able to demonstrate the following comparisons:

Number of Average length Average DRG

cases

of stay

charges

DRG 103--All cases..............................................

633

37.5

$313,583 DRG 103 with code 37.65 reported................................

9

81.3

625,065 DRG 525--All cases..............................................

291

13.66

173,854 DRG 525 with code 37.65 reported................................

110

9.26

206,497 DRG 525 without code 37.65 reported.............................

181

16.34

154,015

Note: This table does not contain the same data that appear in the table in the proposed rule (70 FR 23322). The row containing ``DRG 103 without code 37.65'' had values of ``0'' in all fields. These entries were confusing and therefore deleted.

The above table shows that the 37.8 percent of cases in DRG 525 that reported code 37.65 have average charges that are nearly $33,000 higher than the average charges for all cases in the DRG. However, the average charges for the subset of cases with code 37.65 in DRG 525 ($206,497) are more than $107,086 lower than the average charges for all cases in DRG 103 ($313,583). Furthermore, the average length of stay for the subset of patients in DRG 525 receiving an external heart assist system was 9.26 days compared to 37.5 days for the 633 cases in DRG 103.

We note that the analysis above presents the difference in average charges, not costs. Because hospitals' charges are higher than costs, the difference in hospital costs will be less than the figures shown here.

Moving all cases containing code 37.65 from DRG 525 to DRG 103 would have two consequences. The cases in DRG 103 reporting code 37.65 would be appreciably overpaid, which would be inconsistent with our goal of coherent reimbursement structure within the DRGs. In addition, the relative weight of DRG 103 would ultimately decrease by moving the less resource-intensive external heart procedures into the same DRG with the more expensive heart transplant cases. The net effect would be an underpayment for heart transplant cases. Alternatively, we also reconsidered our position on moving the insertion of an implantable heart assist system (code 37.66) back into DRG 525. However, as shown in the FY 2005 IPPS final rule (69 FR 48929), the resource costs associated with caring for a patient receiving an implantable heart assist system are far more similar to those cases receiving a heart transplant in DRG 103 than they are to cases in DRG 525. For these reasons, we did not propose to make any changes to the structure of either DRG 103 or DRG 525.

Comment: Six commenters mentioned the high cost of the external heart assist device and for treatment for implantation of the device, and requested that CMS increase payment to cover the cost of caring for the patients that can benefit from this technology.

Two commenters agreed with CMS' assessment that the cost associated with implantation of an external heart assist system are considerably less than a heart transplant or insertion of an implantable heart assist system. One commenter echoed CMS' concerns that movement of code 37.65 to DRG 103 would result in overpayment for that service and would result in a decrease of the relative weight of the heart transplant DRG, ultimately resulting in underpayment of heart transplant cases. Both commenters agreed with CMS' decision not to include the implantation of external heart assist systems in DRG 103.

Several commenters noted that significant achievements in the areas of patient selection, implantation technique, and post-implant management have been made surrounding this technology. They added that improvements in the external heart assist device itself have been reported to make the newer devices safer and more durable. One commenter noted that observations from personal experience and research demonstrate that recent improvements to the device have resulted in increased survival rates from 35 percent (the national average) to nearly 50 percent. Several commenters mentioned that, with experience, they have discovered that a longer period of support is required than was originally anticipated for the patient's native heart to recover. The commenters stated that, originally, patients were supported an average of 5 to 7 days, but it has been found that patient outcomes were better with a longer support period, perhaps as long as 30 to 60 days. These commenters cited the increased expenses related to supporting the patient and the major financial commitment on the part of the hospitals choosing to treat this severely ill group of patients as reasons for requesting increased payment for this population of cases.

One commenter offered the following four proposals to address the payment differences between the external heart assist device and an implantable device:

Create a new DRG for patients requiring heart assist devices who also sustained an Acute Myocardial Infarction (AMI) because these patients have higher resource consumption than patients with other diagnoses in MDC 5.

Assign all cases with AMI and a procedure code of 37.65 to DRG 103.

Increase the overall weight of DRG 525 to better align it with ``true hospital charges.''

Allow a second DRG payment or an add-on payment for heart transplantations if recovery of the patient's native heart is first attempted.

Response: We appreciate the commenters' thorough understanding of the IPPS DRG grouping and payment process. We are aware that the external heart assist device cases represent a very resource-intensive group of patients. For this reason, we carefully reviewed the suggestions from the commenter about potential DRG payment policy changes that we could make to address the issue. We reviewed the MedPAR data in DRG 525, using ICD-9-CM codes 410.01 through 410.91 to identify AMIs. In

[[Page 47299]]

addition, we reviewed all cases of patients who received the external heart assist device procedure represented by ICD-9-CM code 37.65. The results are summarized in the following table:

Average length Number of

of stay

Average cases

(days)

charges

DRG 525--Cases with Any Diagnosis of AMI........................

46

8.5

$195,758 DRG 525--Cases of Principal Diagnosis of AMI....................

31

8.9

210,369 DRG 525--Cases with Secondary Diagnosis of AMI..................

15

7.7

165,562 DRG 525--Cases with No Diagnosis of AMI.........................

71

9.2

204,472 DRG 525--All Cases..............................................

291

13.66

206,497

We do not believe that these data demonstrate that the presence of an AMI has significant impact on either the length of stay or the average standardized charges. All cases with AMI have lower lengths of stay than both the average of all cases in DRG 525 (13.66 days) and the 71 cases in which no AMI was documented (9.2 days). Likewise, only those cases with a principal diagnosis of AMI have slightly higher charges than either the group without AMI, or the total of all cases. Because the data do not justify it, we are rejecting the suggestion of creating a new DRG for patients receiving an external heart assist device, as identified by procedure code 37.65, with any diagnosis of AMI.

With respect to the commenter's second suggestion, our data clearly demonstrate in the above table that patients with an AMI and procedure code 37.65 have average standardized charges of $210,369. The first table in this section that was included in the proposed rule shows that cases in DRG 103 have average standardized charges of $313,583. We believe that the relative weight of DRG 103 would eventually decrease by moving all of the less resource-intensive external heart procedures into the same DRG with the more expensive heart transplant cases. For these reasons, we are rejecting the commenter's proposal to assign cases with AMI and code 37.65 to DRG 103.

With regard to the suggestion (received many times) to selectively increase the relative weight of specific DRGs, the DRG relative weights are annually recalibrated based on Medicare hospital discharges using the most current charge information available (FY 2004 MedPAR file for the FY 2006 relative weights). We use a complex mathematical algorithm to determine the relative weights that is fully explained in section II. of this preamble. The DRG relative weights are neither arbitrarily nor capriciously assigned. However, if we adopted the suggestion to select a relative weight for a specific DRG outside of this process, we are concerned that the relative weight determination would be viewed as arbitrary and capricious, and we would lose the advantage of having an objective methodology that bases the relative weight on average hospital charges. For this reason, we are not adopting the commenter's suggestion to select a relative weight for external heart assist device cases outside of our traditional process.

The commenter's fourth suggestion was to make two payments for a single inpatient stay when the patient receives the external heart assist system, recovery of the patient's native heart is attempted and fails, and the patient receives a heart transplant. In cases where the patient received the external heart assist system and later receives a heart transplant, the case is already paid using DRG 103. In this situation, the relative weight for DRG 103 will reflect the average charges for all patients in the DRG, including those described in the scenario presented by the commenter. Thus, to the extent that hospital charges for these patients are already reflected in the relative weight for the DRG, we do not believe that it is necessary for Medicare to make a second payment. To arbitrarily select one DRG, or a group of DRGs, and add an additional DRG payment to those cases is contrary to our stated goal of having a system in which all cases are fairly considered by the same recalibration formula. Therefore, we do not intend to either determine an additional DRG payment or an add-on payment for this category of patients.

We reiterate that our data do not support the argument that patients receiving the external heart assist device have longer lengths of stay than other patients in DRG 525, even though the data show that their average charges are higher, as noted in the above table. In determining the possible reasons for higher average charges and lower lengths of stay, we further examined the Medicare billing data. We found that almost 76 percent of the Medicare beneficiaries receiving the external heart device expired during the hospital stay. Thus, the shorter length of stay and the higher average charges for these patients compared to other patients in DRG 525 are likely explained by the high cost of the device and the fact that these patients are severely ill and frequently expire.

Upon further analysis of the data, we did find that there was a single subgroup of patients who are comparable in resource usage and length of stay to those included in DRG 103. These patients received both the external heart assist device and later had it removed after a lengthy period of rest and recovery. We note that commenters provided information indicating that survival rates are improving for patients receiving more advanced versions of these devices. In addition, commenters provided information indicating that longer periods of support with the external heart assist device are improving patients' survival chances and opportunity to be discharged with their native heart. According to information included with the comments, the data show a 50-percent survival rate with an average total length of stay of 43 days for all AMI heart recovery patients. On average, a surviving patient will receive 31 days of average support time followed by an additional 38 days in the hospital after the device is removed. Based on the commenter's information from a later year than our MedPAR data, it is clear that patients weaned from the external heart assist system have longer lengths of stay and are very different from the average patients having this procedure that are in our FY 2004 data. Given the newness of this procedure, the Medicare charge data included a limited number of patients having the device implanted and removed. However, the Medicare charge data did support that patients receiving both an implant and removal of an external heart assist system in a single hospital stay had an average length of stay exceeding 50 days and average charges of $378,000 that are more comparable to

[[Page 47300]]

patients in DRG 103 than DRG 525. While we did not suggest a change to DRG 103 in the proposed rule, we believe that consideration of the comments is best served by recognizing this unique subset of patients and making a DRG change which acknowledges the increased resources required for improvements in their care.

The commenter has provided us with data showing that with superior patient selection, and increased duration of treatment with an improved device, the patients are more likely to be discharged from the hospital with the native heart intact. While we have limited Medicare data and the data are from a different year than the commenter's data, our data do support that patients having an external heart assist device implanted and removed during the same admission are comparable to in costs and average length of stay to heart transplant and implantable heart assist system patients in DRG 103. While we did not suggest a change to DRG 103 in the proposed rule, we believe that consideration of the comments is best served by recognizing this unique subset of patients, and making a DRG change that acknowledges the increased resources required for improvement in their care. Because we believe that this therapy offers a treatment option to patients who have limited alternatives, we are making a change to the DRG using the limited Medicare data we have available rather than waiting a year to receive more supporting data.

For the reasons stated above, for FY 2006, we are reconfiguring DRG 103 in the following manner: Those patients who have both the implantation of the external VAD (code 37.65) and the explantation of that VAD (code 37.64) prior to the hospital discharge will be assigned to DRG 103. The revised DRG 103 contains the following codes:

33.6, Combined heart-lung transplantation

37.51, Heart transplantation

37.66, Insertion of implantable heart assist system

Or

37.65, Implant of external heart assist system

And

37.64, Removal of heart assist system.

By making this change, Medicare will be making higher payments for patients who receive both an implant and an explant of an external heart assist system during a single hospital stay. Our intent in establishing this policy is to recognize the higher costs of patients who have a longer length of stay and are discharged alive with their native heart. Cases in which a heart transplant also occurs during the same hospitalization episode would continue to be assigned to DRG 103.

In order to accurately monitor these patients and obtain more information on patients with these conditions, we intend to have the Quality Improvement Organizations (QIOs, formerly the PROs) review all cases in DRG 103 under the auspices of their eighth scope of work to determine whether implantation and care during the admission were reasonable and necessary to promote the recovery of the injured myocardium and lead to improvement of the patient's condition. For medical review under this contract, the QIOs determine whether items and services are reasonable and medically necessary and whether the quality of such services meets professionally recognized standards of health care. In addition, in hospitals subject to the IPPS, the QIOs review the validity of diagnostic information, the completeness, adequacy, and quality of care provided, and the appropriateness of admissions and discharges. We will continue to examine the claims data in upcoming years to determine if CMS' consideration surrounding the unique circumstances of these patients and this treatment modality were in the best interest of both the patients and the Medicare program. f. Carotid Artery Stent

Stroke is the third leading cause of death in the United States and the leading cause of serious, long-term disability. Approximately 70 percent of all strokes occur in people age 65 and older. The carotid artery, located in the neck, is the principal artery supplying the head and neck with blood. Accumulation of plaque in the carotid artery can lead to stroke either by decreasing the blood flow to the brain or by having plaque break free and lodge in the brain or in other arteries to the head. The percutaneous transluminal angioplasty (PTA) procedure involves inflating a balloon-like device in the narrowed section of the carotid artery to reopen the vessel. A carotid stent is then deployed in the artery to prevent the vessel from closing or restenosing. A distal filter device (embolic protection device) may also be present, which is intended to prevent pieces of plaque from entering the bloodstream.

Effective July 1, 2001, Medicare covers PTA of the carotid artery concurrent with carotid stent placement when furnished in accordance with the FDA-approved protocols governing Category B Investigational Device Exemption (IDE) clinical trials. PTA of the carotid artery, when provided solely for the purpose of carotid artery dilation concurrent with carotid stent placement, is considered to be a reasonable and necessary service only when provided in the context of such clinical trials and, therefore, is considered a covered service for the purposes of these trials. Performance of PTA in the carotid artery when used to treat obstructive lesions outside of approved protocols governing Category B IDE clinical trials remains a noncovered service. At its April 1, 2004 meeting, the ICD-9-CM Coordination and Maintenance Committee discussed creation of a new code or codes to identify carotid artery stenting, along with a concomitant percutaneous angioplasty or atherectomy (PTA) code for delivery of the stent(s). We established codes for carotid artery stenting procedures for use with discharges occurring on or after October 1, 2004 inpatients who are enrolled in an FDA-approved clinical trial and are using on-label FDA-approved stents and embolic protection devices. These codes are as follows:

00.61 (Percutaneous angioplasty or atherectomy of precerebral (extracranial vessel(s)); and

00.63 (Percutaneous insertion of carotid artery stent(s)).

We assigned procedure code 00.61 to four MDCs and seven DRGs. The most likely scenario is that in which cases are assigned to MDC 1 (Diseases and Disorders of the Nervous System) in DRGs 533 (Extracranial Procedures with CC) and 534 (Extracranial Procedures without CC). Other DRG assignments can be found in Table 6B of the Addendum to the FY 2005 IPPS final rule (69 FR 49624).

In the FY 2005 IPPS final rule, we indicated that we would continue to monitor DRGs 533 and 534 and procedure code 00.61 in combination with procedure code 00.63 in upcoming annual DRG reviews. For the FY 2006 IPPS proposed rule and this final rule, we used proxy codes to evaluate the costs and DRG assignments for carotid artery stenting because codes 00.61 and 00.63 were only approved for use beginning October 1, 2004, and MedPAR data for this period are not yet available. We used procedure code 39.50 (Angioplasty or atherectomy of other noncoronary vessel(s)) in combination with procedure code 39.90 (Insertion of nondrug-eluting peripheral vessel stent(s)) in DRGs 533 and 534 as the proxy codes for carotid artery stenting. For this evaluation, we used principal diagnosis code 433.10 (Occlusion and stenosis of carotid artery, without mention of cerebral

[[Page 47301]]

infarction) to reflect the clinical trial criteria.

The following chart shows our findings:

Number of Average length Average DRG

cases

of stay

charges

DRG 533--All cases..............................................

44,677

3.73

$24,464 DRG 533 with codes 39.50 and 39.90 reported.....................

1,586

3.13

29,737 DRG 534--All cases..............................................

42,493

1.79

15,873 DRG 534 with codes 39.50 and 39.90 reported.....................

1,397

1.54

22,002

The patients receiving a carotid stent (codes 39.50 and 39.90) represented 3.5 percent of all cases in DRG 534. On average, patients receiving a carotid stent had slightly shorter average lengths of stay than other patients in DRGs 533 and 534. While the average charges for patients receiving a carotid artery stent were higher than for other patients in DRG 534, in our view, the small number of cases and the magnitude of the difference in average charges are not sufficient to justify a change in the DRGs.

Because we have a paucity of data for the carotid stent device and its insertion, we believe it is premature to revise the DRG structure at this time. We expect to revisit this analysis once data become available on the new codes for carotid artery stents.

We received 11 comments on our presentation of the carotid stent device issue in the FY 2006 IPPS proposed rule.

Comment: One commenter recommended that CMS include carotid stenting in the DRG for carotid endarterectomy in FY 2006 and ensure that the data it is collecting for setting payment rates in FY 2007 appropriately accounts for the cost of the device.

Response: Code 38.12 (Endarterectomy, other vessels of head and neck) describes the open endarterectomy procedure, and is assigned to DRGs 533 and 534 which is the same DRG assignment as the endovascular endarterectomy. Therefore, both the open endarterectomy and the placement of carotid stent result in assignment to the same DRG, which reflects CMS' policy of placing new codes in predecessor DRGs. We point out that codes 00.61 and 00.63 must be used together to allow payment for carotid stenting. Code 00.63 is not recognized by the GROUPER program as a stand-alone O.R. procedure and, as such, has no impact on DRG assignment. Therefore, we anticipate that the cost of the device will be reflected in the hospital charges.

Comment: One commenter agreed with our presentation in the proposed rule and suggested that we should make no change to the DRG assignment for carotid artery stenting.

Response: We agree and will not be making a change to the DRG assignment for carotid artery stenting.

Comment: Nine commenters encouraged CMS to create two new DRGs for carotid stent procedures and split these new DRGs on the basis of the presence or absence of comorbidities or complications. They believed that, even though the current volume of carotid artery stenting cases appears small, the recent availability of FDA-approved devices, new and ongoing clinical trials, multiple post-market registries, as well as expanded Medicare coverage, will result in a large increase in the number of cases. They also expressed concern that the potential increase in patient volume and their perceived inadequate payment for carotid artery stent cases will create a financial hardship on facilities providing this technology, potentially resulting in decreased Medicare beneficiary access to this beneficial therapy.

Response: We continue to believe that the most appropriate changes to the IPPS and the structure of the DRGs are based on evidence of a significant difference in average costs between technology itself and the DRG where its code is assigned. Because the ICD-9-CM procedure codes are new, we do not have data showing that carotid artery stents are more costly than other cases in DRGs 533 and 534. Further, using codes 39.50 and 39.90 as proxies for carotid artery stenting, we did not observe a substantial difference in average charges between cases using these codes and other cases in the DRGs. For this reason, we do not have sufficient evidence to warrant a DRG change at this time.

In this final rule, we are retaining code 00.61 in DRGs 533 and 534 for FY 2006. We will continue to monitor the Medicare charge data in our annual review of DRGs and the IPPS. g. Extracorporeal Membrane Oxygenation (ECMO)

Extracorporeal membrane oxygenation (ECMO) is a procedure to create a closed chest, heart-lung bypass system by insertion of vascular catheters. Patients receiving this procedure require mechanical ventilation. ECMO is performed for a small number of severely ill patients who are at high risk of dying without this procedure. Most often it is done for neonates with persistent pulmonary hypertension and respiratory failure for whom other treatments have failed, certain severely ill neonates receiving major cardiac procedures or diaphragmatic hernia repair, and certain older children and adults, most of whom are receiving major cardiac procedures.

Prior to the proposed rule, we received several letters from institutions that perform ECMO. The commenters stated that, in the CMS GROUPER logic, this procedure has little or no impact on the DRG assignment in the newborn, pediatric, and adult population. According to these letters, patients receiving ECMO are highly resource intensive and should have a unique DRG that reflects the costs of these resources. The commenters recommended the creation of a new DRG for ECMO with a DRG weight equal to or greater than the DRG weight for tracheostomy.

ECMO is assigned to procedure code 39.65 (Extracorporeal membrane oxygenation). This code is classified as an O.R. procedure and is assigned to DRG 104 (Cardiac Valve and Other Major Cardiothoracic Procedure With Cardiac Catheterization) and DRG 105 (Cardiac Valve and Other Major Cardiothoracic Procedure Without Cardiac Catheterization). When ECMO is performed with other O.R. procedures, the case is assigned to the higher weighted DRG. For example, when ECMO and a tracheostomy are performed during the same admission, the case would be assigned to DRG 541 (Tracheostomy with Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth, and Neck Diagnoses With Major O.R.).

We note that the primary focus of updates to the Medicare DRG classification system is changes relating to the Medicare patient population, not the pediatric patient population.

[[Page 47302]]

Because ECMO is primarily a pediatric procedure and rarely performed in an adult population, we have few cases in our data to use to evaluate resource costs. We are aware that other insurers sometimes use Medicare's rates to make payments. We advise private insurers to make appropriate modifications to our payment system when it is being used for children or other patients who are not generally found in the Medicare population.

To evaluate the appropriateness of payment under the current DRG assignment, we have reviewed the FY 2004 MedPAR data and found 78 ECMO cases in 13 DRGs.

The following table illustrates the results of our findings:

Average Number of Average length Average charges for DRG with code 39.65 reported

cases

of stay

charges for all cases in ECMO cases

the DRG

4..............................................

23

9

$147,766

$120,496 105............................................

21

8

131,700

89,831 541............................................

14

62.9

561,210

273,656 All Other DRGs.................................

20

18.1

308,341

NA

The average charges for all ECMO cases were approximately $258,821, and the average length of stay was approximately 20.7 days. The average charges for the ECMO cases are closer to the average charges for DRG 541 ($273,656) than to the average charges of DRG 104 ($147,766) and DRG 105 ($131,700). Of the 78 ECMO cases, 14 cases are already assigned to DRG 541. We believe that the data indicate that DRG 541 would be a more appropriate DRG assignment for cases where ECMO is performed. We further note that under the All Payer DRG System used in New York State, cases involving ECMO are assigned to the tracheostomy DRG. Thus, the assignment of ECMO cases to the tracheostomy DRG for Medicare would be similar to how these cases are grouped in another DRG system. For these reasons, we proposed to reassign ECMO cases reporting code 39.65 to DRG 541. We also proposed to change the title of DRG 541 to: ``ECMO or Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth and Neck Diagnoses With Major O.R. Procedure''.

Comment: Several commenters supported the proposed modification to ECMO cases reporting code 39.65 to DRG 541.

Response: We appreciate the commenters' support.

Accordingly, in this final rule, we are adopting as final the proposed change to ECMO cases reporting code 39.65 to DRG 541 with minor modification. To further clarify the change, we are changing the title of DRG 541 to ``ECMO or Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth, and Neck With Major O.R.'' This title has been modified since the proposed rule (70 FR 23324) to delete the term ``Diagnoses'' from the title. For consistency purposes, we are also changing the DRG title for DRG 542 from ``Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth, and Neck Diagnoses Without Major O.R. Procedure'' to ``Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth, and Neck Without Major O.R.'' 6. MDC 6 (Diseases and Disorders of the Digestive System): Artificial Anal Sphincter

In the FY 2003 IPPS final rule (67 FR 50242), we created two new codes for procedures involving an artificial anal sphincter, effective for discharges occurring on or after October 1, 2002: Code 49.75 (Implantation or revision of artificial anal sphincter) is used to identify cases involving implantation or revision of an artificial anal sphincter and code 49.76 (Removal of artificial anal sphincter) is used to identify cases involving the removal of the device. In Table 6B of that final rule, we assigned both codes to one of four MDCs, based on principal diagnosis, and one of six DRGs within those MDCs: MDC 6 (Diseases and Disorders of the Digestive System), DRGs 157 and 158 (Anal and Stomal Procedures With and Without CC, respectively); MDC 9 (Diseases and Disorders of the Skin, Subcutaneous Tissue and Breast), DRG 267 (Perianal and Pilonidal Procedures); MDC 21 (Injuries, Poisonings, and Toxic Effects of Drugs), DRGs 442 and 443 (Other O.R. Procedures for Injuries With and Without CC, respectively); and MDC 24 (Multiple Significant Trauma), DRG 486 (Other O.R. Procedures for Multiple Significant Trauma).

In the FY 2004 IPPS final rule (68 FR 45372), we discussed the assignment of these codes in response to a request we received to consider reassignment of these two codes to different MDCs and DRGs. The requester believed that the average charges ($44,000) for these codes warranted reassignment. In the FY 2004 IPPS final rule, we stated that we did not have sufficient MedPAR data available on the reporting of codes 49.75 and 49.76 to make a determination on DRG reassignment of these codes. We agreed that, if warranted, we would give further consideration to the DRG assignments of these codes because it is our customary practice to review DRG assignment(s) for newly created codes to determine clinical coherence and similar resource consumption after we have had the opportunity to collect MedPAR data on utilization, average lengths of stay, average charges, and distribution throughout the system. In the FY 2005 IPPS final rule, we reviewed the FY 2003 MedPAR data for the presence of codes 49.75 and 49.76 and determined that these procedures were not a clinical match with the other procedures in DRGs 157 and 158. Therefore, for FY 2005, we moved procedure codes 49.75 and 49.76 out of DRGs 157 and 158 and into DRGs 146 and 147 (Rectal Resection With and Without CC, respectively). This change had the effect of doubling the payment for the cases with procedure codes 49.75 and 49.76 assigned to DRGs 146 and 147 based on increases in the relative weights. One commenter suggested that we create a new DRG for ``Complex Anal/Rectal Procedure with Implant.'' However, we noted that the DRG structure is a system of averages and is based on groups of patients with similar characteristics. At that time, we indicated that we would continue to monitor procedure codes 49.75 and 49.76 and the DRGs to which they are assigned.

For the FY 2006 proposed rule, we reviewed the FY 2004 MedPAR data for the presence of codes 49.75 and 49.76. We found that these two procedures are still of low incidence. Among the six possible DRG assignments, we found a total of 18 cases reported with codes 49.75 and 49.76 for the implant,

[[Page 47303]]

revision, or removal of the artificial anal sphincter. We found 13 of these cases in DRGs 146 and 147 (compared to 12,558 total cases in these DRGs), and the remaining 5 cases in DRGs 442 and 443 (compared to 19,701 total cases in these DRGs).

We believe the number of cases with codes 49.75 and 49.76 in these DRGs is too low to provide meaningful data of statistical significance. Therefore, we did not propose any further changes to the DRGs for these procedures at this time. Neither did we propose to change the structure of DRGs 146 or 147 at this time.

Comment: One commenter agreed that we should maintain the current DRG assignment for codes 49.75 and 49.76. The commenter recommended that CMS continue to monitor the use of these codes and their DRG assignment.

Response: We acknowledge the support of the commenter and will continue to monitor utilization of the services with codes 49.75 and 49.76.

For FY 2006, we are retaining codes 49.75 and 49.76 within DRGs 146 and 147, as proposed. 7. MDC 8 (Diseases and Disorders of the Musculoskeletal System and Connective Tissue) a. Hip and Knee Replacements

Orthopedic surgeons representing the American Association of Orthopaedic Surgeons (AAOS) requested that we subdivide DRG 209 (Major Joint and Limb Reattachment Procedures of Lower Extremity) in MDC 8 by creating a new DRG for revision of lower joint procedures. The AAOS made a presentation at the October 7-8, 2004 meeting of the ICD-9-CM Coordination and Maintenance Committee meeting. A summary report of this meeting can be found at the CMS Web site: http://www. cms. hhs.

gov/ payment systems/ icd9/. We also received written comments on this request prior to the issuance of the FY 2006 IPPS proposed rule.

The AAOS surgeons stated that cases involving patients who require a revision of a prior replacement of a knee or hip require significantly more resources than cases in which patients receive an initial joint replacement. They pointed out that total joint replacement is one of the most commonly performed and successful operations in orthopedic surgery. The surgeons mentioned that, in 2002, over 300,000 hip replacement and 350,000 knee replacement procedures were performed in the United States. They also pointed out that these procedures are a frequent reason for Medicare hospitalization. The surgeons stated that total joint replacements have been shown to be highly cost-effective procedures, resulting in dramatic improvements in quality of life for patients suffering from disabling arthritic conditions involving the hip or knee. In addition, they reported that the medical literature indicates success rates of greater than 90 percent for implant survivorship, reduction in pain, and improvement in function at a 10- to 15-year followup. However, despite these excellent results with primary total joint replacement, factors related to implant longevity and evolving patient demographics have led to an increase in the volume of revision total joint procedures performed in the United States over the past decade.

Total hip replacement is an operation that is intended to reduce pain and restore function in the hip joint by replacing the arthritic hip joint with a prosthetic ball and socket joint. The prosthetic hip joint consists of a metal alloy femoral component with a modular femoral head made of either metal or ceramic (the ``ball'') that articulates with a metal acetabular component with a modular liner made of either metal, ceramic, or high-density polyethylene (the ``socket'').

The AAOS surgeons stated that, in a normal knee, four ligaments help hold the bones in place so that the joint works properly. When a knee becomes arthritic, these ligaments can become scarred or damaged. During knee replacement surgery, some of these ligaments, as well as the joint surfaces, are substituted or replaced by the new artificial prostheses. Two types of fixation are used to hold the prostheses in place. Cemented designs use polymethyl methacrylate to hold the prostheses in place. Cementless designs rely on bone growing into the surface of the implant for fixation.

The surgeons stated that all hip and knee replacements have an articular bearing surface that is subject to wear (the acetabular bearing surface in the hip and the tibial bearing surface in the knee). Traditionally, these bearing surfaces have been made of metal-on-metal or metal-on-polyethylene, although newer materials (both metals and ceramics) have been used more recently. Earlier hip and knee implant designs had nonmodular bearing surfaces, but later designs included modular articular bearing surfaces to reduce inventory and potentially simplify revision surgery. Wear of the articular bearing surface occurs over time and has been found to be related to many factors, including the age and activity level of the patient. In some cases, wear of the articular bearing surface can produce significant debris particles that can cause peri-prosthetic bone resorption (also known as osteolysis) and mechanical loosening of the prosthesis. Wear of the bearing surface can also lead to instability or prosthetic dislocation, or both, and is a common cause of revision hip or knee replacement surgery.

Depending on the cause of failure of the hip replacement, the type of implants used in the previous surgery, the amount and quality of the patient's remaining bone stock, and factors related to the patient's overall health and anatomy, revision hip replacement surgery can be relatively straightforward or extremely complex. Revision hip replacement can involve replacing any part or all of the implant, including the femoral or acetabular components, and the bearing surface (the femoral head and acetabular liner), and may involve major reconstruction of the bones and soft tissues around the hip. All of these procedures differ significantly in their clinical indications, outcomes, and resource intensity.

The AAOS surgeons provided the following summary of the types of revision knee replacement procedures: Among revision knee replacement procedures, patients who underwent complete revision of all components had longer operative times, higher complication rates, longer lengths of stay, and significantly higher resource utilization, according to studies conducted by the AAOS. Revision of the isolated modular tibial insert component was the next most resource-intensive procedure, and primary total knee replacement was the least resource-intensive of all the procedures studied.

Isolated Modular Tibial Insert Exchange. Isolated removal and exchange of the modular tibial bearing surface involves replacing the modular polyethylene bearing surface without removing the femoral, tibial, or patellar components of the prosthetic joint. Common indications for this procedure include wear of the polyethylene bearing surface or instability (for example, looseness) of the prosthetic knee joint. Patient recovery times are much shorter with this procedure than with removal and exchange of either the tibial, femoral, or patellar components.

Revision of the Tibial Component. Revision of the tibial component involves removal and exchange of the entire tibial component, including both the metal base plate and the modular polyethylene bearing surface. Common indications for tibial component revision are wear of the modular bearing surface, aseptic loosening (often

[[Page 47304]]

associated with osteolysis), or infection. Depending on the amount of associated bone loss and the integrity of the ligaments around the knee, tibial component revision may require the use of specialized implants with stems that extend into the tibial canal and/or the use of metal augments or bone graft to fill bony defects.

Revision of the Femoral Component. Revision of the femoral component involves removal and exchange of the metal implant that covers the end of the thigh-bone (the distal femur). Common indications for femoral component revision are aseptic loosening with or without associated osteolysis/bone loss, or infection. Similar to tibial revision, femoral component revision that is associated with extensive bone loss often involves the use of specialized implants with stems that extend into the femoral canal and/or the use of metal augments or bone graft to fill bony defects.

Revision of the Patellar Component. Complications related to the patella-femoral joint are one of the most common indications for revision knee replacement surgery. Early patellar implant designs had a metal backing covered by high-density polyethylene; these implants were associated with a high rate of failure due to fracture of the relatively thin polyethylene bearing surface. Other common reasons for isolated patellar component revision include poor tracking of the patella in the femoral groove leading to wear and breakage of the implant, fracture of the patella with or without loosening of the patellar implant, rupture of the quadriceps or patellar tendon, and infection.

Revision of All Components (Tibial, Femoral, and Patellar). The most common type of revision knee replacement procedure is a complete total knee revision. A complete revision of all implants is more common in knee replacements than hip replacements because the components of an artificial knee are not compatible across vendors or types of prostheses. Therefore, even if only one of the implants is loose or broken, a complete revision of all components is often required in order to ensure that the implants are compatible. Complete total knee revision often involves extensive surgical approaches, including osteotomizing (for example, cutting) the tibia bone in order to adequately expose the knee joint and gain access to the implants. These procedures often involve extensive bone loss, requiring reconstruction with specialized implants with long stems and metal augments or bone graft to fill bony defects. Depending on the status of the ligaments in the knee, complete total knee revision at times requires implantation of a highly constrained or ``hinged'' knee replacement in order to ensure stability of the knee joint.

Reimplantation from previous resection or cement spacer. In cases of deep infection of a prosthetic knee, removal of the implants with implantation of an antibiotic-impregnated cement spacer, followed by 6 weeks of intravenous antibiotics is often required in order to clear the infection. Revision knee replacement from an antibiotic impregnated cement spacer often involves complex bony reconstruction due to extensive bone loss that occurs as a result of the infection and removal of the often well-fixed implants. As noted above, the clinical outcomes following revision from a spacer are often poor due to limited functional capacity while the spacer is in place, prolonged periods of protected weight bearing (following reconstruction of extensive bony defects), and the possibility of chronic infection.

The surgeons stated that the current ICD-9-CM codes did not adequately capture the complex nature of revisions of hip and knee replacements. Currently, code 81.53 (Revision of hip replacement) captures all ``partial'' and ``total'' revision hip replacement procedures. Code 81.55 (Revision of knee replacement) captures all revision knee replacement procedures. These two codes currently capture a wide variety of procedures that differ in their clinical indications, resource intensity, and clinical outcomes.

An AAOS representative made a presentation at the October 7-8, 2004 ICD-9-CM Coordination and Maintenance Committee. Based on the comments received at the October 7-8, 2004 meeting and subsequent written comments, new ICD-9-CM procedure codes were developed to better capture the variety of ways that revision of hip and knee replacements can be performed: Codes 00.70 through 00.73 and code 81.53 for revisions of hip replacements and codes 00.80 through 00.84 and code 81.55 for revisions of knee replacements. These new and revised procedure codes, which will be effective on October 1, 2005, can be found in Table 6B and Table 6F of this final rule. The commenters stated that claims data using these new and specific codes should provide improved data on these procedures for future DRG modifications.

However, the commenters requested that CMS consider DRG modifications based on current data using the existing revision codes. The commenters reported on a recently completed study comparing detailed hospital resource utilization and clinical characteristics in over 10,000 primary and revision hip and knee replacement procedures at 3 high volume institutions: The Massachusetts General Hospital, the Mayo Clinic, and the University of California at San Francisco. The purpose of this study was to evaluate differences in clinical outcomes and resource utilization among patients who underwent different types of primary and revision hip or knee replacement procedures. The study found significant differences in operative time, complication rates, hospital length of stay, discharge disposition, and resource utilization among patients who underwent different types of revision hip or knee replacement procedures.

Among revision hip replacement procedures, patients who underwent both femoral and acetabular component revision had longer operative times, higher complication rates, longer lengths of stay, significantly higher resource utilization, and were more likely to be discharged to a subacute care facility. Isolated femoral component revision was the next most resource-intensive procedure, followed by isolated acetabular revision. Primary hip replacement was the least resource intensive of all the procedures studied. Similarly, among revision knee replacement procedures, patients who underwent complete revision of all components had longer operative times, higher complication rates, longer lengths of stay, and significantly higher resource utilization. Revision of one component was the next most resource-intensive procedure. Primary total knee replacement was the least resource intensive of all the procedures studied.

In addition, the commenters indicated that the data showed that extensive bone loss around the implants and the presence of a peri- prosthetic fracture were the most significant predictors of higher resource utilization among all revision hip and knee replacement procedures, even when controlling for other significant patient and procedural characteristics.

For the FY 2006 IPPS proposed rule, we examined data in the FY 2004 MedPAR file on the current hip replacement procedures (codes 81.51, 81.52, 81.53) as well as the replacements and revisions of knee replacement procedures (codes 81.54 and 81.55) in DRG 209. We found that revisions were significantly more resource intensive than the original hip and knee replacements. We found average charges for revisions of hip and knee replacements were approximately

[[Page 47305]]

$7,000 higher than average charges for the original joint replacements, as shown in the following charts. The average charges for revisions of hip replacements were 21 percent higher than the average charges for initial hip replacements. The average charges for revisions of knee replacements were 25 percent higher than for initial knee replacements.

Average length DRG

Number of

of stay

Average cases

(days)

charges

209--All cases..................................................

430,776

4.57 $30,695.41 209 With hip replacement codes 81.51 and 81.52 reported.........

181,460

5.21 31,795.84 209 With hip revision code 81.53 reported.......................

20,894

5.57 38,432.04 209 With knee replacement code 81.54 reported...................

209,338

3.92 28,525.66 209 With knee revision code 81.55 reported......................

18,590

4.64 35,671.66

We note that there were no cases in DRG 209 for reattachment of the foot, lower leg, or thigh (codes 84.29, 84.27, and 84.28).

To address the higher resource costs associated with hip and knee revisions relative to the initial joint replacement procedure, we proposed to delete DRG 209, create a proposed new DRG 544 (Major Joint Replacement or Reattachment of Lower Extremity), and create a proposed new DRG 545 (Revision of Hip or Knee Replacement).

We proposed to assign the following codes to the new proposed DRG 544: 81.51, 81.52, 81.54, 81.56, 84.26, 84.27, and 84.28.

We proposed to assign the following codes to the proposed new DRG 545: 00.70, 00.71, 00.72, 00.73, 00.80, 00.81, 00.82, 00.83, 00.84, 81.53, and 81.55.

In response to the FY 2006 IPPS proposed rule, we received the following public comments:

Comment. Four commenters supported our proposal to delete DRG 209 and to create proposed new DRGs 544 and DRG 545. One commenter stated that the proposed rule reveals that the average joint revision charges are $7,000 higher than original joint replacements, which supports the point that joint revision procedures are more resource-intensive than initial replacements.

Another commenter commended CMS for its efforts to provide appropriate payment for revision hip and knee arthroplasty by proposing to split DRG 209 into DRG 544 and 545, and to expand the scope of the relevant ICD-9-CM procedure codes included in these DRGs. The commenter stated that the new codes, in particular, are an important component in aligning hospital reimbursement with hospital costs and patient benefits of total joint arthroplasty. The commenter encouraged CMS to continue its dialogue with industry and providers regarding further DRG changes to primary joint arthroplasty procedures, which represent approximately 90 percent of total hip and knee arthroplasty procedures.

One commenter recommended that CMS consider the number of individual components used in the joint replacement when future DRG revisions are made. The commenter stated the hospital's costs will vary based on the number of parts replaced during the procedure. According to the commenter, we may be overpaying simple head and/or liner exchanges in hips, and patellar/insert exchanges in knees relative to primary hip and knee procedures. The commenter indicated that, with the more specific ICD-9-CM codes, CMS will be able to evaluate further changes in the joint replacement and revision DRGs.

We did not receive any comments that opposed the proposed DRG revisions for hip and knee replacements.

Response: We appreciate the support of the commenters. We will use the data obtained from use of the new codes to consider future DRG revisions for joint replacement and revision procedures.

In this final rule, for FY 2006, we are adopting the DRG revisions relating to hip and knee replacements as proposed. We are deleting DRG 209 and creating new DRG 544 (Major Joint Replacement or Reattachment of Lower Extremity) and new DRG 545 (Revision of Hip or Knee Replacement). The new DRG 544 includes the following code assignments:

81.51, Total hip replacement

81.52, Partial hip replacement

81.54, Total knee replacement

81.56, Total ankle replacement

84.26, Foot reattachment

84.27, Lower leg/ankle reattach

84.28, Thigh reattachment

The new DRG 545 includes the following code assignments:

00.70, Revision of hip replacement, both acetabular and femoral components

00.71, Revision of hip replacement, acetabular component

00.72, Revision of hip replacement, femoral component

00.73, Revision of hip replacement, acetabular liner and/ or femoral head only

00.80, Revision of knee replacement, total (all components)

00.81, Revision of knee replacement, tibial component

00.82, Revision of knee replacement, femoral component

00.83, Revision of knee replacement, patellar component

00.84, Revision of knee replacement, tibial insert (liner)

81.53, Revision of hip replacement, not otherwise specified

81.55, Revision of knee replacement, not otherwise specified

We believe that the creation of the new DRGs for revisions of hip and knee replacements should resolve payment issues for hospitals that perform the more difficult revisions of joint replacements. In addition, as stated earlier, we have worked with the orthopedic community to develop new procedure codes that better capture data on the types of revisions of hip and knee replacements. These new codes will be implemented on October 1, 2005. Once we receive claims data using these new codes, we will review data to determine if additional DRG modifications are needed. This effort may include assigning some of the revision codes, such as 00.83 and 00.84, to a separate DRG. As stated earlier, the AAOS has found that some of the procedures may not be as resource intensive. Therefore, the AAOS has requested that CMS closely examine data from the use of the new codes and consider future revisions. b. Kyphoplasty

In the FY 2005 IPPS final rule (69 FR 48938), we discussed the creation of new codes for vertebroplasty (81.65) and kyphoplasty (81.66), which went into effect on October 1, 2004. Prior to October 1, 2004, both of these surgical procedures were assigned to code 78.49 (Other repair or plastic operation on bone). For FY 2005, we assigned these codes to DRGs 233 and 234 (Other

[[Page 47306]]

Musculoskeletal System and Connective Tissue O.R. Procedure With and Without CC, respectively) in MDC 8 (Table 6B of the FY 2005 final rule). (In the FY 2005 IPPS final rule (69 FR 48938), we indicated that new codes 81.65 and 81.66 were assigned to DRGs 223 and 234. We made a typographical error when indicating that these codes were assigned to DRG 223. Codes 81.65 and 81.66 have been assigned to DRGs 233 and 234.) Last year, we received comments opposing the assignment of code 81.66 to DRGs 233 and 234. The commenters supported the creation of the codes for kyphoplasty and vertebroplasty, but recommended that code 81.66 be assigned to DRGs 497 and 498 (Spinal Fusion Except Cervical With and Without CC, respectively). The commenters stated that kyphoplasty requires special inflatable bone tamps and bone cement and is a significantly more resource intensive procedure than vertebroplasty. The commenters further stated that, while kyphoplasty involves internal fixation of the spinal fracture and restoration of vertebral heights, vertebroplasty involves only fixation. The commenters indicated that hospital costs for kyphoplasty procedures are more similar to resources used in a spinal fusion.

We stated in the FY 2005 IPPS final rule that we did not have data in the MedPAR file on kyphoplasty and vertebroplasty. Prior to October 1, 2004, both procedures were assigned in code 78.49, which was assigned to DRGs 233 and 234 in MDC 8. We stated that we would continue to review this area as part of our annual review of MedPAR data. While we do not have separate data for kyphoplasty because code 81.66 was not established until October 1, 2004, for the FY 2006 IPPS proposed rule, we did examine data on code 78.49, which includes both kyphoplasty and vertebroplasty procedures reported in DRGs 233 and 234. The following chart illustrates our findings:

Number of Average length Average DRG

cases of stay (days) charges

233--All cases..................................................

14,066

6.66 $28,967.78 233 With code 78.49 reported....................................

8,702

5.91 25,402.71 233 Without code 78.49 reported.................................

5,364

7.88 34,571.39 234--All cases..................................................

7,106

2.79 18,954.80 234 With code 78.49 reported....................................

4,437

2.61 18,426.11 234 Without code 78.94 reported.................................

2,669

3.09 19,833.71

We do not believe these data findings support moving cases represented by code 78.49 out of DRGs 233 and 234. While we cannot distinguish cases that are kyphoplasty from cases that are vertebroplasty, cases represented by code 78.49 have lower charges than do other cases within DRGs 233 and 234. Therefore, in the FY 2006 IPPS proposed rule, we did not propose to change the DRG assignment of code 81.66 to DRGs 233 and 234.

Comment: Two commenters supported our proposal not to change the DRG assignment of code 81.66 (Kyphoplasty). Both commenters agreed with our proposal to keep code 81.66 in DRGs 233 and 234. They also agreed that we should wait for bill data using the new kyphoplasty code prior to considering any DRG modification.

Response: We appreciate the commenters' support for our proposal.

In this final rule, for FY 2006, we are retaining the assignment of code 81.66 in DRGs 233 and 234. As we proposed, we will consider whether further changes are warranted once additional hospital charge data are available using the new code. c. Multiple Level Spinal Fusion

On October 1, 2003, the following ICD-9-CM codes were created to identify the number of levels of vertebra fused during a spinal fusion procedure:

81.62, Fusion or refusion of 2-3 vertebrae

81.63, Fusion or refusion of 4-8 vertebrae

81.64, Fusion or refusion of 9 or more vertebrae

Prior to the creation of these codes, we received a comment recommending the establishment of new DRGs that would be differentiated based on the number of vertebrae fused. In the FY 2005 IPPS final rule (69 FR 48936), we stated that we did not yet have any reported cases utilizing these multiple level spinal fusion codes. We stated that we would wait until sufficient data were available prior to making a final determination on whether to create separate DRGs based on the number of vertebrae fused. We also stated that spinal fusion surgery was an area undergoing rapid changes.

Effective October 1, 2004, we created a series of codes that describe a new type of spinal surgery, spinal disc replacement. Our medical advisors describe these procedures as a more conservative approach for back pain than the spinal fusion surgical procedure. These codes are as follows:

84.60, Insertion of spinal disc prosthesis, not otherwise specified

84.61, Insertion of partial spinal disc prosthesis, cervical

84.62, Insertion of total spinal disc prosthesis, cervical

84.63, Insertion of spinal disc prosthesis, thoracic

84.64, Insertion of partial spinal disc prosthesis, lumbosacral

84.65, Insertion of total spinal disc prosthesis, lumbosacral

84.66, Revision or replacement of artificial spinal disc prosthesis, cervical

84.67, Revision or replacement of artificial spinal disc prosthesis, thoracic

84.68, Revision or replacement of artificial spinal disc prosthesis, lumbosacral

84.69, Revision or replacement of artificial spinal disc prosthesis, not otherwise specified

We also created the following two codes effective October 1, 2004, for these new types of spinal surgery that are also a more conservative approach to back pain than is spinal fusion:

81.65, Vertebroplasty

81.66, Kyphoplasty

We do not yet have data in the MedPAR file on these new types of procedures. Therefore, we cannot yet determine what effect these new types of procedures will have on the frequency of spinal fusion procedures.

However, we do have data in the MedPAR file on multiple level spinal procedures for analysis for this year's IPPS rule. We examined data in the FY 2004 MedPAR file on spinal fusion cases in the following DRGs:

DRG 496 (Combined Anterior/Posterior Spinal Fusion)

DRG 497 (Spinal Fusion Except Cervical With CC)

DRG 498 (Spinal Fusion Except Cervical Without CC)

DRG 519 (Cervical Spinal Fusion With CC)

[[Page 47307]]

DRG 520 (Cervical Spinal Fusion Without CC)

Multiple level spinal fusion is captured by code 81.63 (Fusion or refusion of 4-8 vertebrae) and code 81.64 (Fusion or refusion of 9 or more vertebrae). Code 81.62 includes the fusion of 2-3 vertebrae and is not considered a multiple level spinal fusion. Orthopedic surgeons stated at the October 7-8, 2004 ICD-9-CM Coordination and Maintenance Committee meeting that the most simple and common type of spinal fusion involves fusing either 2 or 3 vertebrae. These surgeons stated that there was not a significant difference in resource utilization for cases involving the fusion of 2 versus 3 vertebrae. For this reason, the orthopedic surgeons recommended that fusion of 2 and 3 vertebrae remain grouped into one ICD-9-CM code.

We reviewed the Medicare charge data to determine whether the number of vertebrae fused or specific diagnoses have an effect on average length of stay and resource use for a patient. We found that, while fusing 4 or more levels of the spine results in a small increase in the average length of stay and a somewhat larger increase in average charges for spinal fusion patients, an even greater impact was made by the presence of a principal diagnosis of curvature of the spine or malignancy. The following list of diagnoses describes conditions that have a significant impact on resource use for spinal fusion patients:

170.2, Malignant neoplasm of vertebral column, excluding sacrum and coccyx

198.5, Secondary malignant neoplasm of bone and bone marrow

732.0, Juvenile osteochondrosis of spine

733.13, Pathologic fracture of vertebrae

737.0, Adolescent postural kyphosis

737.10, Kyphosis (acquired) (postural)

737.11, Kyphosis due to radiation

737.12, Kyphosis, postlaminectomy

737.19, Kyphosis (acquired), other

737.20, Lordosis (acquired) (postural)

737.21, Lordosis, postlaminectomy

737.22, Other postsurgical lordosis

737.29, Lordosis (acquired), other

737.30, Scoliosis [and kyphoscoliosis], idiopathic

737.31, Resolving infantile idiopathic scoliosis

737.32, Progressive infantile idiopathic scoliosis

737.33, Scoliosis due to radiation

737.34, Thoracogenic scoliosis

737.39, Other kyphoscoliosis and scoliosis

737.40, Curvature of spine, unspecified

737.41, Curvature of spine associated with other conditions, kyphosis

737.42, Curvature of spine associated with other conditions, lordosis

737.43, Curvature of spine associated with other conditions, scoliosis

737.8, Other curvatures of spine

737.9, Unspecified curvature of spine

754.2, Congenital scoliosis

756.51, Osteogenesis imperfecta

The majority of fusion patients with these diagnoses were in DRGs 497 and 498. The chart below reflects our findings. We also include in the chart statistics for cases in DRGs 497 and 498 with spinal fusion of 4 or more vertebrae and cases with a principal diagnosis of curvature of the spine or bone malignancy.

Average length DRG

Number of

of stay

Average cases

(days)

charges

497.............................................................

27,346

6.08 $64,471.82 498.............................................................

17,943

3.80 48,440.80 497 and 498 With spinal fusions of 4 or more vertebrae reported.

7,881

6.3

77,352.00 497 and 498 With principal diagnosis of curvature of the spine

2,006

8.91 95,315.00 or bone malignancy.............................................

Thus, these diagnoses result in a significant increase in resource use. While the fusing of 4 or more vertebrae resulted in average charges of $77,352, the impact of a principal diagnosis of curvature of the spine or bone malignancy was substantially greater with average charges of $95,315.

Based on this analysis, we proposed to create a new DRG 546 for noncervical spinal fusions with a principal diagnosis of curvature of the spine and malignancies: proposed new DRG 546 (Spinal Fusions Except Cervical With Principal Diagnosis of Curvature of the Spine or Malignancy). We proposed to include in the proposed new DRG cases all noncervical spinal fusions cases previously assigned to DRGs 497 and 498 that have a principal diagnosis of curvature of the spine or malignancy and with the following codes listed above: 170.2, 198.5, 732.0, 733.13, 737.0, 737.10, 737.11, 737.12, 737.19, 737.20, 737.21, 737.22, 737.29, 737.30, 737.31, 737.32, 737.33, 737.34, 737.39, 737.40, 737.41, 737.42, 737.43, 737.8, 737.9, 754.2, and 756.51. We proposed that the proposed DRG 546 would not include cases currently assigned to DRGs 496, 519, or 520 that have a principal diagnosis of curvature of the spine or malignancy and that the structure of DRGs 496, 519, and 520 would remain the same.

As part of our meeting with the AAOS on DRG 209 in February 2005 (discussed under section II.B.6.a. of this preamble), the AAOS offered to work with CMS to analyze clinical issues and make revisions to the spinal fusion DRGs (DRGs 496 through 498 and 519 and 520). Therefore, we limited our proposed changes to the spinal fusion DRGs for FY 2006 to the creation of the proposed DRG 546 discussed above. However, we indicated that we look forward to working with the AAOS to obtain its clinical recommendations concerning our proposed changes and potential additional modifications to the spinal fusion DRGs. We also solicited comments from the public on our proposed changes and how to incorporate new types of spinal procedures such as kyphoplasty and spinal disc prostheses into the spinal fusion DRGs.

Comment: A number of commenters supported our proposal to create new DRG 546 (Spinal Fusions Except Cervical With Principal Diagnosis of Curvature of the Spine or Malignancy) to include all noncervical spinal fusions previously assigned to DRGs 497 and 498 that have a principal diagnosis of curvature of the spine or malignancy. One commenter stated that the addition of new DRG 546, with its higher weight, would help reimburse hospitals more adequately for the resources used in treating patients with significant spinal deformities and other problems. One commenter stated that the cost associated with a multilevel spine fusion when the patient has a diagnosis of curvature of the spine or malignancy exceeds the current Medicare reimbursement.

Several commenters noted that the following four ICD-9-CM diagnosis

[[Page 47308]]

codes are manifestation codes that cannot be reported as a principal diagnosis:

737.40, Curvature of spine, unspecified

737.41, Curvature of spine associated with other conditions, kyphosis

737.42, Curvature of spine associated with other conditions, lordosis

737.43, Curvature of spine associated with other conditions, scoliosis

The commenter pointed out that these codes can only be reported as a secondary diagnosis. Therefore, the commenters stated that our proposed DRG logic for DRG 546 would not work with these four codes.

Response: We appreciate the support of the commenters for the creation of the new DRG 546. We agree that this new DRG would better align Medicare payment with hospital costs for treating these more severe orthopedic cases. We also agree that codes 737.40, 737.41, 737.42, and 737.43 are not to be reported as a principal diagnosis because they are manifestation codes. We inadvertently included them among the list of principal diagnoses that would be assigned to DRG 546. In this final rule, we are removing codes 737.40, 737.41, 737.42, and 737.43 from the list of principal diagnosis codes that would lead to an assignment of DRG 546. However, we will retain these codes as a secondary diagnosis that will result in an assignment to DRG 546 because they describe curvature of the spine. Therefore, patients admitted with an orthopedic diagnosis who receive a spinal fusion will be assigned to DRG 546 if codes 737.40, 737.41, 737.42, and 737.43 are present as a secondary diagnosis. Consistent with this change in the GROUPER logic, we will also remove the term ``principal diagnosis'' from the proposed title so that DRG 546 will be titled ``Spinal Fusions Except Cervical With Curvature of the Spine or Malignancy.''

Comment: One commenter suggested that CMS consider adding the following diagnoses to the list of codes that would be assigned to the new DRG 546:

213.2, Benign neoplasm of bone and articular cartilage; vertebral column, excluding sacrum and coccyx

238.0, Neoplasm of uncertain behavior of other and unspecified sites and tissues; Bone and articular cartilage

239.2, Neoplasms of unspecified nature; Bone, soft tissue, and skin

721.7, Spondylosis and allied disorders; Traumatic spondylopathy

724.3, Other and unspecified disorders of back; Sciatica

732.8, Other specified forms of osteochondropathy

756.19, Anomalies of spine; Other

Response: We discussed these additional diagnosis codes recommended by the commenter with our medical advisors and they agree that the first three listed codes (213.2, 238.0, and 239.2) should be added because they are neoplasm codes. Therefore, they are clinically similar to the other neoplasm codes on our proposed list. Our medical advisors did not support the addition on the latter four codes because they are vague codes that do not necessarily represent significant conditions. Therefore, in this final rule, we are adding codes 213.2, 238.0, 239.2 to our list of conditions in DRG 546. We are not adding codes 721.7, 724.3, 732.8, or 756.19.

After careful consideration of the public comments received, in this final rule, we are establishing a new DRG 546 (Spinal Fusions Except Cervical with Curvature of the Spine or Malignancy). New DRG 546 will be composed of all noncervical spinal fusions previously assigned to DRGs 497 and 498 that have a principal or secondary diagnosis of curvature of the spine or a principal diagnosis of a malignancy. The principal diagnosis codes that will lead to this DRG assignment are the following:

170.2, Malignant neoplasm of vertebral column, excluding sacrum and coccyx

198.5, Secondary malignant neoplasm of bone and bone marrow

213.2, Benign neoplasm of bone and articular cartilage; vertebral column, excluding sacrum and coccyx

238.0, Neoplasm of uncertain behavior of other and unspecified sites and tissues; Bone and articular cartilage

239.2, Neoplasms of unspecified nature; bone, soft tissue, and skin

732.0, Juvenile osteochondrosis of spine

733.13, Pathologic fracture of vertebrae

737.0, Adolescent postural kyphosis

737.10, Kyphosis (acquired) (postural)

737.11, Kyphosis due to radiation

737.12, Kyphosis, postlaminectomy

737.19, Kyphosis (acquired), other

737.20, Lordosis (acquired) (postural)

737.21, Lordosis, postlaminectomy

737.22, Other postsurgical lordosis

737.29, Lordosis (acquired), other

737.30, Scoliosis [and kyphoscoliosis], idiopathic

737.31, Resolving infantile idiopathic scoliosis

737.32, Progressive infantile idiopathic scoliosis

737.33, Scoliosis due to radiation

737.34, Thoracogenic scoliosis

737.39, Other kyphoscoliosis and scoliosis

737.8, Other curvatures of spine

737.9, Unspecified curvature of spine

754.2, Congenital scoliosis

756.51, Osteogenesis imperfecta

The secondary diagnoses that will lead to the new DRG 546 assignment are:

737.40, Curvature of spine, unspecified

737.41, Curvature of spine associated with other conditions, kyphosis

737.42, Curvature of spine associated with other conditions, lordosis

737.43, Curvature of spine associated with other conditions, scoliosis d. CHARITE\TM\ Spinal Disc Replacement Device

As we noted in our discussion of applications for new technology add-on payments for FY 2006 in section II.E. of the IPPS proposed rule (70 FR 23362), the applicant for new technology for CHARITE\TM\ requested a DRG reassignment for cases involving implantation of the CHARITE\TM\ Artificial Disc. CHARITE\TM\ is a prosthetic intervertebral disc. On October 26, 2004, the FDA approved the CHARITE\TM\ Artificial Disc for single level spinal arthroplasty in skeletally mature patients with degenerative disc disease between L4 and S1. The applicant requested a DRG assignment for these cases from DRG 499 (Back and Neck Procedures Except Spinal Fusion With CC) and 500 (Back and Neck Procedures Except Spinal Fusion Without CC) to DRGs 497 (Spinal Fusion Except Cervical With CC) and 498 (Spinal Fusion Except Cervical Without CC). The applicant argued that the costs of an inpatient stay to implant an artificial disc prosthesis are similar to spinal fusion and inclusion in DRGs 497 and 498 should be made consistent with section 1886(d)(5)(K) of the Act that indicates a clear preference for assigning a new technology to a DRG based on similar clinical or anatomical characteristics and costs. As indicated in section II.E. of this final rule, we did not find that CHARITE\TM\ meets the substantial clinical improvement criterion and are not considering a DRG reassignment under the new technology provisions. However, we did evaluate whether to reassign CHARITE\TM\ to a different DRG using the Secretary's authority under section 1886(d)(4) of the Act.

[[Page 47309]]

On October 1, 2004, we created new codes for the insertion of spinal disc prostheses (codes 84.60 through 84.69). In the FY 2005 IPPS proposed and final rules, we described the new DRG assignments for these new codes in Table 6B of the Addendum to those rules. We received a number of comments on the FY 2005 IPPS proposed rule recommending that we change the assignments for these codes from DRG DRGs 499 and 500 to the DRGs for spinal fusion (DRGs 497 and 498). In the FY 2005 IPPS final rule (69 FR 48938), we indicated that DRGs 497 and 498 are limited to spinal fusion procedures. Because the surgery involving the CHARITE\TM\ is not a spinal fusion, we decided not to include this procedure in these DRGs. However, we stated that we would continue to analyze this issue and solicited further public comments on the DRG assignment for spinal disc prostheses.

We received a number of public comments in response to the FY 2006 proposed rule. A summary of the comments and our responses follow.

Comment: One commenter supported our recommendation to keep the CHARITE\TM\ spinal disc procedure code in DRGs 499 and 500. The commenter took no position on CMS' decision on whether to grant add-on payment for new technology for the CHARITE\TM\ spinal disc procedure. However, the commenter stated that until further data becomes publicly available, it would be premature to reassign spinal disc prostheses to DRGs 497 and 498. The commenter stated that waiting for Medicare data would be consistent with the approach CMS used in considering changes to DRGs 497 and 498 for account for multilevel spinal fusion. (We did not propose a change for FY 2006 to account for multilevel spinal fusions because sufficient data were not available in MedPAR under the new multilevel spine fusion procedure codes.) The commenter also stated that the spinal fusion DRGs were well-established based on several years of utilization and accrual of cost experience. Without a fuller understanding of the expected resource use of cases with spinal disc prostheses, the commenter was concerned that reassignment of these procedures to DRGs 497 and 498 may have the potential to cause an inappropriate reduction in future weights for spinal fusion. Therefore, the commenter recommended that spinal disc replacements be kept in DRGs 499 and 500 until data are available to evaluate this change.

Response: We agree with the commenter that our policy is to assign a new procedure code to a DRG based on the assignment of its predecessor code until we have Medicare charge data to evaluate a DRG modification. We also agree that the spinal fusion cases are well- established based on several years of utilization and cost experience. Without Medicare data that shows Medicare charges for CHARITE\TM\ artificial discs in DRGs 499 and 500 and until we receive Medicare charge data using the new procedure codes, it is difficult to evaluate a request for a DRG modification.

Comment: Eight commenters opposed our proposal of keeping CHARITE\TM\ artificial discs in DRGs 499 and 500 until we received Medicare charge data. These commenters recommended that the CHARITE\TM\ spinal disc procedure (code 84.65) be moved out of DRGs 499 and 500 and into the spinal fusion DRGs (DRG 497 and 498). According to the commenters, the current DRG assignment to DRGs 499 and 500 provides a very significant economic disincentive for hospitals to use CHARITE\TM\ in the Medicare population. Based on information submitted with its new technology application, these commenters argued that hospital resources for patients receiving CHARITE\TM\ artificial discs are most closely comparable to patients in DRGs 497 and 498 (the data provided to support the new technology application are discussed in detail in section II.E. of this final rule). The commenters also stated that the Health Service Cost Review Commission (HSCRC) of Maryland developed new artificial disc DRGs for its DRG system.

Response: With respect to the commenter's point regarding the HSCRC, we acknowledge that they recently decided to create new DRGs for artificial disc patients. We understand that the HSCRC established these new DRGs with relative weights that are higher than DRGs 499 and 500 and less than the spinal fusion DRGs (DRGs 497 and 498). We are unaware of the criteria that the HSCRC uses for creating separate DRGs. Currently, we do not have a basis for creating a separate DRG for spinal disc protheses because we have no FY 2004 Medicare charge data that could be used to set the FY 2006 relative weight. Therefore, we are unable to adopt an option similar to that of the HSCRC at this time.

For its new technology application, we note that the applicant supplied cost data for 376 total cases where CHARITE\TM\ was actually used, including 12 cases involving Medicare patients. The data for the 12 Medicare patients did not come from the MedPAR data systems because that information is not yet available due to the fact that give that FDA approval and the code used to identify these patients was not effective until October 2004. Thus, as with all new technology applications, the data supporting whether the technology meets the cost criterion came directly from the applicant and not from Medicare's data systems. While the applicant also supplied data from the FY 2003 MedPAR file, we note that these cases did not actually involve the CHARITE\TM\ artificial disc. Rather the applicant modified the claims data for spinal fusion cases by removing the medical and surgical costs associated with the spinal fusions. The applicant then replaced these costs with costs represented to be those of a typical CHARITE\TM\ artificial disc. These data are acceptable to evaluate whether a new technology meets the cost criterion in a new technology application because, by definition, there is limited or no Medicare data upon which to evaluate a new technology's costs. However, these data do not meet the standards that we apply for making a change to a DRG. That is, we use the predecessor code for a new technology until we have evidence from Medicare's data systems that suggest a change to the DRG assignment is warranted.

As stated previously, we do not have Medicare charge information to evaluate a DRG change at this time. For this reason, we are not making a change to the DRG assignment for CHARIT[Eacute]TM. However, we will consider whether a DRG reassignment for CHARIT[Eacute]TMis warranted for FY 2007, once we have information from Medicare's data system that will assist us in evaluating the cost of these patients. 8. MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)): Severe Sepsis

As we did for FY 2005, we received a request to consider the creation of a separate DRG for the diagnosis of severe sepsis for FY 2006. Severe sepsis is described by ICD-9-CM code 995.92 (Systemic inflammatory response syndrome due to infection with organ dysfunction). Patients admitted with sepsis as a principal diagnosis currently are assigned to DRG 416 (Septicemia Age >17) and DRG 417 (Septicemia Age 0-17) in MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)). The commenter requested that all cases in which severe sepsis is present on admission, as well as those cases in which it develops after admission (which are currently classified elsewhere), be included in this new DRG. We again addressed this issue in the FY 2006 IPPS proposed rule

[[Page 47310]]

(70 FR 23329) as we had in the FY 2005 IPPS final rule (69 FR 48975). In both instances, we did not believe the current clinical definition of severe sepsis is specific enough to identify a meaningful cohort of patients in terms of clinical coherence and resource utilization to warrant a separate DRG. Sepsis is found across hundreds of medical and surgical DRGs, and the term ``organ dysfunction'' implicates numerous currently existing diagnosis codes. While we recognize that Medicare beneficiaries with severe sepsis are quite ill and require extensive hospital resources, we do not believe that they can be identified adequately to justify removing them from all of the other DRGs in which they appear. For this reason, we did not propose a new DRG for severe sepsis for FY 2006.

Comment: Two commenters expressed concerns about the sequencing instructions for severe sepsis. They pointed out that current ICD-9-CM coding guidelines mandate that a code from category 038.x be sequenced as the principal diagnosis followed by code 995.92 for patients admitted in respiratory failure who also have severe sepsis. The commenters expressed concerns that this sequencing instruction results in lower hospital reimbursement for patients with severe sepsis placed on mechanical ventilation. These commenters did not recommend that CMS create a new DRG for patients with severe sepsis. Instead, they suggested that the codes or guidelines, or both, be modified so that other conditions can be sequenced as the principal diagnosis.

Response: We share the concern of the commenters about sequencing guidelines for patients with severe sepsis and respiratory failure. The current ICD-9-CM codes for systemic inflammatory response syndrome (SIRS), codes 995.91 through 995.94, that include severe sepsis mandate these sequencing guidelines. However, the National Center for Health Statistics (NCHS) discussed modifications to these codes at the April 1, 2005 ICD-9-CM Coordination and Maintenance Committee meeting. NCHS has scheduled this topic for further discussion at the September 29-30, 2005 Committee meeting. Suggestions for revising these codes and any resulting guidelines should be sent to Donna Pickett, NCHS, 3311 Toledo Road, Room 2402, Hyattsville, MD 2082, or to the e-mail address dfp4@cdc.gov.

Comment: One commenter expressed disappointment that CMS did not create a new DRG for severe sepsis. The commenter disagreed with our statement that these patients could not be easily identified within our Medicare data. The commenter stated that severe sepsis is a systemic inflammatory syndrome in response to infection that is associated with acute organ dysfunction. The commenter suggested that CMS use the SIRS ICD-9-CM codes for infection plus organ dysfunction along with an ICD- 9-CM procedure code for organ support such as ventilation management (code 96.7x), acute renal replacement (codes 39.95 and 54.98), or vasopressor support (code 00.17), to identify these patients. The commenter recommended that CMS create two new DRGs, one for medical severe sepsis patients with organ support and another for surgical severe sepsis patients with organ support. The commenter recommended that these two DRGs be assigned as pre-MDCs.

Response: There were extensive discussions about the problems in using the current SIRS codes at the March 31-April 1, 2005 ICD-9-CM Coordination and Maintenance Committee meeting. A summary report of this meeting can be found at the Web site: http://www.cdc.gov/nchs/icd9cm. As stated earlier, NCHS has scheduled further discussions on

this topic for the September 29-30 Committee meeting.

Given the considerable confusion among the coding community regarding the use of these codes, we believe it would be premature to consider new DRGs for severe sepsis patients at this time. Therefore, we are not making revisions to the DRG for severe sepsis patients at this time. We will continue to work with NCHS to improve the codes so that our data on these patients improve. We will continue to examine data on these patients as we consider future modifications. 9. MDC 20 (Alcohol/Drug Use and Alcohol/Drug Induced Organic Mental Disorders): Drug-Induced Dementia

In the FY 2005 IPPS final rule (69 FR 48939, August 11, 2004), we discussed a request that CMS modify DRGs 521 through 523 by removing the principal diagnosis code 292.82 (Drug-induced dementia) from these alcohol and drug abuse DRGs. These DRGs are as follows:

DRG 521 (Alcohol/Drug Abuse or Dependence With CC).

DRG 522 (Alcohol/Drug Abuse or Dependence With Rehabilitation Therapy Without CC).

DRG 523 (Alcohol/Drug Abuse or Dependence Without Rehabilitation Therapy Without CC).

The commenter indicated that a patient who has a drug-induced dementia should not be classified to an alcohol/drug DRG. However, the commenter did not propose a new DRG assignment for code 292.82. Our medical advisors evaluated the request and determined that the most appropriate DRG classification for a patient with drug-induced dementia was within MDC 20. The medical advisors indicated that because the dementia is drug induced, it is appropriately classified to DRGs 521 through 523 in MDC 20. Therefore, we did not propose a new DRG classification for the principal diagnosis code 292.82.

In the FY 2005 IPPS final rule, we addressed a comment from an organization representing hospital coders that disagreed with our decision to keep code 292.82 in DRGs 521 through 523. The commenter stated that DRGs 521 through 523 are described as alcohol/drug abuse and dependence DRGs, and that drug-induced dementia can be caused by an adverse effect of a prescribed medication or a poisoning. The commenter did not believe that assignment to DRGs 521 through 523 was appropriate if the drug-induced dementia is due to one of these events and the patient is not alcohol or drug dependent. The commenter recommended that admissions for drug-induced dementia be classified to DRGs 521 through 523 only if there is a secondary diagnosis indicating alcohol/ drug abuse or dependence.

The commenter recommended that drug-induced dementia that is due to the adverse effect of a drug or poisoning be classified to the same DRGs as other types of dementia, such as DRG 429 (Organic Disturbances and Mental Retardation). The commenter believed that when drug-induced dementia is caused by a poisoning, either accidental or intentional, the appropriate poisoning code would be sequenced as the principal diagnosis and, therefore, these cases would likely already be assigned to DRGs 449 and 450 (Poisoning and Toxic Effects of Drugs, Age Greater than 17, With and Without CC, respectively) and DRG 451 (Poisoning and Toxic Effects of Drugs, Age 0-17). The commenter stated that these would be the appropriate DRG assignments for drug-induced dementia due to a poisoning. We received a similar comment from a hospital organization.

In the FY 2005 IPPS final rule, we acknowledged that the commenters raised additional issues surrounding the DRG assignment for code 292.82 that should be considered. The commenters provided alternatives for DRG assignment based on sequencing of the principal diagnosis and reporting of additional secondary diagnoses. We recognized that patients may develop drug-induced dementia from drugs that

[[Page 47311]]

are prescribed, as well as from drugs that are not prescribed. However, because dementia develops as a result of use of a drug, we believed the current DRG assignment to DRGs 521 through 523 remained appropriate. Some commenters have agreed with the current DRG assignment of code 292.82 since the dementia was caused by use of a drug. We agree that if either accidental or intentional poisoning caused the drug-induced dementia, the appropriate poisoning code should be sequenced as the principal diagnosis. As one commenter stated, these cases would be assigned to DRGs 449 through 451. We encouraged hospitals to examine the coding for these types of cases to determine if there were any coding or sequencing errors. As suggested by the commenter, if code 292.82 were reported as a secondary diagnosis and not a principal diagnosis in cases of poisoning or adverse drug reactions, the number of cases on DRGs 521 through 523 would decline.

In the FY 2005 IPPS final rule, we agreed to analyze this area for FY 2006 and to look at the alternative DRG assignments suggested by the commenters. As indicated in the FY 2006 IPPS proposed rule, we examined data from the FY 2004 MedPAR file on cases in DRGs 521 through 523 with a principal diagnosis of code 292.82. We found that there were only 134 cases reported with the principal diagnosis code 292.82 in DRGs 521 through 523 without a diagnosis of drug and alcohol abuse. The average standardized charges for cases with a principal diagnosis of code 292.82 that did not have a secondary diagnosis of drug/alcohol abuse or dependence were $12,244.35, compared to the average standardized charges for all cases in DRG 521, which were $10,543.69. There were no cases in DRG 522 with a principal diagnosis of code 292.82. We found only 24 cases in DRG 523 with a principal diagnosis of code 292.82. Given the small number of cases in DRG 522 and 523, and the similarity in average standardized charges between those cases in DRG 521 with a principal diagnosis of code 292.82 and without a secondary diagnosis of drug/alcohol abuse or dependence to the overall average for all cases in the DRG, we do not believe the data suggest that a modification to DRGs 521 through 523 is warranted. Therefore, we did not propose changes to the current structure of DRGs 521 through 523 for FY 2006.

Comment: One commenter expressed concern that CMS did not propose any DRG change to code 292.82, drug-induced dementia. The commenter stated that a patient admitted with dementia due to an adverse effect of a drug would result in code 292.82, followed by the appropriate E code as a secondary diagnosis, grouping to one of the alcohol and drug abuse DRGs (521 through 523). The commenter indicated an adverse effect of a drug should not be confused with alcohol or drug abuse and recommended that CMS examine the potential impact of not reassigning code 292.82 into a new DRG from both a quality of care and a financial perspective.

Response: We appreciate the commenter's recommendation. However, as we indicated above and in the FY 2006 IPPS proposed rule, drug-induced dementia develops as a result of use of a drug. Therefore, it is appropriate to assign the code to DRGs 521, 522, or 523. As we indicated in the FY 2006 proposed rule (70 FR 23330), we did receive suggestions that drug-induced dementia due to the adverse effects of a drug or poisoning be assigned to DRGs 429, 449, 450, or 451. However, we believe these DRGs should only be assigned when the hospital uses the appropriate poisoning or other codes sequenced as the principal diagnosis. In addition, the data analyzed from the FY 2004 MedPAR file did not support a modification to DRGs 521 through 523. Our data show that hospital charges for patients assigned to DRGs 521 through 523 with a principal diagnosis of code 292.82 and no drug abuse secondary diagnosis were similar to other patients in these DRGs. Given that no other secondary diagnosis codes were used, it is not possible to know whether these patients were more clinically similar to patients in DRGs 426, 449, 450, 451, or 521 through 523. Absent any other diagnoses other than code 292.82, we have no evidence that these patients were clinically different than other patients in DRGs 521 through 523.

After consideration of the comments received, as we proposed, in this final rule we are not changing the DRG assignment for drug-induced dementia (code 292.82) for FY 2006. 10. Medicare Code Editor (MCE) Changes

As explained under section II.B.1. of this preamble, the Medicare Code Editor (MCE) is a software program that detects and reports errors in the coding of Medicare claims data. Patient diagnoses, procedure(s), discharge status, and demographic information go into the Medicare claims processing systems and are subjected to a series of automated screens. The MCE screens are designed to identify cases that require further review before classification into a DRG. a. Newborn Age Edit

In the past, we have discussed and received comments concerning revision of the pediatric portions of the Medicare IPPS DRG classification system, that is, MDC 15 (Newborns and Other Neonates With Conditions Originating in the Perinatal Period). Most recently, we addressed these comments in both the FY 2005 proposed rule (69 FR 28210) and the FY 2005 IPPS final rule (69 FR 48938). In those rules, we indicated that we would be responsive to specific requests for updating MDC 15 on a limited, case-by-case basis.

We have recently received a request through the Open Door Forum to revise the MCE ``newborn age edit'' by removing over 100 codes located in Chapter 15 of ICD-9-CM that are identified as ``newborn'' codes. This request was made because these codes usually cause an edit or denial to be triggered when they are used on children greater than 1 year of age. However, the underlying issue with these particular edits is that other payers have adopted the CMS Medicare Code Editor in a wholesale manner, instead of adapting it for use in their own patient populations.

We acknowledge that Medicare DRGs are sometimes used to classify other patient groups. However, CMS' primary focus of updates to the Medicare DRG classification system is on changes relating to the Medicare patient population, not the pediatric or neonatal patient populations.

There are practical considerations regarding the assumption of a larger role for the Medicare DRGs in the pediatric or neonatal areas, given the difference between the Medicare population and that of newborns and children. There are also challenges surrounding the development of DRG classification systems and applications appropriate to children. We do not have the clinical expertise to make decisions about these patients, and must rely on outside clinicians for advice. In addition, because newborns and other children are generally not eligible for Medicare, we must rely on outside data to make decisions. We recognize that there are evolving alternative classification systems for children and encourage payers to use the CMS MCE as a template while making modifications appropriate for pediatric patients.

Therefore, we would encourage those non-Medicare systems needing a more comprehensive pediatric system of edits to update their systems by choosing from other existing systems or programs that are currently in use. Because of our reluctance to assume expertise in the

[[Page 47312]]

pediatric arena, as we proposed we are not making the commenter's suggested changes to the MCE ``newborn age edit'' for FY 2006.

Comment: One commenter requested that CMS reconsider making the necessary revisions to the ``newborn age edit'' and other pediatric data. The commenter suggested that if CMS continues its current stance regarding the internal level of expertise to develop newborn and pediatric edits, then these edit should be removed from the MCE.

Response: We believe the commenter's recommendation to remove the newborn and pediatric edits from the MCE has merits and will consider it for FY 2007. However, we believe it is important that we have an opportunity to analyze this issue further and consider any comments from interested parties before eliminating these edits. b. Newborn Diagnoses Edit

Last year, in our changes to the MCE, we inadvertently added code 796.6 (Abnormal findings on neonatal screening) to both the MCE edit for ``Maternity Diagnoses--age 12 through 55'', and the MCE edit for ``Diagnoses Allowed for Females Only''. In the FY 2006 IPPS proposed rule, we proposed to remove code 796.6 from these two edits and add it to the ``Newborn Diagnoses'' edit.

We did not receive any comments on this proposal. Therefore, in this final rule, we are adopting the proposal as final without modification. c. Diagnoses Allowed for ``Males Only'' Edit

We have received a request to remove two codes from the ``Diagnoses Allowed for Males Only'' edit, related to androgen insensitivity syndrome (AIS). AIS is a new term for testicular feminization. Code 257.8 (Other testicular dysfunction) is used to describe individuals who, despite having XY chromosomes, develop as females with normal female genitalia and mammary glands. Testicles are present in the same general area as the ovaries, but are undescended and are at risk for development of testicular cancer, so are generally surgically removed. These individuals have been raised as females, and would continue to be considered female, despite their XY chromosome makeup. Therefore, as AIS is coded to 257.8, and has posed a problem associated with the gender edit, in the FY 2006 IPPS proposed rule, we proposed to remove this code from the ``Males Only'' edit in the MCE.

A similar clinical scenario can occur with certain disorders that cause a defective biosynthesis of testicular androgen. This disorder is included in code 257.2 (Other testicular hypofunction). Therefore, we also proposed to remove code 257.2 from the ``Male Only'' gender edit in the MCE.

We did not receive any comments on these proposals. Therefore, in this final rule, we are adopting the proposals as final without modification. d. Tobacco Use Disorder Edit

We have become aware of the possible need to add code 305.1 (Tobacco use disorder) to the MCE in order to make admissions for tobacco use disorder a noncovered Medicare service when code 305.1 is reported as the principal diagnosis. On March 22, 2005, CMS published a final decision memorandum and related national coverage determination (NCD) on smoking cessation counseling services on its Web site: (http://www.cms.hhs.gov/coverage/ ). Among other things, this NCD provides

that: ``Inpatient hospital stays with the principal diagnosis of 305.1, Tobacco Use Disorder, are not reasonable and necessary for the effective delivery of tobacco cessation counseling services. Therefore, we will not cover tobacco cessation services if tobacco cessation is the primary reason for the patient's hospital stay.'' Therefore, in order to maintain internal consistency with CMS programs and decisions, we proposed to add code 305.1 to the MCE edit ``Questionable Admission- Principal Diagnosis Only'' in order to make tobacco use disorder a noncovered admission.

We did not receive any comments on this proposal. Therefore, in this final rule, we are adopting the proposal as final without modification. e. Noncovered Procedure Edit

Effective October 1, 2004, CMS adopted the use of code 00.61 (Percutaneous angioplasty or atherectomy of precerebral (extracranial) vessel(s) (PTA)) and code 00.63 (Percutaneous insertion of carotid artery stent(s). Both codes are to be recorded to indicate the insertion of a carotid artery stent or stents. At the time of the creation of the codes, the coverage indication for carotid artery stenting was only for patients in a clinical trial setting, and diagnostic code V70.7 (Examination of participation in a clinical trial) was required for payment of these cases. However, effective October 12, 2004, Medicare covers PTA of the carotid artery concurrent with the placement of an FDA-approved carotid stent for an FDA-approved indication when furnished in accordance with FDA-approved protocols governing post-approval studies. Therefore, as the coverage indication has changed, we proposed to remove codes 00.61, 00.63, and V70.7 from the MCE noncovered procedure edit.

We did not receive any comments on this proposal. Therefore, in this final rule, we are adopting the proposal as final without modification. f. Error in Non-Covered Procedure Edit--code 36.32

It has come to our attention that an entry in the Non-Covered Procedures section of the MCE was made in error. Procedure code 36.32 (Other transmyocardial revascularization) is covered as a late or last resort for patients with severe (Canadian Cardiovascular Society classification Classes III or IV) angina (stable or unstable). The angina symptoms must be caused by areas of the heart not amenable to surgical therapies. Therefore, as code 36.32 is erroneously in the Non- Covered Procedure edit in the MCE, we are removing it from the edits for FY 2006. 11. Surgical Hierarchies

Some inpatient stays entail multiple surgical procedures, each one of which, occurring by itself, could result in assignment of the case to a different DRG within the MDC to which the principal diagnosis is assigned. Therefore, it is necessary to have a decision rule within the GROUPER by which these cases are assigned to a single DRG. The surgical hierarchy, an ordering of surgical classes from most resource-intensive to least resource-intensive, performs that function. Application of this hierarchy ensures that cases involving multiple surgical procedures are assigned to the DRG associated with the most resource- intensive surgical class.

Because the relative resource intensity of surgical classes can shift as a function of DRG reclassification and recalibrations, we reviewed the surgical hierarchy of each MDC, as we have for previous reclassifications and recalibrations, to determine if the ordering of classes coincides with the intensity of resource utilization.

A surgical class can be composed of one or more DRGs. For example, in MDC 11, the surgical class ``kidney transplant'' consists of a single DRG (DRG 302) and the class ``kidney, ureter and major bladder procedures'' consists of three DRGs (DRGs 303, 304, and 305). Consequently, in many cases, the surgical hierarchy has an impact on more than one DRG. The methodology for determining the most resource- intensive surgical class involves

[[Page 47313]]

weighting the average resources for each DRG by frequency to determine the weighted average resources for each surgical class. For example, assume surgical class A includes DRGs 1 and 2 and surgical class B includes DRGs 3, 4, and 5. Assume also that the average charge of DRG 1 is higher than that of DRG 3, but the average charges of DRGs 4 and 5 are higher than the average charge of DRG 2. To determine whether surgical class A should be higher or lower than surgical class B in the surgical hierarchy, we would weight the average charge of each DRG in the class by frequency (that is, by the number of cases in the DRG) to determine average resource consumption for the surgical class. The surgical classes would then be ordered from the class with the highest average resource utilization to that with the lowest, with the exception of ``other O.R. procedures'' as discussed below.

This methodology may occasionally result in assignment of a case involving multiple procedures to the lower-weighted DRG (in the highest, most resource-intensive surgical class) of the available alternatives. However, given that the logic underlying the surgical hierarchy provides that the GROUPER search for the procedure in the most resource-intensive surgical class, in cases involving multiple procedures, this result is sometimes unavoidable.

We note that, notwithstanding the foregoing discussion, there are a few instances when a surgical class with a lower average charge is ordered above a surgical class with a higher average charge. For example, the ``other O.R. procedures'' surgical class is uniformly ordered last in the surgical hierarchy of each MDC in which it occurs, regardless of the fact that the average charge for the DRG or DRGs in that surgical class may be higher than that for other surgical classes in the MDC. The ``other O.R. procedures'' class is a group of procedures that are only infrequently related to the diagnoses in the MDC, but are still occasionally performed on patients in the MDC with these diagnoses. Therefore, assignment to these surgical classes should only occur if no other surgical class more closely related to the diagnoses in the MDC is appropriate.

A second example occurs when the difference between the average charges for two surgical classes is very small. We have found that small differences generally do not warrant reordering of the hierarchy because, as a result of reassigning cases on the basis of the hierarchy change, the average charges are likely to shift such that the higher- ordered surgical class has a lower average charge than the class ordered below it.

Based on the preliminary recalibration of the DRGs, in the FY 2006 IPPS proposed rule (70 FR 23332), we proposed to revise the surgical hierarchy for MDC 5 (Diseases and Disorders of the Circulatory System) and MDC 8 (Diseases and Disorders of the Musculoskeletal System and Connective Tissue) as follows:

In MDC 5, we proposed reordering--

DRG 116 (Other Permanent Cardiac Pacemaker Implant) above DRG 549 (Percutaneous Cardiovascular Procedure With Drug-Eluting Stent With AMI With CC).

DRG 549 above DRG 550 (Percutaneous Cardiovascular Procedure With Drug-Eluting Stent With AMI Without CC).

DRG 550 above DRG 547 (Percutaneous Cardiovascular Procedure With AMI With CC).

DRG 547 above DRG 548 (Percutaneous Cardiovascular Procedure With AMI Without CC).

DRG 548 above DRG 527 (Percutaneous Cardiovascular Procedure With Drug-Eluting Stent Without AMI).

DRG 527 above DRG 517 (Percutaneous Cardiovascular Procedure With Non-Drug Eluting Stent Without AMI).

DRG 517 above DRG 518 (Percutaneous Cardiovascular Procedure Without Coronary Artery Stent or AMI).

DRG 518 above DRGs 478 and 479 (Other Vascular Procedures With and Without CC, respectively).

Comment: Several commenters agreed with the proposed changes in the surgical hierarchy for MDC 5.

Response: We appreciate the commenters' support. However, because in this final rule we are deleting 9 DRGS and creating 12 new DRGs in MDC 5, as discussed under ``MedPAC Recommendations'' in section IX.A of this preamble, we are reordering the following DRGs in MDC 5:

DRG 106 (Coronary Bypass With PTCA) above DRGs 547 and 548 (Coronary Bypass With Cardiac Catheterization With and Without Major CV Diagnosis, respectively);

DRGs 547-548 above DRGs 549 and 550 (Coronary Bypass Without Cardiac Catheterization With and Without Major CV Diagnosis, respectively);

DRG 113 (Amputation For Circulatory System Disorders Except Upper Limb or Toe) above DRG 551 (Permanent Cardiac Pacemaker Implant With Major CV Diagnosis or AICD Lead or Generator);

DRG 551 above DRG 552 (Other Permanent Cardiac Pacemaker Implant Without Major CV Diagnosis);

DRG 552 above DRG 557 (Percutaneous Cardiovascular Procedure With Drug Eluting Stent With Major CV Diagnosis);

DRG 557 above DRG 555 (Percutaneous Cardiovascular Procedure With Major CV Diagnosis);

DRG 555 above DRG 558 (Percutaneous Cardiovascular Procedure With Drug Eluting Stent Without Major CV Diagnosis);

DRG 558 above DRG 556 (Percutaneous Cardiovascular Procedure Without Major CV Diagnosis);

DRG 556 above DRG 518 (Percutaneous Cardiovascular Procedure Without Coronary Artery Stent Or AMI);

DRG 518 above DRG 553 (Other Vascular Procedures With CC With Major CV Diagnosis);

DRG 553 above DRG 554 (Other Vascular Procedures With CC Without Major CV Diagnosis);

DRG 554 above DRG 479 (Other Vascular Procedures Without CC).

In MDC 8, we proposed to reorder--

DRG 496 (Combined Anterior/Posterior Spinal Fusion) above DRG 546 (Spinal Fusions Except Cervical With Curvature of the Spine or Malignancy).

DRG 546 above DRGs 497 and 498 (Spinal Fusions Except Cervical With and Without CC, respectively).

DRG 217 (Wound Debridement and Skin Graft Except Hand, For Musculoskeletal and Connective Tissue Disease) above DRG 545 (Revision of Hip or Knee Replacement).

DRG 545 above DRG 544 (Major Joint Replacement or Reattachment).

DRG 544 above DRGs 519 and 520 (Cervical Spinal Fusion With and Without CC, respectively).

Comment: Several commenters agreed with the proposed changes in the surgical hierarchy for MDC 8.

Response: We appreciate the commenters' support. Based on a test of the proposed revisions using the March 2005 update of the FY 2004 MedPAR file and the revised GROUPER software, we found that the revisions to MDC 8 are still supported by the data.

Accordingly, in this final rule, we are adopting the proposed change in the surgical hierarchy for MDC 8 as final, without modification. 12. Refinement of Complications and Comorbidities (CC) List

a. Background

As indicated earlier in this preamble, under the IPPS DRG classification system, we have developed a standard list of diagnoses that are considered complications or comorbidities (CCs). Historically, we developed this list using physician panels that classified

[[Page 47314]]

each diagnosis code based on whether the diagnosis, when present as a secondary condition, would be considered a substantial complication or comorbidity. A substantial complication or comorbidity was defined as a condition that, because of its presence with a specific principal diagnosis, would cause an increase in the length of stay by at least 1 day in at least 75 percent of the patients. b. Comprehensive Review of the CC List

In previous years, we have made changes to the standard list of CCs, either by adding new CCs or deleting CCs already on the list, but we have never conducted a comprehensive review of the list. There are currently 3,285 diagnosis codes on the CC list. There are 121-paired DRGs that are split on the presence or absence of a CC.

We have reviewed these paired DRGs and found that the majority of cases that are assigned to DRGs that have a CC split fall into the DRG with CC. While this fact is not new, we have found that a much higher proportion of cases are being grouped to the DRG with a CC than had occurred in the past. In our review of the DRGs included in Table 7b of the September 1, 1987 Federal Register rule (52 FR 33125), we found the following percentages of cases assigned a CC in those DRGs that had a CC split (DRG Definitions Manual, GROUPER Version 5.0 (1986 data)):

Cases with CC: 61.9 percent

Cases without CC: 38.1 percent

When we compared the above 1986 DRG data to the 2004 DRG data that were included in the DRGs Definitions Manual, GROUPER Version 22.0, we found the following:

Cases with CC: 79.9 percent

Cases without CC: 20.1 percent

(We used DRGs Definitions Manual, GROUPER Version 5.0, for this analysis because prior versions of the DRGs Definitions Manual used age as a surrogate for a CC and the split was ``CC and/or age greater than 69''.)

The vast majority of patients being treated in inpatient settings have a CC as currently defined, and we believe that it is possible that the CC distinction has lost much of its ability to differentiate the resource needs of patients. The original definition used to develop the CC list (the presence of a CC would be expected to extend the length of stay of at least 75 percent of the patients who had the CC by at least one day) was used beginning in 1981 and has been part of the IPPS since its inception in 1983. There has been no substantive review of the CC list since its original development. In reviewing this issue, our clinical experts found several diseases that appear to be obvious candidates to be on the CC list, but currently are not:

Code

Code description

2004 count

041.7...................... Pseudomonas Infection in

47,350 Conditions Classified Elsewhere and/or of Unspecified Site. 253.6...................... Disorders of

23,613 Neurohypophysis. 414.12..................... Dissection of Coronary

2,377 Artery. 359.4...................... Toxic Myopathy.............

1,875 031.2...................... Disseminated Disease Due to

1,428 Mycobacteria. 451.83..................... Phlebitis and

376 Thrombophlebitis of Deep Veins of Upper Extremities.

Conversely, our medical experts believe the following conditions are examples of common conditions that are on the CC list, but are not likely lead to higher treatment costs when present as a secondary diagnosis:

Code

Code description

2004 count

424.0...................... Mitral Valve Disorder......

401,359 305.00..................... Alcohol Abuse Unspecified

69,099 Use. 578.1...................... Blood in Stool.............

53,453 723.4...................... Brachial Neuritis/

5,829 Radiculitis, Not Otherwise Specified. 684........................ Impetigo...................

1,230 293.84..................... Anxiety Disorder in

1,153 Conditions Classified Elsewhere.

We note that the above conditions are examples only of why we believe the CC list needs a comprehensive review. In addition to this review, we note that these conditions may be treated differently under several DRG systems currently in use. For instance, ICD-9-CM code 414.12 (Dissection of coronary artery) is listed as a ``Major CC'' under the All Patient (AP) DRGs, GROUPER Version 21.0 and an ``Extreme'' CC under the All Patient Refined (APR) DRGs, GROUPER Version 20.0, but is not listed as a CC at all in GROUPER Version 22.0 of the DRGs Definitions Manual used by Medicare. Similarly, ICD-9-CM code 424.0 (Mitral valve disorder) is a CC under GROUPER Version 22.0 of the DRGs Definitions Manual for Medicare's DRG system, a minor CC under the GROUPER Version 20.0 of the APR-DRGs, and not a CC at all under GROUPER Version 21.0 of the AP-DRGs.

Given the long period of time that has elapsed since the original CC list was developed, the incremental nature of changes to it, and changes in the way inpatient care is delivered, as indicated in the FY 2006 IPPS proposed rule, we are planning a comprehensive and systematic review of the CC list for the IPPS rule for FY 2007. As part of this process, we plan to consider revising the standard for determining when a condition is a CC. For instance, we may use an alternative to classifying a condition as a CC based on how it affects the length of stay of a case. Similar to other aspects of the DRG system, we may consider the effect of a specific secondary diagnosis on the charges or costs of a case to evaluate whether to include the condition on the CC list. Using a statistical algorithm, we may classify each diagnosis based on its effect on hospital charges (or costs) relative to other cases when present as a secondary diagnosis to obtain better information on when a particular condition is likely to increase hospital costs. For example, code 293.84 (Anxiety disorder in conditions classified elsewhere), which is currently listed as a CC, might be removed from the CC list if analysis of the data indicates that the data do not support the fact that it represents a significant increase in resource utilization, and a code such as 359.4 (Toxic myopathy), which is currently not listed as a CC, could be added to the CC list if the data support it. In addition to using hospital

[[Page 47315]]

charge data as a basis for a review, we would expect to supplement the process with review by our medical experts. Further, we may also consider doing a comparison of the Medicare DRG CC list with other DRG systems such as the AP-DRGs and the APR-DRGs to determine how the same secondary diagnoses are treated under these systems.

By performing a comprehensive review of the CC list, we expect to revise the DRG classification system to better reflect resource utilization and remove conditions from the CC list that only have a marginal impact on a hospital's costs. We believe that a comprehensive review of the CC list would be consistent with MedPAC's recommendation that we improve the DRG system to better recognize severity. We will provide more detail about how we expect to undertake this analysis in the future, and any significant structural changes to the CC list will only be adopted after a notice and comment rulemaking that fully explains the methodology we plan to use in conducting this review. In the FY 2006 IPPS proposed rule, we encouraged comment regarding possible ways that more meaningful indicators of clinical severity and their implications for resource use can be incorporated into our comprehensive review and possible restructuring of the CC list.

Comment: Several commenters agreed with CMS that changes in resource utilization and in inpatient hospital care, particularly the focus on decreasing length of stay, may be resulting in the CC distinction not being able to differentiate resource utilization and patient severity as well as it has in the past. Several commenters agreed that it may be valuable to conduct a substantial and comprehensive review of the CC list for the future. While some commenters applauded CMS' efforts to keep refining the DRG system, the commenters believed that review of the CC list can only be taken as an interim step and a more refined DRG system can only be accomplished with more specific clinical classification systems capable of providing more complete information about a patient's condition and the services provided to treat those conditions--namely, ICD-10-CM and ICD-10-PCS. Some commenters suggested waiting to adopt the MedPAC recommendations until these new coding classification systems are implemented.

MedPAC stated that a comprehensive review and revision of the CC list might lead to a desirable improvement in the extent to which payment rates reflect patient severity of illness. However, MedPAC does not expect that even a major revision of the list would greatly improve the extent to which the IPPS payment rates recognize the effects of differences in patient severity of illness. MedPAC noted that the CC distinction is based entirely on the presence or absence of any CC, implicitly assuming that all CCs have equal effects on severity of illness and costs. Even if the CC review process were to correctly identify all secondary diagnoses that significantly affect hospitals' costs, MedPAC's research and CMS' earlier work have shown that simply distinguishing between patients with and without CCs fails to capture large, predictable differences in costs among patients. MedPAC stated that further differentiation is necessary to make the most effective use of information about patients' secondary diagnoses and to help minimize opportunities for hospitals to benefit financially from patient selection.

Response: There has not been a comprehensive review of the CC list in over 20 years. Such a review may indicate that a more focused list will better distinguish the effects of CCs on severity of illness than earlier analysis. Until this comprehensive review and analysis are complete, we will not know whether there is merit in adopting a modification of the CC list or whether it will be necessary to adopt a more comprehensive change to the DRG system such as APR-DRGs. We currently plan to continue with our comprehensive review of the CC list. In addition, we expect shortly to engage a contractor highly experienced with DRG development to study the APR-DRGs over the next year. We appreciate the commenters' suggestions about waiting to adopt MedPAC's recommendations until ICD-10-CM and ICD-10-PCS have been implemented. While we do not have a proposal in place at this time to implement ICD-10-CM and ICD-10-PCS, before adopting any major changes to the DRG system, we will consider the implications of potential future changes to our coding systems as part of our analysis of MedPAC's recommendation.

Comment: Commenters gave numerous suggestions for performing the analysis of the CC list. The suggestions include:

Analyze all diagnosis and procedures codes reported on the claim, not just nine diagnosis codes and six procedure codes.

Examine the impact of multiple CCs on hospital resource consumption and length of stay.

Examine further differentiation beyond simply distinguishing between patients with and without CCs to make the most effective use of information about patients' secondary diagnoses and minimize opportunities for hospitals to benefit financially from patient selection.

Study the need for a general/standard list of CCs that addresses patient conditions across all body systems and a list of special severity conditions that are unique to specific population/ diseases.

Consider abandoning length of stay as an indicator for severity because, in today's clinical environment, length of stay is determined more by postacute care referral dynamics than patient need.

Consider differentiating comorbidities from complications. The former are predictable and can be used to easily affect admission selection.

Compare the existing CC list with those used with other DRG systems.

Conduct the comprehensive review and analysis cautiously, systemically, and thoroughly, using external expertise and maintaining transparency and stakeholder involvement throughout the process, and do not rush the analysis simply to meet the deadline for the FY 2007 IPPS rule.

Use open door forums to inform the public of progress.

Consider combining the cases from each DRG pair in one homogenous DRG. Under such a change, hospitals would still receive the same total reimbursement for the same patients but would have more financial incentive to improve the quality and efficiency of care.

Before inclusion as a CC condition, a diagnosis should meet the following four criteria: (1) The patient group represents a higher cost in that DRG than those without the comorbid condition; (2) the condition cannot be prevented, in any possible way, by superior care in the hospital; (3) the condition is not related to the principal diagnosis; and (4) there is at least some indication that the patient would face inadequate options for finding appropriate medical care without a more appropriate payment.

Response: We appreciate these many suggestions. As we indicated above, we will continue to conduct a thorough review of the CC list. We also will be engaging a contractor shortly to assist us with evaluating APR-DRGs and other mechanisms to better recognize severity in our payment systems. c. CC Exclusions List for FY 2006

In the September 1, 1987 final notice (52 FR 33143) concerning changes to the DRG classification system, we modified the GROUPER logic so that certain diagnoses included on the standard list

[[Page 47316]]

of CCs would not be considered valid CCs in combination with a particular principal diagnosis. We created the CC Exclusions List for the following reasons: (1) To preclude coding of CCs for closely related conditions; (2) to preclude duplicative or inconsistent coding from being treated as CCs; and (3) to ensure that cases are appropriately classified between the complicated and uncomplicated DRGs in a pair. As we indicated above, we developed a list of diagnoses, using physician panels, to include those diagnoses that, when present as a secondary condition, would be considered a substantial complication or comorbidity. In previous years, we have made changes to the list of CCs, either by adding new CCs or deleting CCs already on the list. We did not receive any comments specific to the diagnosis codes on the FY 2006 CC list. Therefore, as we proposed in the FY 2006 IPPS proposed rule, we are not deleting any of the diagnosis codes on the CC list for FY 2006.

In the May 19, 1987 proposed notice (52 FR 18877) and the September 1, 1987 final notice (52 FR 33154), we explained that the excluded secondary diagnoses were established using the following five principles:

Chronic and acute manifestations of the same condition should not be considered CCs for one another.

Specific and nonspecific (that is, not otherwise specified (NOS)) diagnosis codes for the same condition should not be considered CCs for one another.

Codes for the same condition that cannot coexist, such as partial/total, unilateral/bilateral, obstructed/unobstructed, and benign/malignant, should not be considered CCs for one another.

Codes for the same condition in anatomically proximal sites should not be considered CCs for one another.

Closely related conditions should not be considered CCs for one another.

The creation of the CC Exclusions List was a major project involving hundreds of codes. We have continued to review the remaining CCs to identify additional exclusions and to remove diagnoses from the master list that have been shown not to meet the definition of a CC.\1\

\1\ See the FY 1989 final rule (53 FR 38485) September 30, 1988, for the revision made for the discharges occurring in FY 1989; the FY 1990 final rule (54 FR 36552) September 1, 1989, for the FY 1990 revision; the FY 1991 final rule (55 FR 36126) September 4, 1990, for the FY 1991 revision; the FY 1992 final rule (56 FR 43209) August 30, 1991, for the FY 1992 revision; the FY 1993 final rule (57 FR 39753) September 1, 1992, for the FY 1993 revision; the FY 1994 final rule (58 FR 46278) September 1, 1993, for the FY 1994 revisions; the FY 1995 final rule (59 FR 45334), September 1, 1994, for the FY 1995 revisions; the FY 1996 final rule (60 FR 45782) September 1, 1995, for the FY 1996 revisions; the FY 1997 final rule (61 FR 46171), August 30, 1996, for the FY 1997 revisions; the FY 1998 final rule (62 FR 45966), August 29, 1997, for the FY 1998 revisions; the FY 1999 final rule (63 FR 40954), July 31, 1998, for the FY 1999 revisions; the FY 2001 final rule (65 FR 47064), August 1, 2000, for the FY 2001 revisions; the FY 2002 final rule (66 FR 39851) August 1, 2001, for the FY 2002 revisions; the FY 2003 final rule (67 FR 49998), August 1, 2002, for the FY 2003 revisions; the FY 2004 final rule (68 FR 45364) August 1, 2003, for the FY 2004 revisions; and the FY 2005 final rule (69 FR 49848) August 11, 2004, for the FY 2005 revisions. In the FY 2000 final rule (64 FR 41490) July 30, 1999, we did not modify the CC Exclusions List because we did not make any changes to the ICD-9-CM codes for FY 2000.

As proposed, we are making a limited revision of the CC Exclusions List to take into account the changes that will be made in the ICD-9-CM diagnosis coding system effective October 1, 2005. (See section II.B.14. of this preamble for a discussion of ICD-9-CM changes.) We are making these changes in accordance with the principles established when we created the CC Exclusions List in 1987.

We receive one comment that agreed with the revised CC Exclusion List based on the information provided.

Tables 6G and 6H in the Addendum to this final rule contain the revisions to the CC Exclusions List that will be effective for discharges occurring on or after October 1, 2005. Each table shows the principal diagnoses with changes to the excluded CCs. Each of these principal diagnoses is shown with an asterisk, and the additions or deletions to the CC Exclusions List are provided in an indented column immediately following the affected principal diagnosis.

CCs that are added to the list are in Table 6G--Additions to the CC Exclusions List. Beginning with discharges on or after October 1, 2005, the indented diagnoses will not be recognized by the GROUPER as valid CCs for the asterisked principal diagnosis.

CCs that are deleted from the list are in Table 6H--Deletions from the CC Exclusions List. Beginning with discharges on or after October 1, 2005, the indented diagnoses will be recognized by the GROUPER as valid CCs for the asterisked principal diagnosis.

Copies of the original CC Exclusions List applicable to FY 1988 can be obtained from the National Technical Information Service (NTIS) of the Department of Commerce. It is available in hard copy for $152.50 plus shipping and handling. A request for the FY 1988 CC Exclusions List (which should include the identification accession number (PB) 88- 133970) should be made to the following address: National Technical Information Service, United States Department of Commerce, 5285 Port Royal Road, Springfield, VA 22161; or by calling (800) 553-6847.

Users should be aware of the fact that all revisions to the CC Exclusions List (FYs 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2001, 2002, 2003, 2004, and 2005) and those in Tables 6G and 6H of this final rule for FY 2006 must be incorporated into the list purchased from NTIS in order to obtain the CC Exclusions List applicable for discharges occurring on or after October 1, 2005. (Note: There was no CC Exclusions List in FY 2000 because we did not make changes to the ICD-9-CM codes for FY 2000.)

Alternatively, the complete documentation of the GROUPER logic, including the current CC Exclusions List, is available from 3M/Health Information Systems (HIS), which, under contract with CMS, is responsible for updating and maintaining the GROUPER program. The current DRG Definitions Manual, Version 22.0, is available for $225.00, which includes $15.00 for shipping and handling. Version 23.0 of this manual, which will include the final FY 2006 DRG changes, will be available in hard copy for $250.00. Version 23.0 of the manual is also available on a CD for $200.00; a combination hard copy and CD is available for $400.00. These manuals may be obtained by writing 3M/HIS at the following address: 100 Barnes Road, Wallingford, CT 06492; or by calling (203) 949-0303. Please specify the revision or revisions requested. 13. Review of Procedure Codes in DRGs 468, 476, and 477

Each year, we review cases assigned to DRG 468 (Extensive O.R. Procedure Unrelated to Principal Diagnosis), DRG 476 (Prostatic O.R. Procedure Unrelated to Principal Diagnosis), and DRG 477 (Nonextensive O.R. Procedure Unrelated to Principal Diagnosis) to determine whether it would be appropriate to change the procedures assigned among these DRGs.

DRGs 468, 476, and 477 are reserved for those cases in which none of the O.R. procedures performed are related to the principal diagnosis. These DRGs are intended to capture atypical cases, that is, those cases not occurring with sufficient frequency to represent a distinct, recognizable clinical group. DRG 476 is assigned to those discharges in which one or more of the following prostatic procedures are performed and are unrelated to the principal diagnosis:

[[Page 47317]]

60.0, Incision of prostate

60.12, Open biopsy of prostate

60.15, Biopsy of periprostatic tissue

60.18, Other diagnostic procedures on prostate and periprostatic tissue

60.21, Transurethral prostatectomy

60.29, Other transurethral prostatectomy

60.61, Local excision of lesion of prostate

60.69, Prostatectomy, not elsewhere classified

60.81, Incision of periprostatic tissue

60.82, Excision of periprostatic tissue

60.93, Repair of prostate

60.94, Control of (postoperative) hemorrhage of prostate

60.95, Transurethral balloon dilation of the prostatic urethra

60.96, Transurethral destruction of prostate tissue by microwave thermotherapy

60.97, Other transurethral destruction of prostate tissue by other thermotherapy

60.99, Other operations on prostate

All remaining O.R. procedures are assigned to DRGs 468 and 477, with DRG 477 assigned to those discharges in which the only procedures performed are nonextensive procedures that are unrelated to the principal diagnosis.\2\

\2\ The original list of the ICD-9-CM procedure codes for the procedures we consider nonextensive procedures, if performed with an unrelated principal diagnosis, was published in Table 6C in section IV. of the Addendum to the FY 1989 final rule (53 FR 38591). As part of the FY 1991 final rule (55 FR 36135), the FY 1992 final rule (56 FR 43212), the FY 1993 final rule (57 FR 23625), the FY 1994 final rule (58 FR 46279), the FY 1995 final rule (59 FR 45336), the FY 1996 final rule (60 FR 45783), the FY 1997 final rule (61 FR 46173), and the FY 1998 final rule (62 FR 45981), we moved several other procedures from DRG 468 to DRG 477, and some procedures from DRG 477 to DRG 468. No procedures were moved in FY 1999, as noted in the final rule (63 FR 40962); in FY 2000 (64 FR 41496); in FY 2001 (65 FR 47064); or in FY 2002 (66 FR 39852). In the FY 2003 final rule (67 FR 49999) we did not move any procedures from DRG 477. However, we did move procedure codes from DRG 468 and placed them in more clinically coherent DRGs. In the FY 2004 final rule (68 FR 45365), we moved several procedures from DRG 468 to DRGs 476 and 477 because the procedures are nonextensive. In the FY 2005 final rule (69 FR 48950), we moved one procedure from DRG 468 to 477. In addition, we added several existing procedures to DRGs 476 and 477.

a. Moving Procedure Codes from DRG 468 or DRG 477 to MDCs

We annually conduct a review of procedures producing assignment to DRG 468 or DRG 477 on the basis of volume, by procedure, to see if it would be appropriate to move procedure codes out of these DRGs into one of the surgical DRGs for the MDC into which the principal diagnosis falls. The data are arrayed two ways for comparison purposes. We look at a frequency count of each major operative procedure code. We also compare procedures across MDCs by volume of procedure codes within each MDC.

We identify those procedures occurring in conjunction with certain principal diagnoses with sufficient frequency to justify adding them to one of the surgical DRGs for the MDC in which the diagnosis falls. Based on this year's review, we did not identify any procedures in DRGs 468 or 477 that should be removed to one of the surgical DRGs. We did not receive any comments on this provision. Therefore, in this final rule, we are not making any changes for FY 2006. b. Reassignment of Procedures Among DRGs 468, 476, and 477

We also annually review the list of ICD-9-CM procedures that, when in combination with their principal diagnosis code, result in assignment to DRGs 468, 476, and 477, to ascertain if any of those procedures should be reassigned from one of these three DRGs to another of the three DRGs based on average charges and the length of stay. We look at the data for trends such as shifts in treatment practice or reporting practice that would make the resulting DRG assignment illogical. If we find these shifts, we would propose to move cases to keep the DRGs clinically similar or to provide payment for the cases in a similar manner. Generally, we move only those procedures for which we have an adequate number of discharges to analyze the data.

It has come to our attention that procedure code 26.12 (Open biopsy of salivary gland or duct) is assigned to DRG 468 (Extensive O.R. Procedure Unrelated to Principal Diagnosis). We believe this to be an error, as code 26.31 (Partial sialoadenectomy), which is a more extensive procedure than code 26.12, is assigned to DRG 477. Therefore, we proposed to correct this error by moving code 26.12 out of DRG 468 and reassigning it to DRG 477. We received one comment in support of our proposal to move code 26.12 out of DRG 468 and reassign it to DRG 477. Therefore, we are adopting as final our proposal to move procedure code 26.12 out of DRG 468 and reassigning it to DRG 477. We received no comments opposing our plan of not moving any procedure codes from DRG 476 to DRGs 468 or 477 or from DRG 477 to DRG 468. Therefore, as we proposed, we are not moving any procedure codes from DRG 476 to DRGs 468 or 477, or from DRG 477 to DRGs 468 or 476. c. Adding Diagnosis or Procedure Codes to MDCs

Based on our review this year, as we proposed, we are not adding any diagnosis codes to MDCs. We did not receive any comments on our proposal and are therefore not adding any diagnosis codes to any MDCs. 14. Changes to the ICD-9-CM Coding System

As described in section II.B.1. of this preamble, the ICD-9-CM is a coding system used for the reporting of diagnoses and procedures performed on a patient. In September 1985, the ICD-9-CM Coordination and Maintenance Committee was formed. This is a Federal interdepartmental committee, co-chaired by the National Center for Health Statistics (NCHS), the Centers for Disease Control and Prevention, and CMS, charged with maintaining and updating the ICD-9-CM system. The Committee is jointly responsible for approving coding changes, and developing errata, addenda, and other modifications to the ICD-9-CM to reflect newly developed procedures and technologies and newly identified diseases. The Committee is also responsible for promoting the use of Federal and non-Federal educational programs and other communication techniques with a view toward standardizing coding applications and upgrading the quality of the classification system.

The Official Version of the ICD-9-CM contains the list of valid diagnosis and procedure codes. (The Official Version of the ICD-9-CM is available from the Government Printing Office on CD-ROM for $25.00 by calling (202) 512-1800.) The Official Version of the ICD-9-CM is no longer available in printed manual form from the Federal Government; it is only available on CD-ROM. Users who need a paper version are referred to one of the many products available from publishing houses.

The NCHS has lead responsibility for the ICD-9-CM diagnosis codes included in the Tabular List and Alphabetic Index for Diseases, while CMS has lead responsibility for the ICD-9-CM procedure codes included in the Tabular List and Alphabetic Index for Procedures.

The Committee encourages participation in the above process by health-related organizations. In this regard, the Committee holds public meetings for discussion of educational issues and proposed coding changes. These meetings provide an opportunity for representatives of recognized organizations in the coding field, such

[[Page 47318]]

as the American Health Information Management Association (AHIMA), the American Hospital Association (AHA), and various physician specialty groups, as well as individual physicians, health information management professionals, and other members of the public, to contribute ideas on coding matters. After considering the opinions expressed at the public meetings and in writing, the Committee formulates recommendations, which then must be approved by the agencies.

The Committee presented proposals for coding changes for implementation in FY 2006 at a public meeting held on October 7-8, 2004, and finalized the coding changes after consideration of comments received at the meetings and in writing by January 12, 2005. Those coding changes are announced in Tables 6A through 6F of the Addendum to this final rule. The Committee held its 2005 meeting on March 31-April l, 2005. New codes for which there was a consensus of public support and for which complete tabular and indexing changes were made by May 2005 are included in the October 1, 2005 update to ICD-9-CM. Code revisions that were discussed at the March 31-April 1, 2005 Committee meeting were not finalized in time to include them in the FY 2006 IPPS proposed rule. These additional codes are included in Tables 6A through 6F of this final rule and are marked with an asterisk (*).

Copies of the minutes of the procedure codes discussions at the Committee's October 7-8, 2004 meeting can be obtained from the CMS Web site: http://www.cms.hhs.gov/paymentsystems/icd9/. The minutes of the

diagnoses codes discussions at the October 7-8, 2004 meeting are found at: http://www.cdc.gov/nchs/icd9.htm. Paper copies of these minutes are

no longer available and the mailing list has been discontinued. These Web sites also provide detailed information about the Committee, including information on requesting a new code, attending a Committee meeting, and timeline requirements and meeting dates.

We encourage commenters to address suggestions on coding issues involving diagnosis codes to: Donna Pickett, Co-Chairperson, ICD-9-CM Coordination and Maintenance Committee, NCHS, Room 2402, 3311 Toledo Road, Hyattsville, MD 20782. Comments may be sent by E-mail to: dfp4@cdc.gov.

Questions and comments concerning the procedure codes should be addressed to: Patricia E. Brooks, Co-Chairperson, ICD-9-CM Coordination and Maintenance Committee, CMS, Center for Medicare Management, Hospital and Ambulatory Policy Group, Division of Acute Care, C4-08-06, 7500 Security Boulevard, Baltimore, MD 21244-1850. Comments may be sent by E-mail to: Patricia.Brooks1@cms.hhs.gov.

The ICD-9-CM code changes that have been approved will become effective October 1, 2005. The new ICD-9-CM codes are listed, along with their DRG classifications, in Tables 6A and 6B (New Diagnosis Codes and New Procedure Codes, respectively) in the Addendum to this final rule. As we stated above, the code numbers and their titles were presented for public comment at the ICD-9-CM Coordination and Maintenance Committee meetings. Both oral and written comments were considered before the codes were approved. In the FY 2006 IPPS proposed rule, we only solicited comments on the proposed classification of these new codes.

For codes that have been replaced by new or expanded codes, the corresponding new or expanded diagnosis codes are included in Table 6A. New procedure codes are shown in Table 6B. Diagnosis codes that have been replaced by expanded codes or other codes or have been deleted are in Table 6C (Invalid Diagnosis Codes). These invalid diagnosis codes will not be recognized by the GROUPER beginning with discharges occurring on or after October 1, 2005. Table 6D contains invalid procedure codes. These invalid procedure codes will not be recognized by the GROUPER beginning with discharges occurring on or after October 1, 2005. Revisions to diagnosis code titles are in Table 6E (Revised Diagnosis Code Titles), which also includes the DRG assignments for these revised codes. Table 6F includes revised procedure code titles for FY 2006.

In the September 7, 2001 final rule implementing the IPPS new technology add-on payments (66 FR 46906), we indicated we would attempt to include proposals for procedure codes that would describe new technology discussed and approved at the April meeting as part of the code revisions effective the following October. As stated previously, ICD-9-CM codes discussed at the March 31-April 1, 2005 Committee meeting that received consensus and that were finalized are included in Tables 6A through 6F of this final rule.

Section 503(a) of Pub. L. 108-173 included a requirement for updating ICD-9-CM codes twice a year instead of a single update on October 1 of each year. This requirement was included as part of the amendments to the Act relating to recognition of new technology under the IPPS. Section 503(a) amended section 1886(d)(5)(K) of the Act by adding a clause (vii) which states that the ``Secretary shall provide for the addition of new diagnosis and procedure codes in April 1 of each year, but the addition of such codes shall not require the Secretary to adjust the payment (or diagnosis-related group classification) * * * until the fiscal year that begins after such date.'' This requirement improves the recognition of new technologies under the IPPS system by providing information on these new technologies at an earlier date. Data will be available 6 months earlier than would be possible with updates occurring only once a year on October 1.

While section 503(a) states that the addition of new diagnosis and procedure codes on April 1 of each year shall not require the Secretary to adjust the payment, or DRG classification under section 1886(d) of the Act until the fiscal year that begins after such date, we have to update the DRG software and other systems in order to recognize and accept the new codes. We also publicize the code changes and the need for a mid-year systems update by providers to capture the new codes. Hospitals also have to obtain the new code books and encoder updates, and make other system changes in order to capture and report the new codes.

The ICD-9-CM Coordination and Maintenance Committee holds its meetings in the Spring and Fall in order to update the codes and the applicable payment and reporting systems by October 1 of each year. Items are placed on the agenda for the ICD-9-CM Coordination and Maintenance Committee meeting if the request is received at least 2 months prior to the meeting. This requirement allows time for staff to review and research the coding issues and prepare material for discussion at the meeting. It also allows time for the topic to be publicized in meeting announcements in the Federal Register as well as on the CMS Web site. The public decides whether or not to attend the meeting based on the topics listed on the agenda. Final decisions on code title revisions are currently made by March 1 so that these titles can be included in the IPPS proposed rule. A complete addendum describing details of all changes to ICD-9-CM, both tabular and index, are publicized on CMS and NCHS web pages in May of each year. Publishers of coding books and software use this information to modify their products that are used by health care providers. This 5-month time period has proved to be necessary for hospitals and other providers to update their systems.

[[Page 47319]]

A discussion of this timeline and the need for changes are included in March 31-April 1, 2005 ICD-9-CM Coordination and Maintenance Committee minutes. The public agreed that there was a need to hold the fall meetings earlier, in September or October, in order to meet the new implementation dates. The public provided comment that additional time would be needed to update hospital systems and obtain new code books and coding software. There was considerable concern expressed about the impact this new April update would have on providers.

In the FY 2005 IPPS final rule, we implemented section 503(a) by developing a mechanism for approving, in time for the April update, diagnoses and procedure code revisions needed to describe new technologies and medical services for purposes of the new technology add-on payment process. We also established the following process for making these determinations. Topics considered during the Fall ICD-9-CM Coordination and Maintenance Committee meeting are considered for an April 1 update if a strong and convincing case is made by the requester at the Committee's public meeting. The request must identify the reason why a new code is needed in April for purposes of the new technology process. The participants at the meeting and those reviewing the Committee meeting summary report are provided the opportunity to comment on this expedited request. All other topics are considered for the October 1 update. Participants at the Committee meeting are encouraged to comment on all such requests. There were no requests for an expedited April l, 2005 implementation of an ICD-9-CM code at the October 7-8, 2004 Committee meeting. Therefore, there were no new ICD- 9-CM codes implemented on April 1, 2005.

We believe that this process captures the intent of section 503(a). This requirement was included in the provision revising the standards and process for recognizing new technology under the IPPS. In addition, the need for approval of new codes outside the existing cycle (October 1) arises most frequently and most acutely where the new codes will capture new technologies that are (or will be) under consideration for new technology add-on payments. Thus, we believe this provision was intended to expedite data collection through the assignment of new ICD- 9-CM codes for new technologies seeking higher payments.

Current addendum and code title information is published on the CMS Web page at: http://www.cms.hhs.gov/paymentsystems/icd9. Summary tables

showing new, revised, and deleted code titles are also posted on the following CMS Web page: http://www.cms.hhs.gov/medlearn/icd9code.asp.

Information on ICD-9-CM diagnosis codes, along with the Official ICD-9- CM Coding Guidelines, can be found on the Web page at: http://www.cdc.gov/nchs/icd9.htm. Information on new, revised, and deleted

ICD-9-CM codes is also provided to the AHA for publication in the Coding Clinic for ICD-9-CM. AHA also distributes information to publishers and software vendors.

CMS also sends copies of all ICD-9-CM coding changes to its contractors for use in updating their systems and providing education to providers.

These same means of disseminating information on new, revised, and deleted ICD-9-CM codes will be used to notify providers, publishers, software vendors, contractors, and others of any changes to the ICD-9- CM codes that are implemented in April. Currently, code titles are also published in the IPPS proposed and final rules. The code titles are adopted as part of the ICD-9-CM Coordination and Maintenance Committee process. The code titles are not subject to comment in the proposed or final rules. We will continue to publish the October code updates in this manner within the IPPS proposed and final rules. For codes that are implemented in April, we will assign the new procedure code to the same DRG in which its predecessor code was assigned so there will be no DRG impact as far as DRG assignment. This mapping was specified by section 503(a) of Pub. L. 108-173. Any midyear coding updates will be available through the Web sites indicated above and through the Coding Clinic for ICD-9-CM. Publishers and software vendors currently obtain code changes through these sources in order to update their code books and software systems. We will strive to have the April 1 updates available through these Web sites 5 months prior to implementation (that is, early November of the previous year), as is the case for the October 1 updates. Codebook publishers are evaluating how they will provide any code updates to their subscribers. Some publishers may decide to publish mid-year book updates. Others may decide to sell an addendum that lists the changes to the October 1 code book. Coding personnel should contact publishers to determine how they will update their books. CMS and its contractors will also consider developing provider education articles concerning this change to the effective date of certain ICD-9-CM codes.

Comment: Five commenters recommended that CMS modify its DRG GROUPER and instruct fiscal intermediaries to expand the number of diagnoses processed from 9 to 25 and the number of procedures processed from 6 to 25. The commenters were concerned that CMS was not evaluating all reported diagnoses and procedures that could possibly affect a patient's severity of illness or the resources used, or both. The commenters pointed out that the current DRG GROUPER only considers 9 diagnoses and up to 6 procedures; that hospitals submit claims to CMS in electronic format, and that the HIPAA compliant electronic transaction standard, HIPAA 837i, allows up to 25 diagnoses and 25 procedures. The commenters stated that fiscal intermediaries are currently ignoring or omitting the additional codes (beyond 9 diagnoses and 6 procedures) submitted by hospital providers, since these additional diagnoses and procedures are not needed by the GROUPER to assign a DRG. Several commenters stated that, while it is important for inpatient acute hospitals, it is even more crucial for LTCHs whose patients are medically complex and have multiple illnesses beyond the nine diagnoses allowed by CMS. Several commenters further stated that a list of CCs qualifying for comorbidity adjustments for inpatient psychiatric facility services was only recently introduced under the new IRP PPS. Thus, the commenter added, these hospitals have not historically used the software available to sort and rearrange secondary diagnosis cods so that all CCs possibly affecting the DRG grouping are prioritized. One commenter stated that the continued use of more limited diagnosis and procedure codes acts as a disincentive for the reporting of additional codes, and will result in less precise assignment of DRGs.

Response: The commenters are correct that the current Medicare GROUPER does not process codes submitted electronically on the 837i electronic format beyond the first 9 diagnoses and the first 6 procedures. This limitation is not being imposed by the GROUPER. CMS made the decision to process only the first 9 diagnosis codes and first 6 procedure codes. While HIPAA requires CMS to accept up to 25 ICD-9-CM diagnosis and procedure codes on the HIPAA 837i electronic format, it does not require that CMS process that many diagnosis and procedure codes.

As suggested by the commenters, there is value in retaining additional data on patient conditions that would

[[Page 47320]]

result from expanding Medicare's data system so it can accommodate additional diagnosis and procedure codes. We will consider this issue further as we contemplate further refinements to our DRG system to better recognize patient severity. However, while it would be a simple matter to modify our GROUPER software to accept and evaluate 25 diagnosis and 25 procedure codes, extensive lead time to allow for modifications to our internal and contractors' electronic systems would be necessary before we could process and store this additional information. We are unable to move forward with this recommendation without carefully evaluating implementation issues. Nevertheless, we plan to proceed with this evaluation as we consider further changes to our DRG systems.

Comment: Many commenters recommended that CMS act immediately to adopt coordinated implementation of ICD-10-CM and ICD-10-PCS in the United States. Some of these commenters noted that Pub. L. 108-173 (MMA) included report language urging the Secretary to move forward with the implementation of ICD-10 as quickly as possible. The commenters noted that the National Committee on Vital and Health Statistics (NCVHS) raised concerns about the viability of ICD-9-CM in 2003 and stated it was ``increasingly unable to address the needs for accurate data for health care billing, quality assurance, public health reporting, and health services research.'' The commenter further noted that the NCVHS recommended in 2003 that DHHS act expeditiously to initiate the regulatory process for adoption of ICD-10-CM and ICD-10- PCS. The commenter stated that, as of 2005, ``we are still awaiting a process from HHS to begin this important transition.'' While some of the commenters acknowledged the complexities involved with the transition from ICD-9-CM to ICD-10, the commenters still recommended that we act quickly to begin adoption of ICD-10. Other commenters also indicated that the 4-digit structure of ICD-9-CM is limiting the ability of the procedure coding system to identify new procedures and new technologies and it is becoming increasingly outdated. According to these commenters, it is becoming more difficult each year to make changes to the ICD-9-CM coding system because of the availability of new codes. One commenter noted that several participants at the March 31-April 1, 2005 ICD-9-CM Coordination and Maintenance Committee ``appeared to be advocating a higher threshold for the award of new codes based on the ever decreasing number of available codes under ICD- 9-CM.'' Many of the commenters indicated that the coding system's limitations are making it difficult to compare outcomes and efficacy between older and newer technologies, identify costs associated with the new technology, or revise reimbursement policies to appropriately reflect the cost of patient care when new technology is used. One commenter indicated that failure to recognize the looming problems with the ICD-9-CM coding system will impede efforts to meet the President's goal of adopting electronic health records by 2013.

Many of the commenters referred to ICD-10-PCS as the next generation of coding systems. They stated that ICD-10-PCS would modernize and expand CMS' capacity to keep pace with changes in medical practice and technology. In addition, these commenters stated that the structure of ICD-10-PCS would incorporate all new procedures as unique codes that would explicitly identify the technology used to perform the procedure.

Response: We agree that it is becoming increasingly difficult to update ICD-9-CM. However, we are continuing to make revisions to ICD-9- CM and create codes that recognize new medical technology. We continue to update ICD-10-PCS on an annual basis to keep it up to date with changing technology. We agree that it is important to have an accurate and precise coding system for this purpose. However, as noted by many of the commenters, the transition from one coding system to another raises many complex operational issues. The Department will continue to study this matter as we consider whether to adopt ICD-10. 15. Other Issues a. Acute Intermittent Porphyria

Acute intermittent porphyria is a rare metabolic disorder. The condition is described by code 277.1 (Disorders of porphyrin metabolism). Code 277.1 is assigned to DRG 299 (Inborn Errors of Metabolism) under MDC 10 (Endocrine, Nutritional, and Metabolic Diseases and Disorders).

In the FY 2005 final rule (69 FR 48981), we discussed the DRG assignment of acute intermittent porphyria. This discussion was a result of correspondence that we received during the comment period for the FY 2005 proposed rule in which the commenter suggested that Medicare hospitalization payments do not accurately reflect the cost of treatment. At that time, we indicated that we would take this comment into consideration when we analyzed the MedPAR data for this proposed rule for FY 2006.

Our review of the most recent MedPAR data shows a total of 1,370 cases overall in DRG 299, of which 471 had a principal diagnosis coded as 277.1. The average length of stay for all cases in DRG 299 was 5.17 days, while the average length of stay for porphyria cases with code 277.1 was 6.0 days. The average charges for all cases in DRG 299 were $15,891, while the average changes for porphyria cases with code 277.1 were $21,920. Based on our analysis of these data, we did not believe that there is a sufficient difference between the average charges and average length of stay for these cases to justify proposing a change to the DRG assignment for treating this condition.

Comment: One commenter agreed with our proposal not to modify the DRG assignment for acute intermittent porphyria, code 277.1, to DRG 229 due to the minor variance in average charges and length of stay between porphyria cases and other cases in this DRG.

Response: We appreciate the commenter's support of our proposal. Review of the MedPAR data did not demonstrate a significant disparity in the average charges compared to average length of stay.

For FY 2006, as we proposed, we are not modifying the DRG assignment for code 277.1 (Acute intermittent porphyria) to DRG 229. b. Prosthetic Cardiac Support Device (Code 37.41)

Code 37.41 (Implantation of prosthetic cardiac support device around the heart) was addressed in the FY 2006 IPPS proposed rule only as a notification in Table 6B that the new code was being created to describe a prosthetic cardiac support device (70 FR 23594). Code 37.41 was deemed to be an O.R. procedure and was assigned to MDC 5 (Diseases and Disorders of the Circulatory System), DRGs 110 and 111 (Major Cardiovascular Procedures With and Without CC, respectively). This device is being marketed as the CorCapTMCardiac Support Device and is intended to prevent and reverse heart failure by improving the heart's structure and function.

This topic was discussed at the ICD-9-CM Coordination and Maintenance Meeting on October 7, 2004. At that time, there was no specific ICD-9-CM code that more precisely identified this procedure, so coders were advised to use code 37.99 (Other operations on

[[Page 47321]]

heart and pericardium) to describe the operation. Code 37.99 is currently assigned to DRGs 110 and 111.

As is our established pattern, we assign a new code to its predecessor code's DRG until we obtain a pattern of use of the code in the MedPAR data file. After we have evidence-based justification for reassignment of codes within DRGs, we are better able to make decisions about the most appropriate placement of those new codes.

We received 11 comments on this topic as part of the comments on the FY 2006 IPPS proposed rule.

Comment: Most of the commenters responding were cardiovascular surgeons who were principal investigators participating in the United States' CorCapTMclinical trials. All of the commenters requested that we reconsider the assignment of the prosthetic cardiac support device from DRGs 110 and 111 to DRG 108, where the resources

[in DRG 108] more closely approximate those associated with implantation of the device. The commenters stated that procedures in DRG 108 are more clinically similar to the implantation of the prosthetic cardiac support device, being exclusively performed on the internal or external structures of the heart and generally requiring access through a sternotomy.

One commenter likened this procedure to the maze procedure, described by code 37.33 (Excision or destruction of other lesion or tissue of heart, open approach). Another commenter compared it to transmyocardial revascularization, described by code 36.31 (Open chest transmyocardial revascularization). Both of these procedure codes are assigned to DRG 108. Commenters also stated that classification of this procedure to DRGs 110 and 111 would establish a financial disincentive for hospitals to adopt this potentially life-saving and cost-reducing treatment for Medicare beneficiaries suffering from a problem that may otherwise require implantation of a ventricular assist device or heart transplant.

Response: As noted above, we have classified procedure 37.41 to the same DRG as its predecessor code, in accordance with our established policy. Until we have Medicare billing data that will allow us to assess whether the new procedure code has been correctly assigned, our default position is to assign a new procedure code to the same DRG as its predecessor code. Of major concern to CMS is the late June 2005 decision by an FDA advisory panel urging FDA to reject approval of the CorCapTMdevice on the basis that the panel had not seen sufficient evidence of benefit for patients with heart failure. The FDA's concerns included the efficacy of the device in achieving a longer lifespan for patients, and the possibility that the device's benefits did not outweigh the risks of surgery. In addition, the FDA advisory panel had other concerns, including whether the application of this device around the ventricles of the heart might make future heart surgeries more difficult.

Code 37.41 is too new to be included in the MedPAR data. Therefore, we will continue to monitor this prosthetic cardiac support device in future IPPS updates. As noted above, should FDA approve this device and should there be an evidence-based justification for reassignment of codes within these DRGs, we will be open to making changes to the DRG structure. c. Coronary Intravascular Ultrasound (IVUS) (Procedure Code 00.24)

Procedure code 00.24 (Coronary intravascular ultrasound) was addressed in the FY 2005 IPPS proposed rule only as a notification in Table 6B that for FY 2005 a new code had been created to describe this imaging technique (69 FR 49624). Code 00.24 describes ultrasonic imaging within the coronary vessels. It was not assigned ``O.R.'' status within the GROUPER program; that is, the presence or absence of this code does not affect a claim's DRG assignment or payment.

We received one comment on this procedure code as part of the public comments on the FY 2006 IPPS proposed rule.

Comment: One commenter noted that IVUS is an added cost to hospitals. The commenter stated that it has conducted an analysis of coronary IVUS resource use in calendar year 2004 hospital data to determine possible impact. The commenter reported its findings that, in DRGs 516, 517, 526, and 527, cases utilizing IVUS had higher total charges and higher total costs. The commenter requested that CMS perform an analysis of FY 2005 coronary IVUS cases and consider reassigning ICD-9-CM procedure code 00.24 to DRGs where the average resource use most closely approximates the resource use of cases in which an IVUS technique has been employed.

Response: We will perform the requested data analysis using FY 2005 MedPAR data for the FY 2007 annual IPPS update. d. Islet Cell Transplantation

Islet cell transplantation was not a topic addressed in the FY 2006 IPPS proposed rule. The issue of payment for pancreatic islet cell transplantation in clinical trials was addressed in detail in the FY 2005 IPPS final rule (69 FR 48950). At that time, we discussed section 733(b) of Pub. L. 108-173, which provides that Medicare payments, beginning no earlier than October 1, 2004, for the routine costs as well as the costs of the transplantation and appropriate related items and services will be allowed for Medicare beneficiaries who are participating in clinical trials as if such transplantations were covered under Medicare Part A or Part B. In addition, the DRG payment will be supplemented by an add-on payment that includes pre-transplant tests and services, pancreas procurement, and islet isolation services. Cases were assigned to DRG 315 (Other Kidney and Urinary Tract Procedures).

We received one comment on this topic as part of the public comments on the FY 2006 IPPS proposed rule.

Comment: One commenter was concerned that the proposed relative weight for DRG 315 published in the FY 2006 IPPS proposed rule represented a decrease of almost 33 percent. The commenter also indicated that it continues to believe that this procedure is inappropriately classified, and suggested that these cases be reassigned into pre-MDC DRG 513 (Pancreas Transplant). The commenter believed the suggested DRG change is justified because islet cell and pancreas transplants involve substantially similar patient populations. The commenter further pointed out that the transplants both serve the same clinical function--that of freeing the patient from insulin dependence. The commenter requested that CMS identify those admissions in DRG 315 that involve islet cell transplantation and determine the actual costs involved to decide whether islet cell transplant cases should be reclassified to DRG 513.

Response: We do not understand why the commenter believes that the relative weight for DRG 315 decreased by 33 percent. The FY 2006 proposed relative weight (2.0801 (see Table 5 of the FY 2006 proposed rule, 70 FR 23587)) is approximately 0.3 percent less than the FY 2005 relative weight (2.0861 (see Table 5 of the FY 2005 final rule, 69 FR 49603)). We have reviewed the MedPAR data for the first quarter of FY 2005, and have found no cases of islet cell transplantation in DRG 315. Therefore, we do not have a basis for comparison of islet cell transplantation cases to the remainder of the cases in DRG 315. We also take this opportunity to clarify that the DRGs are groupings of cases that are similar both from a clinical perspective

[[Page 47322]]

as well as a resource-intensity perspective. While the commenter's position is that the same clinical endpoint is attempted with both islet cell transplantation and pancreas transplant, the result or endpoint of treatment results is not one of the axis upon which the DRGs are structured. In addtion, the pancreas transplant involves an open abdominal procedure in which one pancreas is surgically removed and a cadaveric pancreas is transplanted. Conversely, islet cells are infused via catheter. Therefore, from the standpoint of clinical similarity, we do not believe that the cases are comparable enough to consider putting the islet cell transplantation into DRG 513.

Comment: The same commenter was concerned about payment for islet cell transplants under a National Institutes of Health (NIH) clinical trial. The commenter believed that the $18,848 islet cell isolation add-on amount is insufficient. This commenter also believed that the data used to calculate the add-on amount were inadequate to form the basis for establishing payment.

Response: The $18,848 isolation add-on amount was based on the best data available, and we remain convinced that it is an appropriate payment for isolating the islet cells from one pancreas. However, we have learned that it typically requires two isolations to acquire enough cells for one infusion. Therefore, while we will maintain the current rate of $18,848 per isolation, we will pay up to two islet isolations per discharge. If only one islet isolation is necessary, Medicare will make an add-on payment of $18,848; if two are necessary, Medicare will make an add-on payment of $37,696. In cases that require two islet isolations, CMS will pay for two pancreata. Pancreata will continue to be paid as a cost pass-through.

We will review the MedPAR data as requested using more complete FY 2005 MedPAR data during our next annual IPPS update for FY 2007.

C. Recalibration of DRG Weights

We are using the same basic methodology for the FY 2006 recalibration as we did for FY 2005 (FY 2005 IPPS final rule (69 FR 48981)). That is, we have recalibrated the DRG weights based on charge data for Medicare discharges using the most current charge information available (the FY 2004 MedPAR file).

The MedPAR file is based on fully coded diagnostic and procedure data for all Medicare inpatient hospital bills. The FY 2004 MedPAR data used in this final rule include discharges occurring between October 1, 2003 and September 30, 2004, based on bills received by CMS through March 31, 2005, from all hospitals subject to the IPPS and short-term acute care hospitals in Maryland (which are under a waiver from the IPPS under section 1814(b)(3) of the Act). The FY 2004 MedPAR file includes data for approximately 12,006,022 Medicare discharges. Discharges for Medicare beneficiaries enrolled in a Medicare+Choice managed care plan are excluded from this analysis. The data excludes CAHs, including hospitals that subsequently became CAHs after the period from which the data were taken.

The methodology used to calculate the DRG relative weights from the FY 2004 MedPAR file is as follows:

To the extent possible, all the claims were regrouped using the DRG classification revisions discussed in section II.B. of this preamble.

The transplant cases that were used to establish the relative weight for heart and heart-lung, liver and/or intestinal, and lung transplants (DRGs 103, 480, and 495) were limited to those Medicare-approved transplant centers that have cases in the FY 2004 MedPAR file. (Medicare coverage for heart, heart-lung, liver and/or intestinal, and lung transplants is limited to those facilities that have received approval from CMS as transplant centers.)

Organ acquisition costs for kidney, heart, heart-lung, liver, lung, pancreas, and intestinal (or multivisceral organs) transplants continue to be paid on a reasonable cost basis. Because these acquisition costs are paid separately from the prospective payment rate, it is necessary to subtract the acquisition charges from the total charges on each transplant bill that showed acquisition charges before computing the average charge for the DRG and before eliminating statistical outliers.

Charges were standardized to remove the effects of differences in area wage levels, indirect medical education and disproportionate share payments, and, for hospitals in Alaska and Hawaii, the applicable cost-of-living adjustment.

The average standardized charge per DRG was calculated by summing the standardized charges for all cases in the DRG and dividing that amount by the number of cases classified in the DRG. A transfer case is counted as a fraction of a case based on the ratio of its transfer payment under the per diem payment methodology to the full DRG payment for nontransfer cases. That is, a transfer case receiving payment under the transfer methodology equal to half of what the case would receive as a nontransfer would be counted as 0.5 of a total case.

Statistical outliers were eliminated by removing all cases that are beyond 3.0 standard deviations from the mean of the log distribution of both the charges per case and the charges per day for each DRG.

The average charge for each DRG was then recomputed (excluding the statistical outliers) and divided by the national average standardized charge per case to determine the relative weight.

The new weights are normalized by an adjustment factor of 1.47462 so that the average case weight after recalibration is equal to the average case weight before recalibration. This adjustment is intended to ensure that recalibration by itself neither increases nor decreases total payments under the IPPS.

When we recalibrated the DRG weights for previous years, we set a threshold of 10 cases as the minimum number of cases required to compute a reasonable weight. We used that same case threshold in recalibrating the DRG weights for FY 2006. Using the FY 2004 MedPAR data set, there are 41 DRGs that contain fewer than 10 cases. We compute the weights for these low-volume DRGs by adjusting the FY 2005 weights of these DRGs by the percentage change in the average weight of the cases in the other DRGs.

Section 1886(d)(4)(C)(iii) of the Act requires that, beginning with FY 1991, reclassification and recalibration changes be made in a manner that assures that the aggregate payments are neither greater than nor less than the aggregate payments that would have been made without the changes. Although normalization is intended to achieve this effect, equating the average case weight after recalibration to the average case weight before recalibration does not necessarily achieve budget neutrality with respect to aggregate payments to hospitals because payments to hospitals are affected by factors other than average case weight. Therefore, as we have done in past years and as discussed in section II.A.4.a. of the Addendum to this final rule, we are making a budget neutrality adjustment to ensure that the requirement of section 1886(d)(4)(C)(iii) of the Act is met.

Comment: One commenter noted that there is a reduction in the proposed weights for DRG 103 (Heart Transplant or Implant of Heart Assist System) and DRG 512 (Simultaneous Pancreas/Kidney Transplant). According to the commenter, the proposed weights represent a 6-percent reduction in DRG 103 and an 11-percent reduction in DRG 512. The commenter inquired as to whether these reductions may have

[[Page 47323]]

resulted from a methodological change in the way organ acquisition costs are addressed in the DRG weighting process.

Response: There is no change in the calculation of the DRG relative weight. Organ acquisition costs for kidney, heart, heart-lung, liver, lung, pancreas, and intestinal (or multivisceral organs) transplants continue to be paid on a reasonable cost basis. Because these acquisition costs are paid separately from the prospective payment rate, it is necessary to subtract the acquisition charges from the total charges on each transplant bill that showed acquisition charges before computing the average charge for the DRG.

As described above, the relative weight for each DRG is calculated by comparing the average charge for cases within each DRG (after removing statistical outliers) with the national average charge per case. Therefore, there are several factors that can cause a shift in the relative weight of a DRG from one fiscal year to the next. For example, even though the average charges of cases within DRG 103 increased from $278,096 in the FY 2005 final rule to $285,317 in the proposed rule, it did not increase by an equal or greater percentage than the national average. As a result, the DRG weight for DRG 103 declined. For DRG 512, the average charges decreased from $85,630 in the FY 2005 final rule to $83,113 in the proposed rule which accounts for the decline in the weight.

Comment: One commenter pointed out three typographical errors in DRG titles in Table 5 (List of Diagnosis Related Groups (DRGs), Relative Weighting Factors, Geometric and Arithmetic Mean Length of Stay) in the Addendum to the FY 2006 IPPS proposed rule. The commenter indicated that the title for DRG 14 should read ``Intracranial Hemorrhage or Cerebral Infarction'' based on the change in FY 2005 IPPS final rule (69 FR 48927) and the title for DRG 315 should read ``Other Kidney & Urinary Tract Procedures'' based on the change in the FY 2003 IPPS final rule (67 FR 49993). The commenter also pointed out a misspelling of the word ``Malignant'' in the title for DRG 276.

Response: The commenter is correct. We have made these corrections in Table 5 in the Addendum to this final rule.

D. LTC-DRG Reclassifications and Relative Weights for LTCHs for FY 2006

1. Background

In the June 6, 2003 LTCH PPS final rule (68 FR 34122), we changed the LTCH PPS annual payment rate update cycle to be effective July 1 through June 30 instead of October 1 through September 30. In addition, because the patient classification system utilized under the LTCH PPS is based directly on the DRGs used under the IPPS for acute care hospitals, in that same final rule, we explained that the annual update of the long-term care diagnosis-related group (LTC-DRG) classifications and relative weights will continue to remain linked to the annual reclassification and recalibration of the CMS-DRGs used under the IPPS. In that same final rule, we specified that we will continue to update the LTC-DRG classifications and relative weights to be effective for discharges occurring on or after October 1 through September 30 each year. Furthermore, we stated that we will publish the annual update of the LTC-DRGs in the proposed and final rules for the IPPS.

In the past, the annual update to the IPPS DRGs has been based on the annual revisions to the ICD-9-CM codes and was effective each October 1. As discussed in the FY 2005 IPPS final rule (69 FR 48954 through 48957) and in the Rate Year (RY) 2006 LTCH PPS final rule (70 FR 24173 through 24175), with the implementation of section 503(a) of Pub. L. 108-173, there is the possibility that one feature of the GROUPER software program may be updated twice during a Federal fiscal year (October 1 and April 1) as required by the statute for the IPPS. Specifically, ICD-9-CM diagnosis and procedure codes for new medical technology may be created and added to existing DRGs in the middle of the Federal fiscal year on April 1. However, this policy change will have no effect on the LTC-DRG relative weights which will continue to be updated only once a year (October 1), nor will there be any impact on Medicare payments under the LTCH PPS. The use of the ICD-9-CM code set is also compliant with the current requirements of the Transactions and Code Sets Standards regulations at 45 CFR Parts 160 and 162, promulgated in accordance with the Health Insurance Portability and Accountability Act of 1996 (HIPAA), Pub. L. 104-191.

As we explained in the FY 2006 IPPS proposed rule (70 FR 23338 through 23339), in the health care industry, historically annual changes to the ICD-9-CM codes were effective for discharges occurring on or after October 1 each year. Thus, the manual and electronic versions of the GROUPER software, which are based on the ICD-9-CM codes, were also revised annually and effective for discharges occurring on or after October 1 each year. As noted above, the patient classification system used under the LTCH PPS (LTC-DRGs) is based on the patient classification system used under the IPPS (CMS-DRGs), which historically had been updated annually and effective for discharges occurring on or after October 1 through September 30 each year. As mentioned above, the ICD-9-CM coding update process has been revised, as discussed in greater detail in the FY 2005 IPPS final rule (69 FR 48954 through 48957) and in section II.B. 14. of this final rule. Specifically, section 503(a) of Pub. L. 108-173 includes a requirement for updating ICD-9-CM codes as often as twice a year instead of the current process of annual updates on October 1 of each year. This requirement is included as part of the amendments to the Act relating to recognition of new medical technology under the IPPS. Section 503(a) of Pub L. 108-173 amended section 1886(d)(5)(K) of the Act by adding a new clause (vii) which states that ``the Secretary shall provide for the addition of new diagnosis and procedure codes in [sic] April 1 of each year, but the addition of such codes shall not require the Secretary to adjust the payment (or diagnosis-related group classification) * * * until the fiscal year that begins after such date.'' This requirement will improve the recognition of new technologies under the IPPS by accounting for those ICD-9-CM codes in the MedPAR claims data at an earlier date. Despite the fact that aspects of the GROUPER software may be updated to recognize any new technology ICD-9-CM codes, as discussed in the RY 2006 LTCH PPS final rule (70 FR 24173 through 24175) and the FY 2006 IPPS proposed rule (70 FR 23338 through 23339), there will be no impact on either LTC-DRG assignments or payments under the LTCH PPS at that time. That is, changes to the LTC-DRGs (such as the creation or deletion of LTC-DRGs) and the relative weights will continue to be updated in the manner and timing (October 1) as they are now.

As noted above and as described in both the RY 2006 LTCH PPS final rule (70 FR 24174) and the FY 2006 IPPS proposed rule (70 FR 23339), updates to the GROUPER for both the IPPS and the LTCH PPS (with respect to relative weights and the creation or deletion of DRGs) are made in the annual IPPS proposed and final rules and are effective each October 1. We explained in the FY 2005 IPPS final rule (69 FR 48955 and 48956), and in section II.B.13. of this preamble, that since we do not publish a midyear IPPS rule, April 1 code updates discussed above

[[Page 47324]]

will not be published in a midyear IPPS rule. Rather, we will assign any new diagnosis or procedure codes to the same DRG in which its predecessor code was assigned, so that there will be no impact on the DRG assignments. Any coding updates will be available through the Web sites indicated in the same rule and provided above in section II.B. of this preamble and through the Coding Clinic for ICD-9-CM. Publishers and software vendors currently obtain code changes through these sources in order to update their code books and software system. If new codes are implemented on April 1, revised code books and software systems, including the GROUPER software program, will be necessary because we must use current ICD-9-CM codes. Therefore, for purposes of the LTCH PPS, because each ICD-9-CM code must be included in the GROUPER algorithm to classify each case into a LTC-DRG, the GROUPER software program used under the LTCH PPS would need to be revised to accommodate any new codes.

As we discussed in the FY 2005 IPPS final rule (69 FR 48956) and in section II.B.14. of this preamble, in implementing section 503(a) of Pub. L. 108-173, there will only be an April 1 update if new technology codes are requested and approved. We note that any new codes created for April 1 implementation will be limited to those diagnosis and procedure code revisions primarily needed to describe new technologies and medical services. However, we reiterate that the process of discussing updates to the ICD-9-CM has been an open process through the ICD-9-CM Coordination and Maintenance Committee since 1995. Requestors will be given the opportunity to present the merits for a new code and make a clear and convincing case for the need to update ICD-9-CM codes for purposes of the IPPS new technology add-on payment process through an April 1 update.

However, as we explained in the FY 2006 IPPS proposed rule (70 FR 23339), at the October 2004 ICD-9-CM Coordination and Maintenance Committee meeting, there were no requests for an April 1, 2005 implementation of ICD-9-CM codes, and the next update to the ICD-9-CM coding system would not occur until October 1, 2005 (FY 2006). Presently, as there were no coding changes suggested for an April 1, 2005 update, the ICD-9-CM coding set implemented on October 1, 2004, will continue through September 30, 2005 (FY 2005). The update to the ICD-9-CM coding system for FY 2006 is discussed above in section II.B.14. of this preamble.

As we proposed in the FY 2006 IPPS proposed rule (70 FR 23339), in this final rule we are making revisions to the LTC-DRG classifications and relative weights, effective October 1, 2005 through September 30, 2006 (FY 2006), using the latest available data. As we proposed in that same IPPS proposed rule, the final LTC-DRGs and relative weights for FY 2006 in this final rule are based on the final IPPS DRGs (GROUPER Version 23.0) discussed in section II. of the preamble to this final rule. 2. Changes in the LTC-DRG Classifications a. Background

Section 123 of Pub. L. 106-113 specifically requires that the PPS for LTCHs be a per discharge system with a DRG-based patient classification system reflecting the differences in patient resources and costs in LTCHs while maintaining budget neutrality. Section 307(b)(1) of Pub. L. 106-554 modified the requirements of section 123 of Pub. L. 106-113 by specifically requiring that the Secretary examine ``the feasibility and the impact of basing payment under such a system

[the LTCH PPS] on the use of existing (or refined) hospital diagnosis- related groups (DRGs) that have been modified to account for different resource use of long-term care hospital patients as well as the use of the most recently available hospital discharge data.''

In accordance with section 307(b)(1) of Pub. L. 106-554 and Sec. 412.515 of our existing regulations, the LTCH PPS uses information from LTCH patient records to classify patient cases into distinct LTC-DRGs based on clinical characteristics and expected resource needs. The LTC- DRGs used as the patient classification component of the LTCH PPS correspond to the DRGs under the IPPS for acute care hospitals. Thus, as we proposed in the FY 2006 IPPS proposed rule (70 FR 23339), in this final rule, we are establishing the use of the IPPS GROUPER Version 23.0 for FY 2006 to process LTCH PPS claims for LTCH discharges occurring from October 1, 2005 through September 30, 2006. The final changes to the CMS-DRG classification system used under the IPPS for FY 2006 (GROUPER Version 23.0) are discussed in section II.B. of the preamble to this final rule.

Under the LTCH PPS, we determine relative weights for each of the DRGs to account for the difference in resource use by patients exhibiting the case complexity and multiple medical problems characteristics of LTCH patients. In a departure from the IPPS, as we discussed in the August 30, 2002 LTCH PPS final rule (67 FR 55985), which implemented the LTCH PPS, and the FY 2006 IPPS proposed rule (70 FR 23340), we use low-volume quintiles in determining the LTC-DRG weights for LTC-DRGs with less than 25 LTCH cases, because LTCHs do not typically treat the full range of diagnoses as do acute care hospitals. Specifically, we group those low-volume LTC-DRGs (LTC-DRGs with fewer than 25 cases) into 5 quintiles based on average charge per discharge. We also adjust for cases in which the stay at the LTCH is less than or equal to five-sixths of the geometric average length of stay; that is, short-stay outlier cases (Sec. 412.529), as discussed below in section II.D.4. of this preamble. b. Patient Classifications into DRGs

Generally, under the LTCH PPS, Medicare payment is made at a predetermined specific rate for each discharge; that is, payment varies by the LTC-DRG to which a beneficiary's stay is assigned. Just as cases are classified for acute care hospitals under the IPPS (see section II.B. of this preamble), cases are classified into LTC-DRGs for payment under the LTCH PPS based on the principal diagnosis, up to eight additional diagnoses, and up to six procedures performed during the stay, as well as age, sex, and discharge status of the patient. The diagnosis and procedure information is reported by the hospital using the ICD-9-CM codes.

As discussed in section II.B. of this preamble, the CMS-DRGs are organized into 25 major diagnostic categories (MDCs), most of which are based on a particular organ system of the body; the remainder involve multiple organ systems (such as MDC 22, Burns). Accordingly, the principal diagnosis determines MDC assignment. Within most MDCs, cases are then divided into surgical DRGs and medical DRGs. Some surgical and medical DRGs are further differentiated based on the presence or absence of CCs. (See section II.B. of this preamble for further discussion of surgical DRGs and medical DRGs.)

Because the assignment of a case to a particular LTC-DRG will help determine the amount that is paid for the case, it is important that the coding is accurate. As used under the IPPS, classifications and terminology used under the LTCH PPS are consistent with the ICD-9-CM and the Uniform Hospital Discharge Data Set (UHDDS), as recommended to the Secretary by the National Committee on Vital and Health Statistics (``Uniform Hospital Discharge Data: Minimum Data Set, National Center for Health Statistics, April

[[Page 47325]]

1980'') and as revised in 1984 by the Health Information Policy Council (HIPC) of the U.S. Department of Health and Human Services. We point out again that the ICD-9-CM coding terminology and the definitions of principal and other diagnoses of the UHDDS are consistent with the requirements of the Transactions and Code Sets Standards under HIPAA (45 CFR Parts 160 and 162).

The emphasis on the need for proper coding cannot be overstated. Inappropriate coding of cases can adversely affect the uniformity of cases in each LTC-DRG and produce inappropriate weighting factors at recalibration and result in inappropriate payments under the LTCH PPS. LTCHs are to follow the same coding guidelines used by acute care hospitals to ensure accuracy and consistency in coding practices. There will be only one LTC-DRG assigned per long-term care hospitalization; it will be assigned at the time of discharge of the patient. Therefore, it is mandatory that the coders continue to report the same principal diagnosis on all claims and include all diagnosis codes that coexist at the time of admission, that are subsequently developed, or that affect the treatment received. Similarly, all procedures performed during that stay are to be reported on each claim.

Upon the discharge of the patient from a LTCH, the LTCH must assign appropriate diagnosis and procedure codes from the ICD-9-CM. Completed claim forms are to be submitted electronically to the LTCH's Medicare fiscal intermediary. Medicare fiscal intermediaries enter the clinical and demographic information into their claims processing systems and subject this information to a series of automated screening processes called the Medicare Code Editor (MCE). These screens are designed to identify cases that require further review before assignment into an LTC-DRG can be made.

After screening through the MCE, each LTCH claim will be classified into the appropriate LTC-DRG by the Medicare LTCH GROUPER. The LTCH GROUPER is specialized computer software and is the same GROUPER used under the IPPS. After the LTC-DRG is assigned, the Medicare fiscal intermediary determines the prospective payment by using the Medicare LTCH PPS PRICER program, which accounts for LTCH hospital-specific adjustments and payment rates. As provided for under the IPPS, we provide an opportunity for the LTCH to review the LTC-DRG assignments made by the fiscal intermediary and to submit additional information within a specified timeframe (Sec. 412.513(c)).

The LTCH GROUPER is used both to classify past cases in order to measure relative hospital resource consumption to establish the LTC-DRG weights and to classify current cases for purposes of determining payment. The records for all Medicare hospital inpatient discharges are maintained in the MedPAR file. The data in this file are used to evaluate possible DRG classification changes and to recalibrate the DRG weights during our annual update (as discussed in section II. of this preamble). The LTC-DRG relative weights are based on data for the population of LTCH discharges, reflecting the fact that LTCH patients represent a different patient-mix than patients in short-term acute care hospitals. 3. Development of the FY 2006 LTC-DRG Relative Weights a. General Overview of Development of the LTC-DRG Relative Weights

As we stated in the August 30, 2002 LTCH PPS final rule (67 FR 55981), one of the primary goals for the implementation of the LTCH PPS is to pay each LTCH an appropriate amount for the efficient delivery of care to Medicare patients. The system must be able to account adequately for each LTCH's case-mix in order to ensure both fair distribution of Medicare payments and access to adequate care for those Medicare patients whose care is more costly. To accomplish these goals, we adjust the LTCH PPS standard Federal prospective payment system rate by the applicable LTC-DRG relative weight in determining payment to LTCHs for each case. Under the LTCH PPS, relative weights for each LTC- DRG are a primary element used to account for the variations in cost per discharge and resource utilization among the payment groups (Sec. 412.515). To ensure that Medicare patients classified to each LTC-DRG have access to an appropriate level of services and to encourage efficiency, we calculate a relative weight for each LTC-DRG that represents the resources needed by an average inpatient LTCH case in that LTC-DRG. For example, cases in an LTC-DRG with a relative weight of 2 will, on average, cost twice as much as cases in an LTC-DRG with a weight of 1. b. Data

In the FY 2006 IPPS proposed rule (70 FR 23341), we proposed to calculate the proposed LTC-DRG relative weights for FY 2006 using total Medicare allowable charges from FY 2004 Medicare hospital bill data from the December 2004 update of the MedPAR file, which were the best available data at that time, and we proposed to use the proposed Version 23.0 of the CMS GROUPER used under the IPPS (as discussed in that same proposed rule) to classify cases. To calculate the LTC-DRG relative weights for FY 2006 in this final rule, we obtained total Medicare allowable charges from FY 2004 Medicare hospital bill data from the March 2005 update of the MedPAR file, which are the most recent available data, and we used the Version 23.0 of the CMS GROUPER used under the IPPS (as discussed in section II.B. of this preamble) to classify cases. In the FY 2006 IPPS proposed rule (70 FR 23341), we stated that ``consistent with the methodology under the IPPS, we are proposing to recalculate the FY 2006 LTC-DRG relative weights based on the best available data.'' For this final rule, we are using the best available data, that is, the March 2005 update of the MedPAR file.

As we discussed in the FY 2006 IPPS proposed rule (70 FR 23341), we have excluded the data from LTCHs that are all-inclusive rate providers and LTCHs that are reimbursed in accordance with demonstration projects authorized under section 402(a) of Pub. L. 90-248 (42 U.S.C. 1395b-1) or section 222(a) of Pub. L. 92-603 (42 U.S.C. 1395b-1). Therefore, in the development of the final FY 2006 LTC-DRG relative weights, we have excluded the data of the 19 all-inclusive rate providers and the 3 LTCHs that are paid in accordance with demonstration projects that had claims in the FY 2004 MedPAR file.

In the FY 2005 IPPS final rule (69 FR 48984), we discussed coding inaccuracies that were found in the claims data for a large chain of LTCHs in the FY 2002 MedPAR file, which were used to determine the LTC- DRG relative weights for FY 2004. As we discussed in the same final rule, after notifying the large chain of LTCHs whose claims contained the coding inaccuracies to request that they resubmit those claims with the correct diagnosis, from an analysis of LTCH claims data from the December 2003 update of the FY 2003 MedPAR file, it appeared that such claims data no longer contain coding errors. Therefore, it was not necessary to correct the FY 2003 MedPAR data for the development of the FY 2005 LTC-DRGs and relative weights established in the same final rule.

As noted above, in the FY 2006 IPPS proposed rule, we proposed to calculate the proposed LTC-DRG relative weights for FY 2006 using the December 2004

[[Page 47326]]

update of the MedPAR file, which were the most recent available data at that time. As stated above, in this final rule, we are using the March 2005 update of the FY 2004 MedPAR file for the determination of the FY 2006 LTC-DRG relative weights as these are the best available data. As we discussed in the FY 2006 IPPS proposed rule (70 FR 23341), based on an analysis of LTCH claims data from the FY 2004 MedPAR file, it appears that such claims data do not contain coding inaccuracies found previously in LTCH claims data. Therefore, it was not necessary to correct the FY 2004 MedPAR data for the development of the FY 2006 LTC- DRGs and relative weights presented in that proposed rule or in this final rule.

Comment: Several commenters cited a study that concluded that the claims data used to develop the proposed LTC-DRG relative weights (that is, the December 2004 update of the FY 2004 MedPAR file) contain irregularities or errors. The commenters' concern was based on a comparison, by a private research group that was commissioned by one of the commenters, of the LTCH FY 2004 MedPAR data to the internal records of one LTCH. The commenters were specifically concerned that the MedPAR data may underrepresent interrupted stay cases and cases during which the beneficiary exhausted Medicare Part A benefits. In addition to the possible underrepresentation of interrupted stay and exhausted benefit cases, these commenters indicated that they had reviewed the FY 2004 MedPAR data used to develop the proposed FY 2006 LTC-DRG relative weights and asserted that there are some cases in the FY 2004 MedPAR file that include overstated or understated charges. They also indicated that there were ``missing'' LTCH cases that they believe should be included in the MedPAR file. The commenters further believed that the missing LTCH cases may be the consequence of ``a high level of suspended claims which were occurring due to the transition [to a different billing system during FY 2004].'' Specifically, the commenters stated that because payment for these suspended claims was received by April 2004, their claims and associated charges for these cases should have been reflected in the December 2004 update of the FY 2004 MedPAR file that was used to compute the proposed FY 2006 LTC-DRG relative weights.

The commenters believed that such errors or irregularities may be the source of the observed decrease in the average charges of many LTC- DRGs. Therefore, they urged CMS to reexamine the MedPAR data to ensure that the charges for all cases are fully accounted for in computing the final FY 2006 LTC-DRG relative weights.

The commenter who commissioned the study gave a number of examples of the alleged irregularities/errors in LTCH claims in the FY 2004 MedPAR file. The commenter's findings from a comparison of one provider's internal records and data reported in the December 2004 update of the FY 2004 MedPAR file, which were used in setting the proposed LTC-DRG relative weights, were extrapolated to all LTCHs and then the proposed FY 2006 LTC-DRG relative weights were recalculated ``to correct for these errors.'' The commenter challenged the integrity of the proposed LTC-DRG relative weights, as well as the final relative weights, which would be based on a more recent update (March 2005) of the FY 2004 MedPAR file, in keeping with our historical practice that uses the best available data for computing payment adjustments for all Medicare PPSs.

Response: After an extensive analysis of the data submitted by one of the commenters, we do not agree with the commenters' assertion that the proposed FY 2006 LTC-DRG relative weights are based on faulty claims data in the FY 2004 MedPAR file. We believe that the use of highly case-specific and interim data drawn from the claims records of one LTCH to challenge the integrity of the LTCH claims in the entire FY 2004 MedPAR file is inappropriate. Our analysis did not reveal systemic problems that would have undermined the data upon which we based the proposed FY 2006 LTC-DRG relative weights or the data upon which we are basing our final FY 2006 LTC-DRG relative weights in this final rule (as discussed above). As indicated by our analysis of the issues presented by the commenter, detailed below, we continue to believe that the March 2005 update of the FY 2004 MedPAR file is the best available data for setting the FY 2006 LTC-DRG relative weights and it accurately reflects LTCH charges per discharge.

The comments were based on the commenters' analysis of one LTCH's data and the results of that analysis were extrapolated to the universe of LTCHs. We reviewed the LTCH data used by the commenter and compared that data to the data in both the December 2004 update of the FY 2004 MedPAR file that were used to determine the proposed FY 2006 LTC-DRG relative weights and in the March 2005 update of the FY 2004 MedPAR file that are being used to determine the final FY 2006 LTC-DRG relative weights in this final rule. The commenter raised four categories of alleged problems: missing discharges related to the exhaustion of Medicare Part A benefits; inaccurate representation of interrupted stay cases; cases not reported in the MedPAR file due to ``an atypical level of suspension of LTCH claims''; and cases with incorrectly reported charges (overstated or understated). Our analysis revealed that rather than being distinct problems, three of the concerns raised by the commenters--the benefits-exhausted cases, the interrupted stay cases, and missing hospital claims--are caused by the same basic problems. That is, the December 2003 update of the FY 2004 MedPAR file did not include some patient claims from the records of the one LTCH in question. Because the MedPAR file represents a total beneficiary stay (total single episode of care) in an inpatient hospital once a beneficiary has been physically discharged from the inpatient hospital, as described below, we evaluated the reasons why such a situation could occur under normal claims processing procedures.

The MedPAR file is a discharge file for inpatient claims and, therefore, during the creation of the MedPAR file, inpatient hospital data without a discharge date would not be included. When a claim is processed for payment calculation, the data from the fiscal intermediary are included in the Medicare Common Working File (CWF), at which time payment authorization or denial will be made and, if authorized, a remittance will be generated to the provider. After the remittance is generated, the National Claims History (NCH) is updated to reflect all of the claims submitted for an entire stay, which may include one claim or multiple claims. The NCH inpatient hospital data are used in the creation of the MedPAR file and all adjustments are resolved prior to the creation of a stay record in the MedPAR file. The creation of the MedPAR file takes all claims submitted for a beneficiary at the same facility and collapses all the data so that one record is created that represents a single record of the entire stay at the facility.

A claim that is correctly coded and submitted timely by the provider will be captured by the specific update of the NCH files, the data source for the MedPAR file. However, if there are issues with the claim, the claim may be suspended. Therefore, even though the hospital will have a record of the stay, until the issue with the claim is resolved, it will not process into the NCH and, therefore, will not be recorded in the MedPAR file. Issues leading to claim suspension may

[[Continued on page 47327]]

From the Federal Register Online via GPO Access [wais.access.gpo.gov] ]

[[pp. 47327-47376]] Medicare Program; Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2006 Rates

[[Continued from page 47326]]

[[Page 47327]]

include submission-systems failures by the provider, including the absence of crucial information or incorrect coding of patient status by the provider. Alternatively, issues may arise during the fiscal intermediary processing of the claim, as a result of data processing problems or broader standard systems issues. The fiscal intermediary may also delay processing the claim pending resolution of policy issues in specific situations. A fiscal intermediary may need to contact a subject-matter specialist at Medicare, for example, for assistance in determining whether a particular atypical patient discharge, treatment, and readmittance scenario would be governed by the payment rules established under either of the interrupted stay policies at Sec. 412.531.

Therefore, there are several reasons why claims could be held in suspension and hence not be ``resolved'' either for payment purposes or for inclusion in the MedPAR file. We understand that, at any one time, there may be as many as 25 percent of a hospital's claims in suspension pending resolution of one or more of the above issues. This statistic is not reflective of any unique problems in the processing procedure but rather is a standard feature of a dynamic claims payment process. In recognition of this fact, and in order to enable a cash flow to a provider where there may be a disproportionate number of unresolved claims in suspension, our regulations at Sec. 412.541(f) provide for accelerated payments, which are reconciled with actual remittances at a future date.

The commenter's first concern was that a substantial number of benefits-exhaust claims from the one LTCH were not included in the March 2004 update of the FY 2004 MedPAR file. Our case-level analysis revealed several reasons for this, which are discussed below. Primarily, we believe that there has been some degree of confusion on by that LTCH as to the policy distinction established under the LTCH PPS between a discharge for payment purposes and a patient's physical discharge. In the August 30, 2002 final rule for the LTCH PPS, we established regulations at Sec. 412.503 specifying that a Medicare patient is considered ``discharged'' for payment purposes when the patient no longer has any Medicare covered days (that is, when Medicare Part A benefits are exhausted). At that point, a LTCH may submit a ``discharge'' claim to its fiscal intermediary and Medicare will issue a payment for covered care (CMS Pub. 100-4 Chapter 1, Section 50.2) delivered until the benefits were exhausted. The patient may continue to receive care at the LTCH, but Medicare Part A will no longer be financially responsible for that treatment. In that same final rule, we also established that we would include data for all inpatient days that a Medicare beneficiary was physically in the LTCH for purposes of meeting the length of stay requirements to qualify as a LTCH as set forth under Sec. 412.23(e) (67 FR 55974) and for developing LTC-DRG relative weights (67 FR 55984). Therefore, for purposes of these two policies, data from the fiscal year during which the patient is physically discharged from the LTCH will include the total day count for the patient's entire stay as well as the total charges for the entire length of stay, including data from noncovered days, even where the Medicare payment to the LTCH was made in a prior fiscal year, based on the earlier bill submitted by the LTCH when the patient's benefits exhausted.

In response to the commenter's allegation that the data from the December 2004 update of the FY 2004 MedPAR file did not capture 16 of 35 benefits-exhaust claims for one specific LTCH, CMS' analysis revealed that 5 of these 16 cases noted by the commenter are, in fact, included in the more recent March 2005 update of the FY 2004 MedPAR file. This indicates that if the bill did not appear on the earlier December update due to a processing suspension, these 5 cases appear in the March 2005 update of the MedPAR file because the issue for which the bill was suspended has been resolved by that time. Furthermore, an additional 7 of the 16 claims that the hospital identified as ``discharged'' represented beneficiaries who were still in the hospital at the end of FY 2004 (September 30, 2004), even though Medicare was no longer making payments for their care (and they had been ``discharged for payment purposes'' under Sec. 412.503). As noted above, only at physical discharge will data be included in the corresponding MedPAR file. Once those 7 patients are discharged physically from the LTCH in question, their data will appear in the MedPAR file for the fiscal year of their discharge. Accordingly, we do not believe that the absence from the March 2005 update of the FY 2004 MedPAR file of the four discharges for this one LTCH represents a systematic and serious underrepresentation of benefits-exhaust cases in the LTCH FY 2004 MedPAR file.

The commenter also claimed that the MedPAR file had inaccurately reported interrupted stay cases, that is, a LTCH stay that has an intervening stay at an acute care hospital for 9 days or less, an IRF for 27 days or less, or a SNF for 45 days or less during the LTCH stay (Sec. 412.531). The one LTCH upon which the commenter bases his concerns had records of 102 interrupted stay cases discharged during FY 2004. Of these, it is claimed that 44 were reported correctly in the December 2004 update of the FY 2004 MedPAR file upon which the proposed LTC-DRG relative weights were based. If an episode of care is governed by the greater than 3 days interruption of stay policy, both segments of the stay at the LTCH are paid as one. The commenter claimed that, in such cases, only one-half of particular interrupted stay cases in that LTCH that were reported were included in the December 2004 update of the FY 2004 MedPAR file. The commenter also claimed that in other interrupted stay cases, the entire stay was absent from the December 2004 update of the FY 2004 MedPAR file. We reviewed the commenter's claims and concluded that most of these cases are included in the recent March 2005 update of the FY 2004 MedPAR file. We believe that these cases were not included in the December 2004 update of the FY 2004 MedPAR file because the provider's final bill was in suspension.

It is likely that the cases appear in the March 2005 update of the FY 2004 MedPAR file because the patient was finally physically discharged or issues relating to the claim were otherwise settled and the claims were no longer in suspension. Other claims reported by the LTCH but still not included in the March 2005 update of the FY 2004 MedPAR file are appropriately not in the MedPAR file because they are still in suspension for various reasons (as noted above and discussed in greater detail below).

As stated above, there may be one or even several valid and appropriate reasons why the interrupted stay cases are suspended. We understand that the initial implementation of certain LTCH PPS system changes resulted in problems, including the mechanics of claim submission. Specifically, for many fiscal intermediaries, the implementation of the 3-day or less interruption of stay policy at Sec. 412.531(a) (69 FR 25690) initially led to submission of overlapping claims, inappropriate payments, recoupment of payments, and subsequent withdrawal and resubmission of claims, and required considerable provider education and resulted in initial suspension of the claims during FY 2004. However, this is no longer a significant problem for fiscal intermediaries. In fact, the fiscal intermediary that services the LTCH cited by the commenter noted that several of its providers worked

[[Page 47328]]

aggressively and in a timely manner to ensure that their claims governed by this policy were being submitted according to CMS instructions, and paid and reported accurately. However, other LTCHs were still working to rectify their claims submission procedures under the new policies or their internal records. Among those LTCHs that apparently had data submission and payment problems, the fiscal intermediary identified the LTCH that was the subject of the commenter's original data collection. Therefore, while we acknowledge that there were initial claims processing difficulties with interrupted stay cases, based on our conversations with the fiscal intermediary that services approximately two-thirds of all LTCHs, as well as with the fiscal intermediary that services the LTCH in question and 10 other LTCHs, we do not believe that there continues to be a significant issue. Furthermore, we believe that, currently, for the vast majority of LTCHs, internal records are consistent with the actual payment adjustments made by their fiscal intermediaries that are reported in the MedPAR file. However, the few LTCHs that experience an inconsistency between their internal records and the data reported in the MedPAR file do so as a result of provider specific billing issues which are in no way indicative of a widespread or even a significant problem with the integrity of the FY 2003 MedPAR data.

As noted above, the commenter believes that ``an atypical level of suspension of LTCH claims'' results from dealing with the FY 2004 conversion from the Arkansas Part A Standard System (APASS) billing system to the Fiscal Intermediary Share System (FISS) billing system. The commenter believed this transition resulted in inaccurate and underreported claims in the FY 2004 MedPAR data. While there were some initial difficulties with the system transition, our analysis of the MedPAR data again indicates that those difficulties have been addressed and, in fact, the MedPAR data accurately reflect provider billings and are reliable.

Based on discussions with the fiscal intermediaries that process the vast majority of LTCH bills, we conclude that, although initially there were some problems with the system's processing of a limited number of claims that were impacted by either the 3-day or less interrupted stay policy (Sec. 412.531(a)) or cases of exhaustion of Medicare benefits, the problems were typically resolved in a timely manner and the claims are reflected in the March 2005 update of the FY 2004 MedPAR file. Furthermore, the fiscal intermediary that serves the LTCH in question also noted experiencing some difficulties with its conversion to the FISS billing system originally, but presently, it is no longer experiencing a significant number of suspended claims as a result of those issues.

We also analyzed the commenter's assertions that, for a number of the LTCH bills in question, the LTCH's internal records of charges included either additional or fewer charges than the amount reported as the charges in the December 2004 update of the FY 2004 MedPAR file. The commenter believed that, because the FY 2004 MedPAR file does not reflect all of the bill's charges for this LTCH, there is a systemic problem that affects the calculations of the FY 2006 LTC-DRG relative weights. We believe the FY 2004 MedPAR file is providing cases with accurate charge data for that fiscal year. Because all Medicare charges that are reported in the MedPAR file are taken directly from claims submitted by providers, in order to further evaluate the commenter's assertion, we requested that the fiscal intermediary serving this LTCH review claims that the commenter alleged exemplified the ``discrepancy'' between the LTCH charges identified in its records and those that appear in the FY 2004 MedPAR file. A comparison of the electronic claims submitted by the LTCH to the fiscal intermediary did not reveal any inconsistencies. That is, the charges on the electronic claims for those cases matched those charges that appeared in the most recent update (March 2005) of the MedPAR file. Therefore, the MedPAR data are consistent with charge data submitted by the LTCH to CMS. Furthermore, as we analyzed each of the commenter's specific allegations of systemic flaws in the FY 2004 MedPAR data, we have concluded that the only way that the actual charges could be higher or lower on the hospital's own records than those charges that appear on the claim in the NCH (upon which the MedPAR file is derived) would be if the provider did not include those charges on the bill submitted to the fiscal intermediary for processing. We note that this issue of a discrepancy between billed charges and the MedPAR data is not an issue for other providers. Therefore, we believe that any inconsistencies between charges for a few cases as listed in the internal records of one LTCH and those reported for those same cases in the FY 2004 MedPAR file are due to internal data reporting practices of a specific LTCH and are not indicative of a widespread problem with the reporting of charges for LTCHs throughout the country in the FY 2004 MedPAR data that affects the final LTC-DRG relative weights.

Based upon our detailed analysis of the commenter's assertions, we believe that there are no systematic errors in the LTCH FY 2004 MedPAR data and we continue to believe it is appropriate to base the FY 2006 LTC-DRG relative weights on the March 2005 update of the FY 2004 MedPAR file. We believe that the December 2004 update of the FY 2004 MedPAR file that we used to determine the proposed LTC-DRG relative weights for FY 2006 in the FY 2006 IPPS proposed rule reflected the best available data at that time. Moreover, we maintain that calculating the final LTC-DRG payment weights set forth in this final rule using the March 2005 update of the FY 2004 MedPAR file eliminates most of the issues raised by the commenter, even with the specific claims submitted by the one LTCH cited by the commenter. Furthermore, based on our analysis, we conclude that many of the issues experienced by that LTCH were unique to that hospital and were not systemic issues.

In summary, as explained above, we do not believe there is evidence to support the contention that there is a systemic flaw in the LTCH FY 2004 MedPAR data or the integrity of the FY 2006 final LTC-DRG relative weights. Rather, we believe that extrapolation to the entire universe of LTCHs of the issues of one particular LTCH with its own submission and reporting history as proof of the unreliability of our FY 2004 MedPAR data is both misleading and inaccurate. Therefore, in this final rule, we are using the LTCH claims data from the March 2005 update of the FY 2004 MedPAR file to determine the FY 2006 LTC-DRG relative weights using the methodology described below. c. Hospital-Specific Relative Value Methodology

By nature, LTCHs often specialize in certain areas, such as ventilator-dependent patients and rehabilitation and wound care. Some case types (DRGs) may be treated, to a large extent, in hospitals that have, from a perspective of charges, relatively high (or low) charges. This nonarbitrary distribution of cases with relatively high (or low) charges in specific LTC-DRGs has the potential to inappropriately distort the measure of average charges. To account for the fact that cases may not be randomly distributed across LTCHs, we use a hospital- specific relative value method to calculate the LTC-DRG relative weights instead of the

[[Page 47329]]

methodology used to determine the DRG relative weights under the IPPS described in section II.C. of this preamble. We believe this method will remove this hospital-specific source of bias in measuring LTCH average charges. Specifically, we reduce the impact of the variation in charges across providers on any particular LTC-DRG relative weight by converting each LTCH's charge for a case to a relative value based on that LTCH's average charge.

Under the hospital-specific relative value method, we standardize charges for each LTCH by converting its charges for each case to hospital-specific relative charge values and then adjusting those values for the LTCH's case-mix. The adjustment for case-mix is needed to rescale the hospital-specific relative charge values (which, by definition, averages 1.0 for each LTCH). The average relative weight for a LTCH is its case-mix, so it is reasonable to scale each LTCH's average relative charge value by its case-mix. In this way, each LTCH's relative charge value is adjusted by its case-mix to an average that reflects the complexity of the cases it treats relative to the complexity of the cases treated by all other LTCHs (the average case- mix of all LTCHs).

In accordance with the methodology established under Sec. 412.523, as implemented in the August 30, 2002 LTCH PPS final rule (67 FR 55989 through 55991), we standardize charges for each case by first dividing the adjusted charge for the case (adjusted for short-stay outliers under Sec. 412.529 as described in section II.D.4. (step 3) of this preamble) by the average adjusted charge for all cases at the LTCH in which the case was treated. Short-stay outliers under Sec. 412.529 are cases with a length of stay that is less than or equal to five-sixths the average length of stay of the LTC-DRG. The average adjusted charge reflects the average intensity of the health care services delivered by a particular LTCH and the average cost level of that LTCH. The resulting ratio is multiplied by that LTCH's case-mix index to determine the standardized charge for the case.

Multiplying by the LTCH's case-mix index accounts for the fact that the same relative charges are given greater weight in a LTCH with higher average costs than they would at a LTCH with low average costs which is needed to adjust each LTCH's relative charge value to reflect its case-mix relative to the average case-mix for all LTCHs. Because we standardize charges in this manner, we count charges for a Medicare patient at a LTCH with high average charges as less resource intensive than they would be at a LTCH with low average charges. For example, a $10,000 charge for a case in a LTCH with an average adjusted charge of $17,500 reflects a higher level of relative resource use than a $10,000 charge for a case in a LTCH with the same case-mix, but an average adjusted charge of $35,000. We believe that the adjusted charge of an individual case more accurately reflects actual resource use for an individual LTCH because the variation in charges due to systematic differences in the markup of charges among LTCHs is taken into account. d. Low-Volume LTC-DRGs

In order to account for LTC-DRGs with low-volume (that is, with fewer than 25 LTCH cases), in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55984), we group those ``low-volume LTC-DRGs'' (that is, DRGs that contained between 1 and 24 cases annually) into one of five categories (quintiles) based on average charges, for the purposes of determining relative weights. In the FY 2006 IPPS proposed rule (70 FR 23341), we stated that we would continue to employ this treatment of low volume LTC-DRGs in determining the FY 2006 LTC-DRG relative weights using the best available LTCH data. In that same proposed rule, using LTCH cases from the December 2004 update of the FY 2004 MedPAR file, we identified 172 LTC-DRGs that contained between 1 and 24 cases. For this final rule, using LTCH cases from the March 2005 update of the FY 2004 MedPAR file, we identified 171 LTC-DRGs that contained between 1 and 24 cases. This list of LTC-DRGs was then divided into one of the 5 low-volume quintiles, each containing a minimum of 34 LTC-DRGs (171/5 = 34 with 1 LTC-DRG as the remainder). In accordance with our established methodology, we then make an assignment to a specific low-volume quintile by sorting the low-volume LTC-DRGs in ascending order by average charge. For this final rule, this results in an assignment to a specific low volume quintile of the sorted 171 low-volume LTC-DRGs by ascending order by average charge. Because the number of LTC-DRGs with less than 25 LTCH cases is not evenly divisible by five, the average charge of the low-volume LTC-DRG was used to determine which low-volume quintile received the additional LTC-DRG. After sorting the 171 low- volume LTC-DRGs in ascending order, we group the first fifth of low- volume LTC-DRGs with the lowest average charge into Quintile 1. The highest average charge cases are grouped into Quintile 5. Since the average charge of the 69th LTC-DRG in the sorted list is closer to the 68th LTC-DRG's average charge (assigned to Quintile 2) than to the average charge of the 70th LTC-DRG in the sorted list (to be assigned to Quintile 3), we placed it into Quintile 2. This process was repeated through the remaining low-volume LTC-DRGs so that 1 low-volume quintile contains 35 LTC-DRGs and 4 low-volume quintiles contain 34 LTC-DRGs.

In order to determine the relative weights for the LTC-DRGs with low volume for FY 2006, in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55984), we used the five low-volume quintiles described above. The composition of each of the five low-volume quintiles shown in the chart below was used in determining the LTC-DRG relative weights for FY 2006. We determined a relative weight and (geometric) average length of stay for each of the five low-volume quintiles using the formula that we apply to the regular LTC-DRGs (25 or more cases), as described below in section II.D.4. of this preamble. We assigned the same relative weight and average length of stay to each of the LTC-DRGs that make up that low- volume quintile. We note that, as this system is dynamic, it is possible that the number and specific type of LTC-DRGs with a low volume of LTCH cases will vary in the future. We use the best available claims data in the MedPAR file to identify low-volume LTC-DRGs and to calculate the relative weights based on our methodology.

Composition of Low-Volume Quintiles for FY 2006

LTC-DRG

Description

QUINTILE 1

17................................... NONSPECIFIC CEREBROVASCULAR DISORDERS W/O CC. 25................................... SEIZURE & HEADACHE AGE >17 W/O CC.

[[Page 47330]]

65................................... DYSEQUILIBRIUM. 69................................... OTITIS MEDIA & URI AGE >17 W/O CC. 86................................... PLEURAL EFFUSION W/O CC. 95................................... PNEUMOTHORAX W/O CC. 102.................................. OTHER RESPIRATORY SYSTEM DIAGNOSES W/O CC. 133.................................. ATHEROSCLEROSIS W/O CC. 140.................................. ANGINA PECTORIS. 142 ***.............................. SYNCOPE & COLLAPSE W/O CC. 143.................................. CHEST PAIN. 171.................................. OTHER DIGESTIVE SYSTEM O.R. PROCEDURES W/O CC. 175.................................. G.I. HEMORRHAGE W/O CC. 219.................................. LOWER EXTREM & HUMER PROC EXCEPT HIP, FOOT, FEMUR AGE >17 W/O CC. 237.................................. SPRAINS, STRAINS, & DISLOCATIONS OF HIP, PELVIS & THIGH. 241.................................. CONNECTIVE TISSUE DISORDERS W/O CC. 246.................................. NON-SPECIFIC ARTHROPATHIES. 251.................................. FX, SPRN, STRN & DISL OF FOREARM, HAND, FOOT AGE >17 W/O CC. 262.................................. BREAST BIOPSY & LOCAL EXCISION FOR NON-MALIGNANCY. 273.................................. MAJOR SKIN DISORDERS W/O CC. 281.................................. TRAUMA TO THE SKIN, SUBCUT TISS & BREAST AGE >17 W/O CC. 284.................................. MINOR SKIN DISORDERS W/O CC. 301.................................. ENDOCRINE DISORDERS W/O CC. 305.................................. KIDNEY, URETER & MAJOR BLADDER PROC FOR NON-NEOPL W/O CC. 312.................................. URETHRAL PROCEDURES, AGE >17 W CC. 319.................................. KIDNEY & URINARY TRACT NEOPLASMS W/O CC. 328.................................. URETHRAL STRICTURE AGE >17 W CC. 344.................................. OTHER MALE REPRODUCTIVE SYSTEM O.R. PROCEDURES FOR MALIGNANCY. 428.................................. DISORDERS OF PERSONALITY & IMPULSE CONTROL. 431.................................. CHILDHOOD MENTAL DISORDERS. 441.................................. HAND PROCEDURES FOR INJURIES. 445.................................. TRAUMATIC INJURY AGE >17 W/O CC. 509.................................. FULL THICKNESS BURN W/O SKIN GRFT OR INH INJ W/O CC OR SIG TRAUMA. 511.................................. NON-EXTENSIVE BURNS W/O CC OR SIGNIFICANT TRAUMA .

QUINTILE 2

11................................... NERVOUS SYSTEM NEOPLASMS W/O CC. 29................................... TRAUMATIC STUPOR & COMA, COMA 17 W/O CC. 44................................... ACUTE MAJOR EYE INFECTIONS. 46................................... OTHER DISORDERS OF THE EYE AGE >17 W CC. 83................................... MAJOR CHEST TRAUMA W CC. 93................................... INTERSTITIAL LUNG DISEASE W/O CC. 97................................... BRONCHITIS & ASTHMA AGE >17 W/O CC. 122.................................. CIRCULATORY DISORDERS W AMI W/O MAJOR COMP, DISCHARGED ALIVE. 128.................................. DEEP VEIN THROMBOPHLEBITIS. 136.................................. CARDIAC CONGENITAL & VALVULAR DISORDERS AGE >17 W/O CC. 139.................................. CARDIAC ARRHYTHMIA & CONDUCTION DISORDERS W/O CC. 151.................................. PERITONEAL ADHESIOLYSIS W/O CC. 173.................................. DIGESTIVE MALIGNANCY W/O CC. 206.................................. DISORDERS OF LIVER EXCEPT MALIG, CIRR, ALC HEPA W/O CC. 208.................................. DISORDERS OF THE BILIARY TRACT W/ O CC. 250.................................. FX, SPRN, STRN & DISL OF FOREARM, HAND, FOOT AGE >17 W CC. 254.................................. FX, SPRN, STRN & DISL OF UPARM, LOWLEG EX FOOT AGE >17 W/O CC. 259.................................. SUBTOTAL MASTECTOMY FOR MALIGNANCY W CC. 276.................................. NON-MALIGANT BREAST DISORDERS. 293.................................. OTHER ENDOCRINE, NUTRIT & METAB O.R. PROC W/O CC. 306.................................. PROSTATECTOMY W CC. 325.................................. KIDNEY & URINARY TRACT SIGNS & SYMPTOMS AGE >17 W CC. 332.................................. OTHER KIDNEY & URINARY TRACT DIAGNOSES AGE >17 W/O CC. 334.................................. MAJOR MALE PELVIC PROCEDURES W CC. 336.................................. TRANSURETHRAL PROSTATECTOMY W CC. 347.................................. MALIGNANCY, MALE REPRODUCTIVE SYSTEM, W/O CC. 348.................................. BENIGN PROSTATIC HYPERTROPHY W CC. 399.................................. RETICULOENDOTHELIAL & IMMUNITY DISORDERS W/O CC. 404.................................. LYMPHOMA & NON-ACUTE LEUKEMIA W/O CC. 425.................................. ACUTE ADJUSTMENT REACTION & PSYCHOLOGICAL DYSFUNCTION. 432.................................. OTHER MENTAL DISORDER DIAGNOSES. 433.................................. ALCOHOL/DRUG ABUSE OR DEPENDENCE, LEFT AMA. 447.................................. ALLERGIC REACTIONS AGE >17. 484.................................. CRANIOTOMY FOR MULTIPLE SIGNIFICANT TRAUMA. 503.................................. KNEE PROCEDURES W/O PDX OF INFECTION.

[[Page 47331]]

QUINTILE 3

8.................................... PERIPH & CRANIAL NERVE & OTHER NERV SYST PROC W/O CC. 21................................... VIRAL MENINGITIS. 31................................... CONCUSSION AGE >17 W CC. 61................................... MYRINGOTOMY W TUBE INSERTION AGE >17. 67................................... EPIGLOTTITIS. 100.................................. RESPIRATORY SIGNS & SYMPTOMS W/O CC. 110.................................. MAJOR CARDIOVASCULAR PROCEDURES W CC. 119.................................. VEIN LIGATION & STRIPPING. 125.................................. CIRCULATORY DISORDERS EXCEPT AMI, W CARD CATH W/O COMPLEX DIAG. 152.................................. MINOR SMALL & LARGE BOWEL PROCEDURES W CC. 177.................................. UNCOMPLICATED PEPTIC ULCER W CC. 178.................................. UNCOMPLICATED PEPTIC ULCER W/O CC. 181.................................. G.I. OBSTRUCTION W/O CC. 185.................................. DENTAL & ORAL DIS EXCEPT EXTRACTIONS & RESTORATIONS, AGE >17. 193.................................. BILIARY TRACT PROC EXCEPT ONLY CHOLECYST W OR W/O C.D.E. W CC. 195.................................. CHOLECYSTECTOMY W C.D.E. W CC. 197.................................. CHOLECYSTECTOMY EXCEPT BY LAPAROSCOPE W/O C.D.E. W CC. 223.................................. MAJOR SHOULDER/ELBOW PROC, OR OTHER UPPER EXTREMITY PROC W CC. 227.................................. SOFT TISSUE PROCEDURES W/O CC. 235.................................. FRACTURES OF FEMUR. 266.................................. SKIN GRAFT &/OR DEBRID EXCEPT FOR SKIN ULCER OR CELLULITIS W/O CC. 270.................................. OTHER SKIN, SUBCUT TISS & BREAST PROC W/O CC. 274.................................. MALIGNANT BREAST DISORDERS W CC. 295.................................. DIABETES AGE 0-35. 308.................................. MINOR BLADDER PROCEDURES W CC. 369.................................. MENSTRUAL & OTHER FEMALE REPRODUCTIVE SYSTEM DISORDERS. 424.................................. O.R. PROCEDURE W PRINCIPAL DIAGNOSES OF MENTAL ILLNESS. 443.................................. OTHER O.R. PROCEDURES FOR INJURIES W/O CC. 449.................................. POISONING & TOXIC EFFECTS OF DRUGS AGE >17 W CC. 454.................................. OTHER INJURY, POISONING & TOXIC EFFECT DIAG W CC. 467.................................. OTHER FACTORS INFLUENCING HEALTH STATUS. 507.................................. FULL THICKNESS BURN W SKIN GRFT OR INHAL INJ W/O CC OR SIG TRAUMA. 531.................................. SPINAL PROCEDURES WITH CC. 532.................................. SPINAL PROCEDURES WITHOUT CC.

QUINTILE 4

22................................... HYPERTENSIVE ENCEPHALOPATHY. 40................................... EXTRAOCULAR PROCEDURES EXCEPT ORBIT AGE >17. 63................................... OTHER EAR, NOSE, MOUTH & THROAT O.R. PROCEDURES. 117.................................. CARDIAC PACEMAKER REVISION EXCEPT DEVICE REPLACEMENT. 118.................................. CARDIAC PACEMAKER DEVICE REPLACEMENT. 124.................................. CIRCULATORY DISORDERS EXCEPT AMI, W CARD CATH & COMPLEX DIAG. 150.................................. PERITONEAL ADHESIOLYSIS W CC. 157.................................. ANAL & STOMAL PROCEDURES W CC. 168.................................. MOUTH PROCEDURES W CC. 191.................................. PANCREAS, LIVER & SHUNT PROCEDURES W CC. 211.................................. HIP & FEMUR PROCEDURES EXCEPT MAJOR JOINT AGE >17 W/O CC. 216.................................. BIOPSIES OF MUSCULOSKELETAL SYSTEM & CONNECTIVE TISSUE. 228.................................. MAJOR THUMB OR JOINT PROC, OR OTH HAND OR WRIST PROC W CC. 288.................................. O.R. PROCEDURES FOR OBESITY. 299.................................. INBORN ERRORS OF METABOLISM. 303.................................. KIDNEY, URETER & MAJOR BLADDER PROCEDURES FOR NEOPLASM. 310.................................. TRANSURETHRAL PROCEDURES W CC. 323.................................. URINARY STONES W CC, &/OR ESW LITHOTRIPSY. 339.................................. TESTES PROCEDURES, NON-MALIGNANCY AGE >17. 341.................................. PENIS PROCEDURES. 360.................................. VAGINA, CERVIX & VULVA PROCEDURES. 406.................................. MYELOPROLIF DISORD OR POORLY DIFF NEOPL W MAJ O.R. PROC W CC. 408.................................. MYELOPROLIF DISORD OR POORLY DIFF NEOPL W OTHER O.R. PROC. 419.................................. FEVER OF UNKNOWN ORIGIN AGE >17 W CC. 476.................................. PROSTATIC O.R. PROCEDURE UNRELATED TO PRINCIPAL DIAGNOSIS. 497.................................. SPINAL FUSION W CC. 500.................................. BACK & NECK PROCEDURES EXCEPT SPINAL FUSION W/O CC. 502.................................. KNEE PROCEDURES W PDX OF INFECTION W/O CC. 505.................................. EXTENSIVE BURN OR FULL THICKNESS BURNS WITH MECH VENT 96+ HOURS WITHOUT SKIN GRAFT. 506.................................. FULL THICKNESS BURN W SKIN GRAFT OR INHAL INJ W CC OR SIG TRAUMA. 539.................................. LYMPHOMA AND LEUKEMIA WITH MAJOR O.R. PROCEDURE WITH CC. 551.................................. PERMANENT CARDIAC PACEMAKER IMPLANT WITH MAJOR CV DIAGNOSIS OR AICD LEAD OR GNRTR. 552.................................. OTHER PERMANENT CARDIAC PACEMAKER IMPLANT WITHOUT MAJOR CV DIAGNOSIS.

[[Page 47332]]

555.................................. PERCUTANEOUS CARDIOVASCULAR PROC WITH MAJOR CV DIAGNOSIS. 556*................................. PERCUTANEOUS CARDIOVASCULAR PROC WITH NON-DRUG-ELUTING STENT WITHOUT MAJOR CV DIAGNOSIS. 557*................................. PERCUTANEOUS CARDIOVASCULAR PROC WITH DRUG-ELUTING STENT WITH MAJOR CV DIAGNOSIS.

QUINTILE 5

1.................................... CRANIOTOMY AGE >17 W CC. 75................................... MAJOR CHEST PROCEDURES. 77................................... OTHER RESP SYSTEM O.R. PROCEDURES W/O CC. 154.................................. STOMACH, ESOPHAGEAL & DUODENAL PROCEDURES AGE >17 W CC. 161.................................. INGUINAL & FEMORAL HERNIA PROCEDURES AGE >17 W CC. 200.................................. HEPATOBILIARY DIAGNOSTIC PROCEDURE FOR NON-MALIGNANCY. 210.................................. HIP & FEMUR PROCEDURES EXCEPT MAJOR JOINT AGE >17 W CC. 218.................................. LOWER EXTREM & HUMER PROC EXCEPT HIP, FOOT, FEMUR AGE >17 W CC. 230.................................. LOCAL EXCISION & REMOVAL OF INT FIX DEVICES OF HIP & FEMUR. 268.................................. SKIN, SUBCUTANEOUS TISSUE & BREAST PLASTIC PROCEDURES. 290.................................. THYROID PROCEDURES. 304.................................. KIDNEY, URETER & MAJOR BLADDER PROC FOR NON-NEOPL W CC. 345.................................. OTHER MALE REPRODUCTIVE SYSTEM O.R. PROC EXCEPT FOR MALIGNANCY. 364.................................. D&C, CONIZATION EXCEPT FOR MALIGNANCY. 365.................................. OTHER FEMALE REPRODUCTIVE SYSTEM O.R. PROCEDURES. 394.................................. OTHER O.R. PROCEDURES OF THE BLOOD AND BLOOD FORMING ORGANS. 401.................................. LYMPHOMA & NON-ACUTE LEUKEMIA W OTHER O.R. PROC W CC. 471.................................. BILATERAL OR MULTIPLE MAJOR JOINT PROCS OF LOWER EXTREMITY. 482.................................. TRACHEOSTOMY FOR FACE, MOUTH & NECK DIAGNOSES. 486.................................. OTHER O.R. PROCEDURES FOR MULTIPLE SIGNIFICANT TRAUMA. 488.................................. HIV W EXTENSIVE O.R. PROCEDURE. 491.................................. MAJOR JOINT & LIMB REATTACHMENT PROCEDURES OF UPPER EXTREMITY. 493.................................. LAPAROSCOPIC CHOLECYSTECTOMY W/O C.D.E. W CC. 499.................................. BACK & NECK PROCEDURES EXCEPT SPINAL FUSION W CC. 501.................................. KNEE PROCEDURES W PDX OF INFECTION W CC. 515.................................. CARDIAC DEFIBRILATOR IMPLANT W/O CARDIAC CATH. 519.................................. CERVICAL SPINAL FUSION W CC. 529.................................. VENTRICULAR SHUNT PROCEDURES W CC. 533.................................. EXTRACRANIAL VASCULAR PROCEDURES WITH CC. 543.................................. CRANIOTOMY W IMPLANT OF CHEMO AGENT OR ACUTE COMPLEX CNS PDX. 544.................................. MAJOR JOINT REPLACEMENT OR REATTACHMENT OF LOWER EXTREMITY. 545.................................. REVISION OF HIP OR KNEE REPLACEMENT. 556 **............................... PERCUTANEOUS CARDIOVASCULAR PROC WITH NON-DRUG-ELUTING STENT WITHOUT MAJOR CV DIAGNOSIS. 557 **............................... PERCUTANEOUS CARDIOVASCULAR PROC WITH DRUG-ELUTING STENT WITH MAJOR CV DIAGNOSIS.

* One of the original 171 low-volume LTC-DRGs initially assigned to a different low-volume quintile; reassigned to this low-volume quintile in addressing nonmonotonicity (see step 4 below). ** One of the original 171 low-volume LTC-DRGs initially assigned to this low-volume quintile; reassigned to a different low-volume quintile in addressing nonmonotonicity (see step 4 below). *** One of the original 171 low-volume LTC-DRGs initially assigned to this low-volume quintile; removed from this low-volume quintile in addressing nonmonotonicity (see step 4 below).

4. Steps for Determining the FY 2006 LTC-DRG Relative Weights

As we noted in the FY 2006 IPPS proposed rule (70 FR 23346), the FY 2006 LTC-DRG relative weights are determined in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55989 through 55991). In summary, LTCH cases must be grouped in the appropriate LTC-DRG, while taking into account the low-volume LTC-DRGs as described above, before the FY 2006 LTC-DRG relative weights can be determined. After grouping the cases in the appropriate LTC-DRG, we calculated the relative weights for FY 2006 in this final rule by first removing statistical outliers and cases with a length of stay of 7 days or less, as discussed in greater detail below. Next, we adjusted the number of cases in each LTC-DRG for the effect of short-stay outlier cases under Sec. 412.529, as also discussed in greater detail below. The short-stay adjusted discharges and corresponding charges are used to calculate ``relative adjusted weights'' in each LTC-DRG using the hospital-specific relative value method described above.

Comment: A few commenters expressed concern regarding what they believed to be a proposed change in the methodology to compute the LTC- DRG relative weights. Specifically, they asserted that removing statistical outlier cases and cases with a length of stay of 7 days or less may inappropriately remove too many cases from the relative weight calculations. The commenters believed that, by narrowing the universe of cases used to compute the LTC-DRG relative weights, the principle of averaging that is a fundamental feature of a PPS would be eroded or distorted.

Response: We did not propose any policy change in the methodology for determining the LTC-DRG relative weights for FY 2006 in the FY 2006 IPPS proposed rule. The commenters are mistaken in their belief that we did. Rather, the six steps for determining the proposed FY 2006 LTC-DRG relative weights presented in the FY 2006 IPPS proposed rule (70 FR 23346 through 23353) are the same steps that we have used to determine the LTC-DRG relative weights since the implementation of the LTCH PPS in FY 2003 (August 30, 2002 LTCH IPPS final rule (67 FR 55989 through 55991)). In every final rule in which we have updated the LTC-DRG

[[Page 47333]]

relative weights since the October 1, 2002 implementation of the LTCH PPS (68 FR 45375 through 45385, and 69 FR 48989 through 49000), we reiterated the same steps of our established methodology to determine the annual update to the LTC-DRG relative weights. We continue to believe that this methodology continues to be valid, and we do not find any reason at this time to revise it.

As we explained in the FY 2006 IPPS proposed rule (70 FR 23346), we believe it is appropriate to remove statistical outlier cases and cases with a length of stay of 7 days or less because including those LTCH cases in the calculation of the relative weights could result in an inaccurate relative weight, and therefore an inappropriate payment amount, that does not truly reflect relative resource use among the LTC-DRGs. Specifically, we continue to believe that statistical outlier cases may represent aberrations in the data that distort the measure of average resource use and that, as we explained above, including them in the calculation of the relative weights could result in an inappropriate payment amount.

In the RY 2006 LTCH PPS final rule (70 FR 23346) and as we discussed in greater detail in the FY 2005 IPPS final rule (69 FR 48990), we also explained that, generally, cases with a length of stay 7 days or less are not representative of either typical or perhaps even appropriate LTCH patients. Furthermore, in general, in a hospital established solely to treat very long-stay patients, and with a payment system calibrated to reflect the costs incurred in treating such patients, stays of 7 days or less would not fully receive or benefit from treatment or the range of resource use that is typical in a LTCH stay, and full resources are often not used in the earlier stages of admission to a LTCH. We continue to believe that, if we were to include stays of 7 days or less in the computation of the LTC-DRG relative weights, the value of many relative weights would decrease and, therefore, payments would decrease to a level that may no longer be appropriate. Specifically, because LTCH cases with very short lengths of stay (that is, 7 days or less) do not use the same amount or type of resources as typical LTCH inlier cases (that is, cases in which Medicare covered days exceed five-sixths of the geometric average length of stay for the LTC-DRG) and the patient is discharged prior to receiving a LTCH PPS high-cost outlier payment, our simulations indicate that including these cases would significantly bias payments against LTCH inlier cases to a point where LTCH inlier cases would be underpaid (69 FR 48990). Thus, we do not believe that it would be appropriate to compromise the integrity of the payment determination for those LTCH cases that actually benefit from and receive a full course of treatment at a LTCH, in order to include data from these very short-stays. Consequently, we disagree with the commenters that removing aberrant LTCH cases (that is, statistical outlier cases and cases with a length of stay of 7 days or less) undermines the averaging principle upon which PPSs are developed.

Although we did not propose any change in the methodology for determining the LTC-DRG relative weights for FY 2006, we disagree with the assertions that removing statistical outlier cases and cases with a length of stay of 7 days or less inappropriately narrows the universe of cases used to compute the LTC-DRG relative weights, resulting in a distortion of the principle of averaging. Rather, because each LTC-DRG relative weight represents the average resources required to treat cases in that particular LTC-DRG, relative to the average resources used to treat cases in all LTC-DRGs, we believe that, by removing cases that do not represent the ``average resource use'' of the mix of LTCH cases within a DRG (that is, statistical outlier cases and cases with a length of stay of 7 days or less), for the reasons explained above, we are preserving the integrity of a system that is based on averages. Therefore, in establishing the FY 2006 LTC-DRG relative weights in this final rule, we have continued to remove statistical outlier cases and cases with a length of stay of 7 days or less from the MedPAR data used to compute the FY 2006 LTC-DRG relative weights.

Comment: Four commenters believed the estimated decrease in LTCH PPS payments resulting from the proposed changes to the LTC-DRG relative weights is inconsistent with the statutory mandate that the LTCH PPS be maintained in a budget neutral manner. These commenters recommended that we apply a budget neutrality adjustment to the LTC-DRG relative weights in order to mitigate the estimated LTCH PPS payment reductions that we estimated would result from the proposed changes to the LTC-DRG relative weights for FY 2006. Two of those commenters cited the statutory language authorizing the establishment of the LTCH PPS and argued that the language requires that the LTCH PPS continue to operate under ``budget neutrality.'' They further asserted that, although we did not interpret this language as mandating budget neutrality beyond the initial year of the LTCH PPS, the Secretary should use his or her broad discretionary authority to assure ``the same level of payments projected in the FY 2006 LTCH update regulation'' by making a budget neutrality adjustment in developing the FY 2006 LTC-DRG relative weights.

Response: We understand that these commenters are concerned about the estimated decrease in payments under LTCH PPS based upon changes in the LTC-DRG relative weights for FY 2006. However, we believe that this issue is distinct from the Secretary's budget neutrality obligation under the statute for the first year of implementation of the LTCH PPS. After the first year of the LTCH PPS, the statute gives the Secretary broad authority to determine the appropriateness of system updates and matters such as annual updates and policy changes. As we discussed in the FY 2005 IPPS final rule (69 FR 48999), with respect to budget neutrality, we interpreted section 123(a)(1) of Pub. L. 106-113 to require that total payments under the LTCH PPS during FY 2003 will be projected to equal estimated payments that would have been made for LTCHs' operating and capital-related inpatient hospital costs had the LTCH PPS not have been implemented. Thus we believe the statute's mandate for budget neutrality applies only to the first year of implementation of the LTCH PPS (that is, FY 2003). Consistent with the broad discretional authority conferred upon the Secretary under section 123(a)(1) of Pub. L. 103-116, as amended by section 307 of Pub. L. 106- 554, the Secretary is exercising his broad authority to make updates the LTCH PPS in a nonbudget neutral manner after FY 2003 for various components of the LTCH PPS, including the annual update of the LTC-DRG classifications and relative weights.

Consistent with this budget neutrality requirement for the first year of implementation of the LTCH PPS, under Sec. 412.523(d)(2) of the regulations, an adjustment is made in determining the standard Federal rate for FY 2003 so that aggregate payments under the LTCH PPS are estimated to equal the amount that would have been paid to LTCHs under the reasonable cost-based (TEFRA) payment system if the LTCH PPS were not implemented. Therefore, in the August 30, 2002 LTCH PPS final rule (67 FR 56027 through 56037), which implemented the LTCH PPS, in order to maintain budget neutrality, we adjusted the LTCH PPS Federal rate for FY 2003 so that aggregate payments under the LTCH PPS are estimated to equal the amount that would have been paid to LTCHs under the reasonable cost-based (TEFRA) payment system

[[Page 47334]]

had the LTCH PPS not been implemented.

As we stated in the FY 2005 IPPS final rule (70 FR 48999 through 49000), we continue to believe that section 123 of the Pub. L. 106-113 does not require that the annual update to the LTC-DRG classifications and relative weights maintain budget neutrality. We believe we have satisfied the budget neutrality requirement of section 123 of the Pub. L. 106-113 by establishing the LTCH PPS Federal rate for FY 2003 under Sec. 412.523(d)(2) so that aggregate payment under the LTCH PPS are projected equal to estimated aggregate payments under the reasonable cost-based payment system if the LTCH PPS were not implemented. Therefore, we disagree with the commenters that a budget neutrality adjustment to the LTC-DRG relative weights or to the LTCH PPS Federal rate is required by statute or as a result of the annual update to the LTC-DRGs under Sec. 412.517 for FY 2006.

We agree with the commenters that, under section 123 of the BBRA and section 307 of the BIPA, the Secretary generally has broad authority in developing the LTCH PPS, including whether and how to make adjustments to the LTCH PPS. As we discussed in the RY 2006 LTCH PPS final rule (70 FR 24188), we will consider whether it is appropriate for us to propose a budget neutrality adjustment in the annual update of some aspects of the LTCH PPS under our broad discretionary authority under the statute to provide ``appropriate adjustments'' to the LTCH PPS. As several commenters noted, LTCHs are still transitioning to a PPS and, while coding practices continue to improve, the FY 2004 claims data may ``not yet fully reflect the nature and types of services, staff, and other resources'' that LTCH provide to their patients. In the RY 2005 LTCH PPS final rule, we indicated that, until the 5-year transition from reasonable cost-based reimbursement to prospective payment is complete, we believe it may not be appropriate to update any aspects of the LTCH PPS in a budget neutral manner. As noted above, the most recent available LTCH PPS claims data are from discharges occurring during FY 2004. These LTCH claims data are from the second year of the LTCH PPS (FY 2004), which is the only first full year since the LTCH PPS was implemented for cost reporting periods beginning on or after October 1, 2002 (FY 2003). Because it is still early in the 5- year LTCH PPS transition period, we continue to believe that it is inappropriate to update any aspects of the LTCH PPS in a budget neutral manner. A primary reason for waiting until after the transition is complete before evaluating aspects of the LTCH PPS, including the budget neutrality issue, is that the data available to analyze such issues are very limited because the LTCH PPS is still relatively new and there is a lag time in data availability. As several commenters pointed out, the FY 2004 MedPAR data are the first full year of LTCH PPS data since the LTCH PPS was implemented for cost reporting periods beginning on or after October 1, 2002 (FY 2003). In addition, the fact that a number of LTCHs were and some are still transitioning to 100 percent of the Federal prospective payment rate may make the available data on which to base a budget neutrality adjustment even less appropriate because LTCHs may still be modifying their behavior based on their transition to prospective payment and, therefore, our data may not yet fully reflect any operational changes LTCHs may have made in response to prospective payment. We continue to believe that, once we have progressed further through the 5-year transition period, we will have a better opportunity to evaluate the impacts of the implementation of this new payment system based on a number of years of LTCH PPS data, which will most appropriately reflect LTCHs' experience under a PPS.

For the reasons stated above, we do not believe that a budget neutrality adjustment to the FY 2006 LTC-DRG relative weights or to the LTCH PPS Federal rate is necessary or appropriate. Accordingly, in developing the FY 2006 LTC-DRGs and relative weights shown in Table 11 of the Addendum of this final rule, we have not applied an adjustment for budget neutrality nor are we adjusting the 2006 LTCH PPS rate year Federal rate established in the 2006 LTCH PPS final rule (70 FR 24180) to account for the estimated change in LTCH PPS payments that will result from the annual update to the LTC-DRG classifications and relative weights for FY 2006.

Comment: Several commenters recommended implementing a ``dampening policy,'' similar to that which was implemented for the Ambulatory Payment Classification (APC) changes under the Hospital Outpatient PPS (OPPS) in CY 2003, which would reduce the decrease in any relative weight in excess of a threshold (for example, 15 percent) by half, to mitigate instability in LTCH PPS payments because of the ``significant/ substantial'' decrease in many of the relative weights.

Response: A ``dampening policy,'' as recommended by the commenters, would limit the decrease in any of the LTC-DRG relative weights to a maximum amount, which would reduce the estimated decrease in LTCH PPS payments that we projected in the FY 2006 IPPS proposed rule as a result of the proposed changes to the LTC-DRG relative weights for FY 2006 (70 FR 23667). The commenters believed that the estimated decrease in the LTCH PPS payments resulting from the proposed changes to the LTC-DRGs for FY 2006 would create a ``destabilizing effect'' on LTCH PPS payments. For the reasons discussed below, we do not believe the estimated decrease in LTCH PPS payments resulting from the changes we are making to the LTC-DRG relative weights for FY 2006 in this final rule will lead to instability in LTCH PPS payments, and therefore, we are not implementing a ``dampening policy,'' as recommended by the commenters.

As discussed in the November 1, 2002 OPPS final rule (67 FR 66749 through 66750), we believed it was appropriate to implement the ``dampening policy'' under the OPPS referenced by the commenters because many of the decreases in payment rates for some of the APCs appeared to be linked to ``changes in the methodology for those drugs and devices that will no longer be eligible for pass-through payments; miscoding; restructuring of APCs (in which movement of a single code from one APC to another may change the median cost of both APCs), or use of data from the period following the implementation of the OPPS.'' Although Medicare payment for both hospital outpatient services and inpatient LTCH services are reimbursed under a PPS (respectively), there are significant distinctions between the two payment systems. For instance, under the LTCH PPS, a single per LTC-DRG payment is made for all inpatient hospital services provided to a patient for each stay, where in contrast, under the OPPS, payments based on APCs may include distinct payment methodologies for certain drugs and devices that are eligible for pass-through payments. Thus, there are significant distinctions between the two payment systems that warrant different considerations when evaluating the need for a ``dampening policy.'' Below we discuss the reasons we believe that a ``dampening policy'' to mitigate the effects of the changes in the LTC-DRG relative weights for FY 2006 on LTCH PPS payments are not necessary or appropriate.

As noted by the commenters, many of the proposed FY 2006 LTC-DRG relative weights decreased in

[[Page 47335]]

comparison to the FY 2005 LTC-DRG relative weights, which would result in an aggregate estimated decrease in FY 2006 LTCH PPS payments. As we explained in the FY 2006 IPPS proposed rule (70 FR 23667), we continue to observe an increase of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year. The addition of these lower charge cases results in a decrease in the many of the LTC-DRG relative weights from FY 2005 to FY 2006. This decrease in many of the LTC-DRG relative weights, in turn, will result in an estimated decrease in LTCH PPS payments. As we explained in that same proposed rule, contributing to this increased number of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year are improvements in coding practices, which are typically found when moving from a reasonable cost-based payment system to a PPS. A further analysis of the LTCH claims in the March 2005 update of the FY 2004 MedPAR data, which we used to determine the FY 2006 LTC-DRG relative weights in this final rule, continue to show an increase of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year. As we explained the FY 2006 IPPS proposed rule (70 FR 23667), the impact of including cases with relatively lower charges into LTC-DRGs that had a relatively higher relative weight in the version 22.0 (FY 2005) GROUPER is a decrease in the average relative weight for those LTC-DRGs, which, in turn, results in an estimated aggregate decrease in LTCH PPS payments.

A few commenters acknowledged that with the move from cost-based reimbursement to a PPS, LTCHs' coding practices are still undergoing refinement. Specifically, two commenters stated that ``the LTCH PPS, in its third year of implementation, is still in transition; the initial 5-year phase-in will end September 2006. During this time of transition, LTCH coding and data are still undergoing improvement.'' Therefore, it is not unreasonable to observe relatively significant changes (either higher or lower) in the average charge for many LTC- DRGs as LTCHs' behavior coding continues to change in response to the implementation of a PPS. As the transition progresses, we expect that LTCH's behavior will result in fewer nonuniform changes in the average charge of many LTC-DRGs, which may impact the LTC-DRG relative weights from year to year.

As we discussed above, we believe that there are no systemic errors in the LTCH FY 2004 MedPAR data, and we believe that the increase of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights that we observed in the FY 2004 LTCH claims data (which results in a decrease in the many of the LTC-DRG relative weights) accurately represents current LTCH costs. Specifically, an analysis of a comparison of the FY 2003 LTCH claims data (used to develop the FY 2005 LTC-DRG relative weights) and the FY 2004 LTCH claims data (used to develop the FY 2006 LTC-DRG relative weights) shows that, of the 155 LTC-DRGs that are used on a ``regular basis'' (that is, nationally, LTCHs discharge, in total, 25 or more of these cases annually), about 30 percent of those LTC-DRGs have experienced a decrease in the average charge per case, which generally results in a lower relative weight. In addition, about 45 percent of those LTC-DRGs have experienced an increase in the average charge that is less than the increase (16 percent) in the overall average charge across all LTC- DRGs. In general, the LTC-DRG relative weights are determined by dividing the average charge for each LTC-DRG by the average charge across all LTC-DRGs. Accordingly, those LTC-DRGs with an increase in average charge of less than 16 percent (that is, the increase in average charge across all LTC-DRGs) will also experience a reduction in their relative weight because the average charge for each of those LTC- DRGs is being divided by a bigger number (that is, the average charge across all LTC-DRGs). Therefore, because we believe the FY 2004 LTCH claims data used to determine the FY 2006 LTC-DRG relative weights accurately reflect the resources used by LTCHs to treat their patients, and these data show either a decrease in the average charge of the LTC- DRG or an increase in the average charge of the LTC-DRG that is less than the overall increase in the average charge across all LTC-DRGs, we believe that the decrease in many of the LTC-DRG relative weights is appropriate.

The LTC-DRG relative weights are designed to reflect the average of resources used to treat representative cases of the discharges within each LTC-DRG. As we discussed in greater detail above, after our extensive analysis of the FY 2004 MedPAR data, which we used to determine the FY 2006 LTC-DRG relative weights, we concluded that there are no systematic errors in that data. Therefore, we continue to believe it is appropriate to base the FY 2006 LTC-DRG relative weights on LTCH claims data in the FY 2004 MedPAR file. Furthermore, we believe that the decrease in many of the LTC-DRG relative weights is appropriate and is reflective of the changing behaviors of LTCHs' response to a PPS environment. As we discussed above, we believe that the LTCH claims data in the FY 2004 MedPAR file accurately reflects the resources that are expended to treat LTCH patients in each LTC-DRG. Although many of the LTC-DRG relative weights (and consequently aggregate LTCH PPS payments, excluding the update to the LTCH PPS Federal rate effective July 1, 2005 (70 FR 24217) will be lower in FY 2006 as compared to FY 2005, we do not believe that the payment rates for those LTC-DRGs are inappropriate based on the LTCH claims data in the FY 2004 MedPAR files. Rather, we believe that the lower LTC-DRG relative weights (and consequently a reduction in aggregate LTCH PPS payments) are appropriate, given that the average resources used to treat a LTCH patient in a particular LTC-DRG are less than the average resources used to treat a LTCH patient in a particular LTC-DRG based on FY 2003 LTCH claims data. Therefore, we do not agree with the commenters' assertion that the changes to the LTC-DRG relative weights for FY 2006 will result in instability in LTCH PPS payments. Rather, we believe that the changes to the LTC-DRG relative weights for FY 2006 will result in appropriate payments for the resources used to treat LTCH patients in a particular LTC-DRG. Accordingly, for the reasons discussed above, we are not implementing a ``dampening policy'' in determining the FY 2006 LTC-DRG relative weights in this final rule. We also note that the 4.2 percent decrease in LTCH PPS payments estimated as a result of the changes we are making to the LTC-DRGs and relative weights in this final rule for FY 2006 (see section VII. of the Addendum to this final rule) is partially offset by the projected 5.7 percent increase in LTCH PPS payments estimated based on the updated rates and factors effective for discharges occurring on or after July 1, 2005 established in the FY 2006 LTCH PPS final rule (70 FR 24217).

Below we discuss in detail the steps for calculating the FY 2006 LTC-DRG relative weights as presented in the FY 2006 IPPS proposed rule (70 FR 23346 through 23353). We note that, as we stated above in section II.D.3.b. of this preamble, as we proposed, we have excluded the data of all-inclusive rate LTCHs and LTCHs that are paid in accordance with demonstration projects

[[Page 47336]]

that had claims in the FY 2004 MedPAR file.

Step 1--Remove statistical outliers.

The first step in the calculation of the FY 2006 LTC-DRG relative weights is to remove statistical outlier cases. We define statistical outliers as cases that are outside of 3.0 standard deviations from the mean of the log distribution of both charges per case and the charges per day for each LTC-DRG. These statistical outliers are removed prior to calculating the relative weights. As noted above, we believe that they may represent aberrations in the data that distort the measure of average resource use. Including those LTCH cases in the calculation of the relative weights could result in an inaccurate relative weight that does not truly reflect relative resource use among the LTC-DRGs.

Step 2--Remove cases with a length of stay of 7 days or less.

The FY 2006 LTC-DRG relative weights reflect the average of resources used on representative cases of a specific type. Generally, cases with a length of stay 7 days or less do not belong in a LTCH because these stays do not fully receive or benefit from treatment that is typical in a LTCH stay, and full resources are often not used in the earlier stages of admission to a LTCH. As explained above, if we were to include stays of 7 days or less in the computation of the FY 2006 LTC-DRG relative weights, the value of many relative weights would decrease and, therefore, payments would decrease to a level that may no longer be appropriate.

We do not believe that it would be appropriate to compromise the integrity of the payment determination for those LTCH cases that actually benefit from and receive a full course of treatment at a LTCH, in order to include data from these very short-stays. Thus, as explained above, in determining the FY 2006 LTC-DRG relative weights, we remove LTCH cases with a length of stay of 7 days or less.

Step 3--Adjust charges for the effects of short-stay outliers.

After removing cases with a length of stay of 7 days or less, we are left with cases that have a length of stay of greater than or equal to 8 days. The next step in the calculation of the FY 2006 LTC-DRG relative weights is to adjust each LTCH's charges per discharge for those remaining cases for the effects of short-stay outliers as defined in Sec. 412.529(a). (However, we note that even if a case was removed in Step 2 (that is, cases with a length of stay of 7 days or less), it was paid as a short-stay outlier if its length of stay was less than or equal to five-sixths of the average length of stay of the LTC-DRG, in accordance with Sec. 412.529.)

We make this adjustment by counting a short-stay outlier as a fraction of a discharge based on the ratio of the length of stay of the case to the average length of stay for the LTC-DRG for nonshort-stay outlier cases. This has the effect of proportionately reducing the impact of the lower charges for the short-stay outlier cases in calculating the average charge for the LTC-DRG. This process produces the same result as if the actual charges per discharge of a short-stay outlier case were adjusted to what they would have been had the patient's length of stay been equal to the average length of stay of the LTC-DRG.

As we explained in the FY 2006 IPPS proposed rule (70 FR 23346 through 23347), counting short-stay outlier cases as full discharges with no adjustment in determining the LTC-DRG relative weights would lower the LTC-DRG relative weight for affected LTC-DRGs because the relatively lower charges of the short-stay outlier cases would bring down the average charge for all cases within a LTC-DRG. This would result in an ``underpayment'' to nonshort-stay outlier cases and an ``overpayment'' to short-stay outlier cases. Therefore, in this final rule, we adjust for short-stay outlier cases under Sec. 412.529 in this manner because it results in more appropriate payments for all LTCH cases.

Step 4--Calculate the FY 2006 LTC-DRG relative weights on an iterative basis.

The process of calculating the LTC-DRG relative weights using the hospital-specific relative value methodology is iterative. First, for each LTCH case, we calculate a hospital-specific relative charge value by dividing the short-stay outlier adjusted charge per discharge (see step 3) of the LTCH case (after removing the statistical outliers (see step 1)) and LTCH cases with a length of stay of 7 days or less (see step 2) by the average charge per discharge for the LTCH in which the case occurred. The resulting ratio is then multiplied by the LTCH's case-mix index to produce an adjusted hospital-specific relative charge value for the case. An initial case-mix index value of 1.0 is used for each LTCH.

For each LTC-DRG, the FY 2006 LTC-DRG relative weight is calculated by dividing the average of the adjusted hospital-specific relative charge values (from above) for the LTC-DRG by the overall average hospital-specific relative charge value across all cases for all LTCHs. Using these recalculated LTC-DRG relative weights, each LTCH's average relative weight for all of its cases (case-mix) is calculated by dividing the sum of all the LTCH's LTC-DRG relative weights by its total number of cases. The LTCHs' hospital-specific relative charge values above are multiplied by these hospital-specific case-mix indexes. These hospital-specific case-mix adjusted relative charge values are then used to calculate a new set of LTC-DRG relative weights across all LTCHs. In this final rule, this iterative process is continued until there is convergence between the weights produced at adjacent steps, for example, when the maximum difference is less than 0.0001.

Step 5--Adjust the FY 2006 LTC-DRG relative weights to account for nonmonotonically increasing relative weights.

As explained in section II.B. of this preamble, the FY 2006 CMS DRGs, on which the FY 2006 LTC-DRGs are based, contain ``pairs'' that are differentiated based on the presence or absence of CCs. The LTC- DRGs with CCs are defined by certain secondary diagnoses not related to or inherently a part of the disease process identified by the principal diagnosis, but the presence of additional diagnoses does not automatically generate a CC. As we discussed in the FY 2005 IPPS final rule (69 FR 48991), the value of monotonically increasing relative weights rises as the resource use increases (for example, from uncomplicated to more complicated). The presence of CCs in a LTC-DRG means that cases classified into a ``without CC'' LTC-DRG are expected to have lower resource use (and lower costs). In other words, resource use (and costs) are expected to decrease across ``with CC/without CC'' pairs of LTC-DRGs.

For a case to be assigned to a LTC-DRG with CCs, more coded information is called for (that is, at least one relevant secondary diagnosis), than for a case to be assigned to a LTC-DRG ``without CCs'' (which is based on only one principal diagnosis and no relevant secondary diagnoses). Currently, the LTCH claims data include both accurately coded cases without complications and cases that have complications (and cost more), but were not coded completely. Both types of cases are grouped to a LTC-DRG ``without CCs'' when only the principal diagnosis was coded. Since the LTCH PPS was only implemented for cost reporting periods beginning on or after October 1, 2002 (FY 2003), and LTCHs were previously paid under cost-based reimbursement, which is not based on patient diagnoses, coding by LTCHs for

[[Page 47337]]

these cases may not have been as detailed as possible.

Thus, in developing the FY 2003 LTC-DRG relative weights for the LTCH PPS based on FY 2001 claims data, as we discussed in the August 30, 2002 LTCH PPS final rule (67 FR 55990), we found on occasion that the data suggested that cases classified to the LTC-DRG ``with CCs'' of a ``with CC''/``without CC'' pair had a lower average charge than the corresponding LTC-DRG ``without CCs.'' Similarly, as discussed in the FY 2005 IPPS final rule (69 FR 48991 through 48992), based on FY 2003 claims data, we also found on occasion that the data suggested that cases classified to the LTC-DRG ``with CCs'' of a ``with CC''/``without CC'' pair have a lower average charge than the corresponding LTC-DRG ``without CCs'' for the FY 2005 LTC-DRG relative weights.

We believe this anomaly may be due to coding that may not have fully reflected all comorbidities that were present. Specifically, LTCHs may have failed to code relevant secondary diagnoses, which resulted in cases that actually had CCs being classified into a ``without CC'' LTC-DRG. It would not be appropriate to pay a lower amount for the ``with CC'' LTC-DRG because, in general, cases classified into a ``with CC'' LTC-DRG are expected to have higher resource use (and higher cost) as discussed above. Therefore, previously when we determined the LTC-DRG relative weights in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55990), we grouped both the cases ``with CCs'' and ``without CCs'' together for the purpose of calculating the LTC-DRG relative weights since the implementation of the LTCH PPS in FY 2003. As we stated in that same final rule, we will continue to employ this methodology to account for nonmonotonically increasing relative weights until we have adequate data to calculate appropriate separate weights for these anomalous LTC-DRG pairs. We expect that, as was the case when we first implemented the IPPS, this problem will be self-correcting, as LTCHs submit more completely coded data in the future.

There are three types of ``with CC'' and ``without CC'' pairs that could be nonmonotonic; that is, where the ``without CC'' LTC-DRG would have a higher average charge than the ``with CC'' LTC-DRG. For this final rule, using the LTCH cases in the March 2005 update of the FY 2004 MedPAR file (the best available data at this time), we identified three types of nonmonotonic LTC-DRG pairs. As we stated in the August 30, 2002 LTCH PPS final rule (67 FR 55990), we believe this anomaly may be due to coding inaccuracies and expect that, as was the case when we first implemented the acute care hospital IPPS, this problem will be self-correcting, as LTCHs submit more completely coded data in the future.

The first category of nonmonotonically increasing relative weights for FY 2006 LTC-DRG pairs ``with and without CCs'' contains one pair of LTC-DRGs in which both the LTC-DRG ``with CCs'' and the LTC-DRG ``without CCs'' had 25 or more LTCH cases and, therefore, did not fall into one of the 5 low-volume quintiles. For those nonmonotonic LTC-DRG pairs, we combine the LTCH cases and compute a new relative weight based on the case-weighted average of the combined LTCH cases of the LTC-DRGs. The case-weighted average charge is determined by dividing the total charges for all LTCH cases by the total number of LTCH cases for the combined LTC-DRG. This new relative weight is then assigned to both of the LTC-DRGs in the pair. In this final rule, for FY 2006, LTC- DRGs 553 and 554 fall into this category.

The second category of nonmonotonically increasing relative weights for LTC-DRG pairs ``with and without CCs'' consists of one pair of LTC- DRGs that has fewer than 25 cases, and each LTC-DRG is grouped to different low-volume quintiles in which the ``without CC'' LTC-DRG is in a higher-weighted low-volume quintile than the ``with CC'' LTC-DRG. For those pairs, we combine the LTCH cases and determine the case- weighted average charge for all LTCH cases. The case-weighted average charge is determined by dividing the total charges for all LTCH cases by the total number of LTCH cases for the combined LTC-DRG. Based on the case-weighted average LTCH charge, we determine within which low- volume quintile the ``combined LTC-DRG'' is grouped. Both LTC-DRGs in the pair are then grouped into the same low-volume quintile, and thus have the same relative weight. In this final rule, for FY 2006, LTC- DRGs 555, 556 and 557 fall into this category. (We note, 3 LTC-DRGs make up this non-monotonic ``pair'' of LTC-DRGs because these percutaneous cardiovascular procedure DRGs are further split depending on the presence or absence of a drug eluting stint and the presence or absence of a major ``CV'' (cardiovascular) diagnosis, which is similar to the adjustment for non-monotonicity for DRGs 521, 522 and 523 in the development of the FY 2005 LTC-DRG relative weights (69 FR 78922).

The third category of nonmonotonically increasing relative weights for LTC-DRG pairs ``with and without CCs'' consists of one pair of LTC- DRGs where one of the LTC-DRGs has fewer than 25 LTCH cases and is grouped to a low-volume quintile and the other LTC-DRG has 25 or more LTCH cases and has its own LTC-DRG relative weight, and the LTC-DRG ``without CCs'' has the higher relative weight. We removed the low- volume LTC-DRG from the low-volume quintile and combined it with the other LTC-DRG for the computation of a new relative weight for each of these LTC-DRGs. This new relative weight is assigned to both LTC-DRGs, so they each have the same relative weight. In this final rule, for FY 2006, LTC-DRGs 142 and 143 fall into this category.

Step 6--Determine a FY 2006 LTC-DRG relative weight for LTC-DRGs with no LTCH cases.

As we stated above, we determine the relative weight for each LTC- DRG using charges reported in the March 2005 update of the FY 2004 MedPAR file. Of the 526 LTC-DRGs for FY 2006, we identified 196 LTC- DRGs for which there were no LTCH cases in the database. That is, based on data from the FY 2004 MedPAR file used in this final rule, no patients who would have been classified to those LTC-DRGs were treated in LTCHs during FY 2004 and, therefore, no charge data were reported for those LTC-DRGs. Thus, in the process of determining the LTC-DRG relative weights, we are unable to determine weights for these 196 LTC- DRGs using the methodology described in steps 1 through 5 above. However, because patients with a number of the diagnoses under these LTC-DRGs may be treated at LTCHs beginning in FY 2006, we assign relative weights to each of the 196 ``no volume'' LTC-DRGs based on clinical similarity and relative costliness to one of the remaining 330 (526 - 196 = 330) LTC-DRGs for which we are able to determine relative weights, based on FY 2004 claims data.

As there are currently no LTCH cases in these ``no volume'' LTC- DRGs, we determined relative weights for the 196 LTC-DRGs with no LTCH cases in the FY 2004 MedPAR file used in this final rule by grouping them to the appropriate low-volume quintile. This methodology is consistent with our methodology used in determining relative weights to account for the low-volume LTC-DRGs described above.

Our methodology for determining relative weights for the ``no volume'' LTC-DRGs is as follows: We crosswalk

[[Page 47338]]

the no volume LTC-DRGs by matching them to other similar LTC-DRGs for which there were LTCH cases in the FY 2004 MedPAR file based on clinical similarity and intensity of use of resources as determined by care provided during the period of time surrounding surgery, surgical approach (if applicable), length of time of surgical procedure, post- operative care, and length of stay. We assign the relative weight for the applicable low-volume quintile to the no volume LTC-DRG if the LTC- DRG to which it is crosswalked is grouped to one of the low-volume quintiles. If the LTC-DRG to which the no volume LTC-DRG is crosswalked is not one of the LTC-DRGs to be grouped to one of the low-volume quintiles, we compare the relative weight of the LTC-DRG to which the no volume LTC-DRG is crosswalked to the relative weights of each of the five quintiles and we assign the no volume LTC-DRG the relative weight of the low-volume quintile with the closest weight. For this final rule, a list of the no volume FY 2006 LTC-DRGs and the FY 2006 LTC-DRG to which it is crosswalked in order to determine the appropriate low- volume quintile for the assignment of a relative weight for FY 2006 is shown in the chart below.

No Volume LTC-DRG Crosswalk and Quintile Assignment for FY 2006

Low-volume LTC-DRG

Description Cross-walked quintile LTC-DRG

assignment

2............... CRANIOTOMY AGE >17 W/

1 Quintile 5. O CC. 3............... CRANIOTOMY AGE 0-17.

1 Quintile 5. 6............... CARPAL TUNNEL

251 Quintile 1. RELEASE. 26.............. SEIZURE & HEADACHE

25 Quintile 1. AGE 0-17. 30.............. TRAUMATIC STUPOR &

29 Quintile 2. COMA, COMA 17 W/

25 Quintile 1. O CC. 33.............. CONCUSSION AGE 0-17.

25 Quintile 1. 36.............. RETINAL PROCEDURES..

40 Quintile 4. 37.............. ORBITAL PROCEDURES..

40 Quintile 4. 38.............. PRIMARY IRIS

40 Quintile 4. PROCEDURES. 39.............. LENS PROCEDURES WITH

40 Quintile 4. OR WITHOUT VITRECTOMY. 41.............. EXTRAOCULAR

40 Quintile 4. PROCEDURES EXCEPT ORBIT AGE 0-17. 42.............. INTRAOCULAR

40 Quintile 4. PROCEDURES EXCEPT RETINA, IRIS & LENS. 43.............. HYPHEMA.............

40 Quintile 4. 45.............. NEUROLOGICAL EYE

40 Quintile 4. DISORDERS. 47.............. OTHER DISORDERS OF

40 Quintile 4. THE EYE AGE >17 W/O CC. 48.............. OTHER DISORDERS OF

40 Quintile 4. THE EYE AGE 0-17. 49.............. ESOPHAGITIS,

64 Quintile 4. GASTROENT & MISC DIGEST DISORDERS AGE 0-17. 50.............. SIALOADENECTOMY.....

63 Quintile 4. 51.............. SALIVARY GLAND

63 Quintile 4. PROCEDURES EXCEPT SIALOADENECTOMY. 52.............. CLEFT LIP & PALATE

63 Quintile 4. REPAIR. 53.............. SINUS & MASTOID

63 Quintile 4. PROCEDURES AGE >17. 54.............. SINUS & MASTOID

63 Quintile 4. PROCEDURES AGE 0-17. 55.............. MISCELLANEOUS EAR,

63 Quintile 4. NOSE, MOUTH & THROAT PROCEDURES. 56.............. RHINOPLASTY.........

63 Quintile 4. 57.............. T&A PROC, EXCEPT

69 Quintile 1. TONSILLECTOMY &/OR ADENOIDECTOMY ONLY, AGE >17. 58.............. T&A PROC, EXCEPT

69 Quintile 1. TONSILLECTOMY &/OR ADENOIDECTOMY ONLY, AGE 0-17. 59.............. TONSILLECTOMY &/OR

69 Quintile 1. ADENOIDECTOMY ONLY, AGE >17. 60.............. TONSILLECTOMY &/OR

69 Quintile 1. ADENOIDECTOMY ONLY, AGE 0-17. 62.............. MYRINGOTOMY W TUBE

69 Quintile 1. INSERTION AGE 0-17. 66.............. EPISTAXIS...........

69 Quintile 1. 70.............. OTITIS MEDIA & URI

69 Quintile 1. AGE 0-17. 71.............. LARYNGOTRACHEITIS...

97 Quintile 2. 72.............. OTHER DIGESTIVE

73 Quintile 3. SYSTEM DIAGNOSES AGE 0-17. 74.............. OTHER EAR, NOSE,

69 Quintile 1. MOUTH & THROAT DIAGNOSES AGE 0-17. 81.............. RESPIRATORY

69 Quintile 1. INFECTIONS & INFLAMMATIONS AGE 0- 17. 84.............. MAJOR CHEST TRAUMA W/

93 Quintile 2. O CC. 91.............. ALCOHOL/DRUG ABUSE

90 Quintile 1. OR DEPENDENCE W REHABILITATION THERAPY W/O CC. 98.............. BRONCHITIS & ASTHMA

97 Quintile 2. AGE 0-17. 104............. CARDIAC VALVE &

110 Quintile 3. OTHER MAJOR CARDIOTHORACIC PROC W CARDIAC CATH. 105............. CARDIAC VALVE &

110 Quintile 3. OTHER MAJOR CARDIOTHORACIC PROC W/O CARDIAC CATH. 106............. CORONARY BYPASS W

110 Quintile 3. PTCA. 107............. CORONARY BYPASS W

110 Quintile 3. CARDIAC CATH. 108............. OTHER CARDIOTHORACIC

110 Quintile 3. PROCEDURES. 109............. CORONARY BYPASS W/O

110 Quintile 3. PTCA OR CARDIAC CATH. 111............. MAJOR CARDIOVASCULAR

110 Quintile 3. PROCEDURES W/O CC. 129............. CARDIAC ARREST,

110 Quintile 3. UNEXPLAINED. 137............. CARDIAC CONGENITAL &

136 Quintile 2. VALVULAR DISORDERS AGE 0-17. 146............. ALCOHOL/DRUG ABUSE

148 Quintile 5. OR DEPENDENCE W/O REHABILITATION THERAPY W/O CC. 147............. NASAL TRAUMA &

148 Quintile 5. DEFORMITY. 149............. FX, SPRN, STRN &

176 Quintile 3. DISL OF FOREARM, HAND, FOOT AGE 0-17. 153............. MINOR SMALL & LARGE

152 Quintile 3. BOWEL PROCEDURES W/ O CC. 155............. STOMACH, ESOPHAGEAL

154 Quintile 5. & DUODENAL PROCEDURES AGE >17 W/O CC. 156............. STOMACH, ESOPHAGEAL

154 Quintile 5. & DUODENAL PROCEDURES AGE 0-17. 158............. ANAL & STOMAL

157 Quintile 4. PROCEDURES W/O CC. 159............. HERNIA PROCEDURES

177 Quintile 3. EXCEPT INGUINAL & FEMORAL AGE >17 W CC. 160............. HERNIA PROCEDURES

177 Quintile 3. EXCEPT INGUINAL & FEMORAL AGE >17 W/O CC.

[[Page 47339]]

162............. INGUINAL & FEMORAL

178 Quintile 3. HERNIA PROCEDURES AGE >17 W/O CC. 163............. HERNIA PROCEDURES

178 Quintile 3. AGE 0-17. 164............. NUTRITIONAL & MISC

148 Quintile 5. METABOLIC DISORDERS AGE 0-17. 165............. CESAREAN SECTION W/O

148 Quintile 5. CC. 166............. LYMPHOMA & NON-ACUTE

148 Quintile 5. LEUKEMIA W OTHER O.R. PROC W/O CC. 167............. MAJOR SMALL & LARGE

148 Quintile 5. BOWEL PROCEDURES W/ O CC. 169............. MOUTH PROCEDURES W/O

185 Quintile 3. CC. 184............. FX, SPRN, STRN &

183 Quintile 1. DISL OF UPARM,LOWLEG EX FOOT AGE 0-17. 186............. DENTAL & ORAL DIS

185 Quintile 3. EXCEPT EXTRACTIONS & RESTORATIONS, AGE 0-17. 187............. DENTAL EXTRACTIONS &

185 Quintile 3. RESTORATIONS. 190............. PERIANAL & PILONIDAL

189 Quintile 1. PROCEDURES. 192............. PANCREAS, LIVER &

191 Quintile 4. SHUNT PROCEDURES W/ O CC. 194............. BILIARY TRACT PROC

193 Quintile 3. EXCEPT ONLY CHOLECYST W OR W/O C.D.E. W/O CC. 196............. CHOLECYSTECTOMY W

197 Quintile 3. C.D.E. W/O CC. 198............. CHOLECYSTECTOMY

197 Quintile 3. EXCEPT BY LAPAROSCOPE W/O C.D.E. W/O CC. 199............. HEPATOBILIARY

200 Quintile 5. DIAGNOSTIC PROCEDURE FOR MALIGNANCY. 212............. HIP & FEMUR

210 Quintile 5. PROCEDURES EXCEPT MAJOR JOINT AGE 0- 17. 220............. LOWER EXTREM & HUMER

218 Quintile 5. PROC EXCEPT HIP,FOOT,FEMUR AGE 0-17. 224............. SHOULDER,ELBOW OR

227 Quintile 3. FOREARM PROC,EXC MAJOR JOINT PROC, W/ O CC. 229............. HAND OR WRIST PROC,

237 Quintile 1. EXCEPT MAJOR JOINT PROC, W/O CC. 232............. ARTHROSCOPY.........

237 Quintile 1. 234............. OTHER MUSCULOSKELET

237 Quintile 1. SYS & CONN TISS O.R. PROC W/O CC. 252............. SEPTICEMIA AGE 0-17.

253 Quintile 3. 255............. LIMB REATTACHMENT,

253 Quintile 3. HIP AND FEMUR PROC FOR MULTIPLE SIGNIFICANT TR. 257............. TOTAL MASTECTOMY FOR

274 Quintile 3. MALIGNANCY W CC. 258............. TOTAL MASTECTOMY FOR

274 Quintile 3. MALIGNANCY W/O CC. 260............. SUBTOTAL MASTECTOMY

274 Quintile 3. FOR MALIGNANCY W/O CC. 261............. BREAST PROC FOR NON-

274 Quintile 3. MALIGNANCY EXCEPT BIOPSY & LOCAL EXCISION. 267............. MAJOR HEAD & NECK

271 Quintile 3. PROCEDURES. 275............. MALIGNANT BREAST

274 Quintile 3. DISORDERS W/O CC. 279............. CELLULITIS AGE 0-17.

273 Quintile 1. 282............. TRAUMA TO THE SKIN,

281 Quintile 1. SUBCUT TISS & BREAST AGE 0-17. 286............. EXTREME IMMATURITY..

292 Quintile 5. 289............. PARATHYROID

63 Quintile 4. PROCEDURES. 291............. THYROGLOSSAL

63 Quintile 4. PROCEDURES. 298............. PREMATURITY W MAJOR

297 Quintile 2. PROBLEMS. 307............. PROSTATECTOMY W/O CC

306 Quintile 2. 309............. MINOR BLADDER

308 Quintile 3. PROCEDURES W/O CC. 311............. TRANSURETHRAL

310 Quintile 4. PROCEDURES W/O CC. 313............. URETHRAL PROCEDURES,

312 Quintile 1. AGE >17 W/O CC. 314............. URETHRAL PROCEDURES,

305 Quintile 1. AGE 0-17. 322............. FULL TERM NEONATE W

321 Quintile 1. MAJOR PROBLEMS. 324............. NEONATE W OTHER

321 Quintile 1. SIGNIFICANT PROBLEMS. 326............. RECTAL RESECTION W/O

321 Quintile 1. CC. 327............. RECTAL RESECTION W

321 Quintile 1. CC. 329............. URETHRAL STRICTURE

305 Quintile 1. AGE >17 W/O CC. 330............. URETHRAL STRICTURE

305 Quintile 1. AGE 0-17. 333............. OTHER KIDNEY &

332 Quintile 2. URINARY TRACT DIAGNOSES AGE 0-17. 335............. MAJOR MALE PELVIC

345 Quintile 5. PROCEDURES W/O CC. 337............. TRANSURETHRAL

306 Quintile 2. PROSTATECTOMY W/O CC. 338............. TESTES PROCEDURES,

336 Quintile 2. FOR MALIGNANCY. 340............. TESTES PROCEDURES,

339 Quintile 4. NON-MALIGNANCY AGE 0-17. 342............. CIRCUMCISION AGE >17

339 Quintile 4. 343............. CIRCUMCISION AGE 0-

339 Quintile 4. 17. 349............. BENIGN PROSTATIC

339 Quintile 4. HYPERTROPHY W/O CC. 351............. STERILIZATION, MALE.

339 Quintile 4. 353............. PELVIC EVISCERATION,

339 Quintile 4. RADICAL HYSTERECTOMY & RADICAL VULVECTOMY. 354............. UTERINE,ADNEXA PROC

339 Quintile 4. FOR NON-OVARIAN/ ADNEXAL MALIG W CC. 355............. UTERINE,ADNEXA PROC

339 Quintile 4. FOR NON-OVARIAN/ ADNEXAL MALIG W/O CC. 356............. FEMALE REPRODUCTIVE

339 Quintile 4. SYSTEM RECONSTRUCTIVE PROCEDURES. 357............. UTERINE & ADNEXA

339 Quintile 4. PROC FOR OVARIAN OR ADNEXAL MALIGNANCY. 358............. UTERINE & ADNEXA

339 Quintile 4. PROC FOR NON- MALIGNANCY W CC. 359............. UTERINE & ADNEXA

339 Quintile 4. PROC FOR NON- MALIGNANCY W/O CC. 361............. LAPAROSCOPY &

110 Quintile 3. INCISIONAL TUBAL INTERRUPTION. 362............. ENDOSCOPIC TUBAL

110 Quintile 3. INTERRUPTION. 363............. D&C, CONIZATION &

110 Quintile 3. RADIO-IMPLANT, FOR MALIGNANCY. 367............. MALIGNANCY, FEMALE

110 Quintile 3. REPRODUCTIVE SYSTEM W/O CC. 370............. CESAREAN SECTION W

369 Quintile 3. CC. 371............. APPENDECTOMY W

368 Quintile 2. COMPLICATED PRINCIPAL DIAG W CC. 372............. VAGINAL DELIVERY W

110 Quintile 3. COMPLICATING DIAGNOSES.

[[Page 47340]]

373............. VAGINAL DELIVERY W/O

110 Quintile 3. COMPLICATING DIAGNOSES. 374............. VAGINAL DELIVERY W

110 Quintile 3. STERILIZATION &/OR D&C. 375............. VAGINAL DELIVERY W

110 Quintile 3. O.R. PROC EXCEPT STERIL &/OR D&C. 376............. POSTPARTUM & POST

110 Quintile 3. ABORTION DIAGNOSES W/O O.R. PROCEDURE. 377............. POSTPARTUM & POST

110 Quintile 3. ABORTION DIAGNOSES W O.R. PROCEDURE. 378............. ECTOPIC PREGNANCY...

369 Quintile 3. 379............. THREATENED ABORTION.

110 Quintile 3. 380............. ABORTION W/O D&C....

110 Quintile 3. 381............. ABORTION W D&C,

110 Quintile 3. ASPIRATION CURETTAGE OR HYSTEROTOMY. 382............. FALSE LABOR.........

110 Quintile 3. 383............. OTHER ANTEPARTUM

110 Quintile 3. DIAGNOSES W MEDICAL COMPLICATIONS. 384............. OTHER ANTEPARTUM

110 Quintile 3. DIAGNOSES W/O MEDICAL COMPLICATIONS. 385............. NEONATES, DIED OR

110 Quintile 3. TRANSFERRED TO ANOTHER ACUTE CARE FACILITY. 386............. KIDNEY & URINARY

87 Quintile 4. TRACT INFECTIONS AGE 0-17. 387............. URINARY STONES W/O

87 Quintile 4. CC. 388............. PREMATURITY W/O

110 Quintile 3. MAJOR PROBLEMS. 389............. KIDNEY & URINARY

87 Quintile 4. TRACT SIGNS & SYMPTOMS AGE 0-17. 390............. VIRAL ILLNESS &

87 Quintile 4. FEVER OF UNKNOWN ORIGIN AGE 0-17. 391............. NORMAL NEWBORN......

110 Quintile 3. 392............. SPLENECTOMY AGE >17.

197 Quintile 3. 393............. SPLENECTOMY AGE 0-17

197 Quintile 3. 396............. RED BLOOD CELL

399 Quintile 2. DISORDERS AGE 0-17. 402............. APPENDECTOMY W

395 Quintile 2. COMPLICATED PRINCIPAL DIAG W/O CC. 405............. ACUTE LEUKEMIA W/O

404 Quintile 2. MAJOR O.R. PROCEDURE AGE 0-17. 407............. MYELOPROLIF DISORD

408 Quintile 4. OR POORLY DIFF NEOPL W MAJ O.R.PROC W/O CC. 411............. HISTORY OF

110 Quintile 3. MALIGNANCY W/O ENDOSCOPY. 412............. HISTORY OF

110 Quintile 3. MALIGNANCY W ENDOSCOPY. 414............. OTHER MYELOPROLIF

399 Quintile 2. DIS OR POORLY DIFF NEOPL DIAG W/O CC. 417............. APPENDECTOMY W/O

416 Quintile 3. COMPLICATED PRINCIPAL DIAG W/O CC. 420............. FEVER OF UNKNOWN

419 Quintile 4. ORIGIN AGE >17 W/O CC. 422............. ADRENAL & PITUITARY

419 Quintile 4. PROCEDURES. 446............. TRAUMATIC INJURY AGE

445 Quintile 1. 0-17. 448............. ALLERGIC REACTIONS

447 Quintile 2. AGE 0-17. 450............. POISONING & TOXIC

449 Quintile 3. EFFECTS OF DRUGS AGE >17 W/O CC. 451............. POISONING & TOXIC

449 Quintile 3. EFFECTS OF DRUGS AGE 0-17. 455............. OTHER INJURY,

449 Quintile 3. POISONING & TOXIC EFFECT DIAG W/O CC. 478............. OTHER VASCULAR

110 Quintile 3. PROCEDURES W CC. 479............. OTHER VASCULAR

110 Quintile 3. PROCEDURES W/O CC. 481............. BONE MARROW

394 Quintile 5. TRANSPLANT. 485............. OTHER HEART ASSIST

487 Quintile 4. SYSTEM IMPLANT. 492............. APPENDECTOMY W/O

410 Quintile 4. COMPLICATED PRINCIPAL DIAG W CC. 494............. LAPAROSCOPIC

493 Quintile 5. CHOLECYSTECTOMY W/O C.D.E. W/O CC. 496............. COMBINED ANTERIOR/

497 Quintile 4. POSTERIOR SPINAL FUSION. 498............. SPINAL FUSION W/O CC

497 Quintile 4. 504............. CHEMOTHERAPY W ACUTE

468 Quintile 5. LEUKEMIA AS SECONDARY DIAGNOSIS. 518............. PERCUTANEOUS

125 Quintile 3. CARDIVASCULAR PROC W/O CORONARY ARTERY STENT OR AMI. 520............. CERVICAL SPINAL

497 Quintile 4. FUSION W/O CC. 522............. CARDIAC DEFIB

521 Quintile 1. IMPLANT W CARDIAC CATH W AMI/HF/SHOCK. 523............. CARDIAC DEFIB

521 Quintile 1. IMPLANT W CARDIAC CATH W/O AMI/HF/ SHOCK. 525............. EXTENSIVE BURN OR

468 Quintile 5. FULL THICKNESS BURNS WITH MECH VENT 96+ HOURS WITH SKIN GRAFT. 528............. INTRACRANIAL

1 Quintile 5. VASCULAR PROC W PDX HEMORRHAGE. 530............. VENTRICULAR SHUNT

529 Quintile 5. PROCEDURES W/O CC. 534............. EXTRACRANIAL

500 Quintile 4. VASCULAR PROCEDURES WITHOUT CC. 535............. ACUTE ISCHEMIC

515 Quintile 5. STROKE WITH USE OF THROMBOLYTIC AGENT. 536............. KIDNEY & URINARY

515 Quintile 5. TRACT SIGNS & SYMPTOMS AGE >17 W/ O CC. 538............. LOCAL EXCISION AND

228 Quintile 4. REMOVAL OF INTERNAL FIXATION DEVICES EXCEPT HIP AND FEMUR WITHOUT CC. 540............. LYMPHOMA AND

399 Quintile 2. LEUKEMIA WITH MAJOR O.R. PROCEDURE WITHOUT CC. 546............. SPINAL FUSION EXCEPT

499 Quintile 5. CERVICAL WITH CURVATURE OF SPINE OR MALIGNANCY.

To illustrate this methodology for determining the relative weights for the 201 LTC-DRGs with no LTCH cases, we are providing the following examples, which refer to the no volume LTC-DRGs crosswalk information for FY 2006 provided in the chart above.

Example 1: There were no cases in the FY 2004 MedPAR file used for this final rule for LTC-DRG 163 (Hernia Procedures Age 0-17). Since the procedure is similar in resource use and the length and complexity of the procedures and the length of stay are similar, we determined that LTC-DRG 178 (Uncomplicated Peptic Ulcer Without CC), which is assigned to low-

[[Page 47341]]

volume Quintile 3 for the purpose of determining the FY 2006 relative weights, would display similar clinical and resource use. Therefore, we assign the same relative weight of LTC-DRG 178 of 0.7637 (Quintile 3) for FY 2006 (Table 11 in the Addendum to this final rule) to LTC-DRG 163.

Example 2: There were no LTCH cases in the FY 2004 MedPAR file used in this final rule for LTC-DRG 91 (Simple Pneumonia and Pleurisy Age 0- 17). Since the severity of illness in patients with bronchitis and asthma is similar in patients regardless of age, we determined that LTC-DRG 90 (Simple Pneumonia and Pleurisy Age >17 Without CC) would display similar clinical and resource use characteristics and have a similar length of stay to LTC-DRG 91. There were over 25 cases in LTC- DRG 90. Therefore, it would not be assigned to a low-volume quintile for the purpose of determining the LTC-DRG relative weights. However, under our established methodology, LTC-DRG 91, with no LTCH cases, would need to be grouped to a low-volume quintile. We determined that the low-volume quintile with the closest weight to LTC-DRG 90 (0.4970) (refer to Table 11 in the Addendum to this final rule) would be low- volume Quintile 1 (0.4499) (refer to Table 11 in the Addendum to this final rule). Therefore, we assign LTC-DRG 91 a relative weight of 0.4499 for FY 2006.

Furthermore, we are establishing LTC-DRG relative weights of 0.0000 for heart, kidney, liver, lung, pancreas, and simultaneous pancreas/ kidney transplants (LTC-DRGs 103, 302, 480, 495, 512, and 513, respectively) for FY 2006 because Medicare will only cover these procedures if they are performed at a hospital that has been certified for the specific procedures by Medicare and presently no LTCH has been so certified.

Based on our research, we found that most LTCHs only perform minor surgeries, such as minor small and large bowel procedures, to the extent any surgeries are performed at all. Given the extensive criteria that must be met to become certified as a transplant center for Medicare, we believe it is unlikely that any LTCHs would become certified as a transplant center. In fact, in the nearly 20 years since the implementation of the IPPS, there has never been a LTCH that even expressed an interest in becoming a transplant center.

However, if in the future a LTCH applies for certification as a Medicare-approved transplant center, we believe that the application and approval procedure would allow sufficient time for us to determine appropriate weights for the LTC-DRGs affected. At the present time, we would only include these six transplant LTC-DRGs in the GROUPER program for administrative purposes. Because we use the same GROUPER program for LTCHs as is used under the IPPS, removing these LTC-DRGs would be administratively burdensome.

Again, we note that as this system is dynamic, it is entirely possible that the number of LTC-DRGs with a zero volume of LTCH cases based on the system will vary in the future. We used the best most recent available claims data in the MedPAR file to identify zero volume LTC-DRGs and to determine the relative weights in this final rule.

Table 11 in the Addendum to this final rule lists the LTC-DRGs and their respective relative weights, geometric mean length of stay, and five-sixths of the geometric mean length of stay (to assist in the determination of short-stay outlier payments under Sec. 412.529) for FY 2006. 5. Other Public Comments Relating to the LTCH PPS Payment Policies

Comment: One commenter submitted comments that addressed aspects of the existing LTCH PPS, including the hospital-within-hospital policy, which was discussed in the FY 2005 IPPS final rule (69 FR 49191), and the June 2004 MedPAC recommendations concerning the definition of LTCHs, which was discussed in the RY 2006 LTCH PPS final rule (70 FR 5757), for which we did not propose LTCH policy changes in the FY 2006 IPPS proposed rule.

Response: Because those comments pertain to specific aspects of the existing LTCH PPS that were not specific proposed changes to the LTCH PPS presented in the FY 2006 IPPS proposed rule, we are not responding to them at this time. Rather, we believe it is more appropriate to address the issues in the annual LTCH PPS proposed and final rules. We will consider the issues raised in those comments in the context of future rulemaking for the LTCH PPS.

E. Add-On Payments for New Services and Technologies

1. Background

Sections 1886(d)(5)(K) and (L) of the Act establish a process of identifying and ensuring adequate payment for new medical services and technologies under the IPPS. Section 1886(d)(5)(K)(vi) of the Act specifies that a medical service or technology will be considered new if it meets criteria established by the Secretary after notice and opportunity for public comment. Section 1886(d)(5)(K)(ii)(I) of the Act specifies that the process must apply to a new medical service or technology if, ``based on the estimated costs incurred with respect to discharges involving such service or technology, the DRG prospective payment rate otherwise applicable to such discharges under this subsection is inadequate.''

The regulations implementing this provision establish three criteria for new medical services and techniques to receive an additional payment. First, Sec. 412.87(b)(2) defines when a specific medical service or technology will be considered new for purposes of new medical service or technology add-on payments. The statutory provision contemplated the special payment treatment for new medical services or technologies until such time as data are available to reflect the cost of the technology in the DRG weights through recalibration. There is a lag of 2 to 3 years from the point a new medical service or technology is first introduced on the market and when data reflecting the use of the medical service or technology are used to calculate the DRG weights. For example, data from discharges occurring during FY 2004 are used to calculate the FY 2006 DRG weights in this final rule. Section 412.87(b)(2) provides that a ``medical service or technology may be considered new within 2 or 3 years after the point at which data begin to become available reflecting the ICD-9- CM code assigned to the new medical service or technology (depending on when a new code is assigned and data on the new medical service or technology become available for DRG recalibration). After CMS has recalibrated the DRGs, based on available data, to reflect the costs of an otherwise new medical service or technology, the medical service or technology will no longer be considered `new' under the criterion for this section.''

The 2-year to 3-year period during which a technology or medical service can be considered new would ordinarily begin with FDA approval, unless there was some documented delay in bringing the product onto the market after that approval (for instance, component production or drug production had been postponed until FDA approval due to shelf life concerns or manufacturing issues). After the DRGs have been recalibrated to reflect the costs of an otherwise new medical service or technology, the special add-on payment for new medical services or technology ceases (Sec. 412.87(b)(2)). For example, an approved new technology that received FDA approval in October 2004 and entered the market at that time may be eligible to receive add-on payments as a new technology until FY 2007

[[Page 47342]]

(discharges occurring before October 1, 2006), when data reflecting the costs of the technology would be used to recalibrate the DRG weights. Because the FY 2007 DRG weights will be calculated using FY 2005 MedPAR data, the costs of such a new technology would likely be reflected in the FY 2007 DRG weights.

Section 412.87(b)(3) further provides that, to receive special payment treatment, new medical services or technologies must be inadequately paid otherwise under the DRG system. To assess whether technologies would be inadequately paid under the DRGs, we establish thresholds to evaluate applicants for new technology add-on payments. In the FY 2004 IPPS final rule (68 FR 45385, August 1, 2003), we established the threshold at the geometric mean standardized charge for all cases in the DRG plus 75 percent of 1 standard deviation above the geometric mean standardized charge (based on the logarithmic values of the charges and transformed back to charges) for all cases in the DRG to which the new medical service or technology is assigned (or the case-weighted average of all relevant DRGs, if the new medical service or technology occurs in many different DRGs). Table 10 in the Addendum to the FY 2004 IPPS final rule (68 FR 45648) listed the qualifying threshold by DRG, based on the discharge data that we used to calculate the FY 2004 DRG weights.

However, section 503(b)(1) of Pub. L. 108-173 amended section 1886(d)(5)(K)(ii)(I) of the Act to provide for ``applying a threshold * * * that is the lesser of 75 percent of the standardized amount (increased to reflect the difference between cost and charges) or 75 percent of 1 standard deviation for the diagnosis-related group involved.'' The provisions of section 503(b)(1) apply to classification for fiscal years beginning with FY 2005. We updated Table 10 from the Federal Register document that corrects the FY 2004 final rule (68 FR 57753, October 6, 2003), which contains the thresholds that we used to evaluate applications for new service or technology add-on payments for FY 2005, using the section 503(b)(1) measures stated above, and posted these new thresholds on our Web site at: http://www.cms.hhs.gov/providers/hipps/newtech.asp. In the FY 2005 IPPS final rule (in Table

10 of the Addendum), we included the final thresholds that are being used to evaluate applicants for new technology add-on payments for FY 2006. (Refer to section IV.D. of the preamble to the FY 2005 IPPS final rule (69 FR 49084, August 11, 2004) for a discussion of a revision of the regulations to incorporate the change made by section 503(b)(1) of Pub. L. 108-173.)

Section 412.87(b)(1) of our existing regulations provides that a new technology is an appropriate candidate for an additional payment when it represents an advance in medical technology that substantially improves, relative to technologies previously available, the diagnosis or treatment of Medicare beneficiaries. For example, a new technology represents a substantial clinical improvement when it reduces mortality, decreases the number of hospitalizations or physician visits, or reduces recovery time compared to the technologies previously available. (See the September 7, 2001 final rule, 66 FR 46902, for a complete discussion of this criterion.)

The new medical service or technology add-on payment policy provides additional payments for cases with high costs involving eligible new medical services or technologies while preserving some of the incentives under the average-based payment system. The payment mechanism is based on the cost to hospitals for the new medical service or technology. Under Sec. 412.88, Medicare pays a marginal cost factor of 50 percent for the costs of a new medical service or technology in excess of the full DRG payment. If the actual costs of a new medical service or technology case exceed the DRG payment by more than the 50- percent marginal cost factor of the new medical service or technology, Medicare payment is limited to the DRG payment plus 50 percent of the estimated costs of the new technology.

The report language accompanying section 533 of Pub. L. 106-554 indicated Congressional intent that the Secretary implement the new mechanism on a budget neutral basis (H.R. Conf. Rep. No. 106-1033, 106th Cong., 2nd Sess. at 897 (2000)). Section 1886(d)(4)(C)(iii) of the Act requires that the adjustments to annual DRG classifications and relative weights must be made in a manner that ensures that aggregate payments to hospitals are not affected. Therefore, in the past, we accounted for projected payments under the new medical service and technology provision during the upcoming fiscal year at the same time we estimated the payment effect of changes to the DRG classifications and recalibration. The impact of additional payments under this provision was then included in the budget neutrality factor, which was applied to the standardized amounts and the hospital-specific amounts.

Section 1886(d)(5)(K)(ii)(III) of the Act, as amended by section 503(d)(2) of Pub. L. 108-173, provides that there shall be no reduction or adjustment in aggregate payments under the IPPS due to add-on payments for new medical services and technologies. Therefore, add-on payments for new medical services or technologies for FY 2005 and later years will not be budget neutral.

Applicants for add-on payments for new medical services or technologies for FY 2007 must submit a formal request, including a full description of the clinical applications of the medical service or technology and the results of any clinical evaluations demonstrating that the new medical service or technology represents a substantial clinical improvement, along with a significant sample of data to demonstrate the medical service or technology meets the high-cost threshold, no later than October 15, 2005. Applicants must submit a complete database no later than December 30, 2005. Complete application information, along with final deadlines for submitting a full application, will be available after publication of this final rule at our Web site: http://www.cms.hhs.gov/providers/hipps/default.asp. To

allow interested parties to identify the new medical services or technologies under review before the publication of the proposed rule for FY 2007, the Web site will also list the tracking forms completed by each applicant. 2. Public Input Before Publication of a Notice of Proposed Rulemaking on Add-On Payments

Section 1886(d)(5)(K)(viii) of the Act, as amended by section 503(b)(2) of Pub. L. 108-173, provides for a mechanism for public input before publication of a notice of proposed rulemaking regarding whether a medical service or technology represents a substantial improvement or advancement. The process for evaluating new medical service and technology applications requires the Secretary to--

Provide, before publication of a proposed rule, for public input regarding whether a new service or technology represents an advance in medical technology that substantially improves the diagnosis or treatment of Medicare beneficiaries.

Make public and periodically update a list of the services and technologies for which an application for add-on payments is pending.

Accept comments, recommendations, and data from the public regarding whether a service or technology represents a substantial improvement.

[[Page 47343]]

Provide, before publication of a proposed rule, for a meeting at which organizations representing hospitals, physicians, manufacturers, and any other interested party may present comments, recommendations, and data regarding whether a new service or technology represents a substantial clinical improvement to the clinical staff of CMS.

In order to provide an opportunity for public input regarding add- on payments for new medical services and technologies for FY 2006 before publication of the FY 2006 IPPS proposed rule, we published a notice in the Federal Register on December 30, 2004 (69 FR 78466) and held a town hall meeting at the CMS Headquarters Office in Baltimore, MD, on February 23, 2005. In the announcement notice for the meeting, we stated that the opinions and alternatives provided during the meeting would assist us in our evaluations of applications by allowing public discussions of the substantial clinical improvement criteria for each of the FY 2006 new medical service and technology add-on payment applications before the publication of the FY 2006 IPPS proposed rule.

Approximately 45 participants registered and attended in person, while additional participants listened over an open telephone line. The participants focused on presenting data on the substantial clinical improvement aspect of their products, as well as the need for additional payments to ensure access to Medicare beneficiaries. In addition, we received written comments regarding the substantial clinical improvement criterion for the applicants. We considered these comments in our evaluation of each new application for FY 2006 in the proposed rule and in this final rule. We have summarized these comments or, if applicable, indicated that no comments were received, at the end of the discussion of the individual applications.

Section 1886(d)(5)(K)(ix) of the Act, as added by section 503(c) of Pub. L. 108-173, requires that, before establishing any add-on payment for a new medical service or technology, the Secretary shall seek to identify one or more DRGs associated with the new technology, based on similar clinical or anatomical characteristics and the costs of the technology and assign the new technology into a DRG where the average costs of care most closely approximate the costs of care using the new technology. No add-on payment shall be made with respect to such a new technology.

At the time an application for new technology add-on payments is submitted, the DRGs associated with the new technology are identified. We only determine that a new DRG assignment is necessary or a new technology add-on payment is appropriate when the reimbursement under these currently assigned DRGs is not adequate for this new technology. The criterion for this determination is the cost threshold, which we discuss below. We discuss the assignments of several new technologies within the DRG payment system in section II.B. of this final rule.

In this final rule, we evaluate whether new technology add-on payments will continue in FY 2006 for the three technologies that currently receive such payments. In addition, we present our evaluations of eight applications for add-on payments in FY 2006. The eight applications for FY 2006 include two applications for products that were denied new technology add-on payments for FY 2005.

Comment: Commenters argued that CMS' interpretation of the newness criterion is inconsistent with the statute and that, as a result, CMS is prematurely denying eligibility for many technologies. Commenters believed that instead of basing the newness criterion on FDA approval or market availability, CMS should start the 2-3 year period that a technology can be considered new from the later of the date that the technology is assigned an ICD-9-CM code or is approved by the FDA. Commenters argued that neither the statutory language nor the regulatory language refers to the date of FDA approval in determining whether a technology is new. One commenter further argued that CMS should ensure a maximum period of eligibility for new technology add-on payments that takes into account a ``host of `newness' factors'' such as production and distribution, negotiation with hospitals, and physician education programs. The commenter proposed that CMS determine newness, based on the latest of the following dates:

Date of ICD-9-CM code assignment;

Date of FDA approval plus six months; or

The time/date at which 50 percent of the Fiscal Intermediaries are processing claims that include the technology in question.

The commenter further recommended that, given the numerous challenges of bringing a device to market, CMS should extend the period that a product is considered new from two to three years to four or five years.

Response: Section 1886(d)(5)(K)(vi) of the Act provides the Secretary with broad discretion to define a ``new medical service or technology.'' As we have indicated in prior rules (for example, see 66 FR 46914, September 7, 2001), we believe that a product should be considered new 2 to 3 years from the date a product becomes available on the market (generally from the date of FDA approval unless an applicant can demonstrate that there was a delay in making the product available on the market). Once a product becomes available on the market, hospitals that use the new technology will begin including charges for the product on their bills under either an existing or new ICD-9-CM code. These charges will be used to set the DRG relative weights two years later (that is, FY 2004 charge data are being used to set the FY 2006 DRG relative weights). Therefore, 2 to 3 years after the technology is available on the market, there will be a full year of Medicare charge data used to set the relative weights that will reflect the cost of the device. We note that a manufacturer can reasonably predict when a product will become available on the market and, if warranted, could request a new ICD-9-CM code in order to distinctly identify the new technology in our data. In the FY 2005 final rule (69 FR 49002), we provided a detailed explanation for why using the date on which a specific ICD-9-CM code is assigned to a technology is not an appropriate test of newness. In that rule, we noted that, in many instances, a technology may have been in use for several years, or even several decades, prior to the assignment of a new code (69 FR 49003). Thus, we believe it is appropriate to continue to determine newness based on the date on which a product becomes available for use in the Medicare population and the date when hospitals can begin to use either an existing or new ICD-9-CM code to bill for the new service or technology.

Comment: One commenter indicated that, because Medicare does not pay for devices during clinical trials, ``little or no internal Medicare claims data exist upon which to base an initial DRG assignment for new technologies.'' To address this issue, commenters suggested that CMS should accept external data while maintaining confidentiality for proprietary data. Other commenters indicated that CMS decisions regarding substantial clinical improvement have been largely subjective and made without stakeholder input. Commenters requested that CMS include ``a consistent and reasonable set of requirements for manufacturers of novel technologies to meet'' in order to be eligible for new technology add-on

[[Page 47344]]

payments. Several commenters indicated the process for applying for new technology add-on payments is particularly burdensome for smaller companies. Commenters urged CMS to provide a preliminary assessment of substantial clinical improvement for each technology in the proposed rule, in order for the public to respond CMS' findings during the public comment period.

Response: With respect to the comment about the lack of Medicare claims data for making a DRG assignment for a new medical product, we believe that the new technology process is intended to address precisely this issue. In our evaluation of a new technology application, we consider any external data provided by the applicant to make judgments as to whether a product meets the three criteria we have established either to assign a new technology to a different DRG or to approve a new technology for add-on payments. In addition, while we generally do not pay for an experimental device itself when used as part of a clinical trial, a hospital is not precluded from including an existing or a newly assigned ICD-9-CM code or V-Codes on its bill for Medicare covered services. Thus, we have been able to successfully track devices that are (or were) in clinical trials in our MedPAR data, and have used these data to determine whether several new technologies have met the cost threshold for new technology payment. We addressed the concerns over submissions of external data and proprietary information in the FY 2005 final rule (69 FR 49004, August 11, 2004). As indicated in that rule, we are continuing to consider this issue, but we are not making any changes to our policy on the submission of external data and proprietary information at this time.

We disagree that determinations regarding applications for add-on payments are made without stakeholder input. There is ample opportunity for applicants and other interested parties to make their views known to us throughout the application process, at the public meeting, as well as during the comment period on the proposed rule. We have had numerous meetings with applicants where they have addressed our concerns and/or brought further information to our attention on the merits of their technology. Our initial new technology final rule (66 FR 46914, September 7, 2001) provides the specific guidelines we consider to determine whether a technology is a substantial clinical improvement. In that final rule, we indicated that, in order to meet the substantial clinical improvement criteria, a new technology must be able to offer a new treatment option for a patient population unresponsive to, or ineligible for, currently available treatments; diagnose a previously undetectable condition or allow for earlier diagnosis; or significantly improve clinical outcomes. We provided seven potential measures to evaluate this third standard. While our regulations provide specific criteria for evaluating substantial clinical improvement, by its very nature, this process involves judgment. Before making a final judgment about substantial clinical improvement, we carefully consider all of the information that is provided to us in a new technology application, as well as the viewpoints expressed through the public meeting, during the comment period, and in meetings with individual applicants.

We do not believe that our criteria present an inordinately cumbersome burden for smaller companies that want to apply for new technology add-on payments. Several small companies have already approached us seeking advice on how to apply for new technology add-on payments FY 2007 and later years. We encourage potential applicants to contact us before their technology is available on the market to become familiar with the new technology application process.

With respect to providing preliminary determinations of substantial clinical improvement in the proposed rule, we addressed this issue in the FY 2006 proposed rule (70 FR 23359). We indicated that our decision about new technology add-on payments follows a logical sequence of determinations, moving from the newness criterion, to the cost criterion and finally to the substantial clinical improvement criterion. Therefore, we are reluctant to import substantial clinical improvement considerations into the logically prior decisions about whether technologies satisfy the newness and cost criteria. We acknowledge that an applicant seeking new technology payment for a product expected to receive FDA approval between the proposed and final rule has an interest in knowing CMS' findings about substantial clinical improvement. Nevertheless, we believe that FDA approval of a product is a logical prior determination because substantial clinical improvement is a higher standard to meet than either of the FDA standards for allowing a product on the market. If a product does not meet the FDA standards for a pre-market (``safe and effective'') or humanitarian device exemption (``safe'') approval, it cannot be a substantial clinical improvement. While we do not believe a determination about substantial clinical improvement should be made prior to FDA approval, two applicants have received FDA approval for their products since the publication of the proposed rule. We met with these two applicants during the public comment period to discuss our concerns about substantial clinical improvement. As indicated below, we are approving both of these technologies for new technology add-on payments beginning in FY 2006. 3. FY 2006 Status of Technology Approved for FY 2005 Add-On Payments a. INFUSE[supreg] (Bone Morphogenetic Proteins (BMPs) for Spinal Fusions)

INFUSETMwas approved by FDA for use on July 2, 2002, and became available on the market immediately thereafter. In the FY 2004 IPPS final rule (68 FR 45388), we approved INFUSE[supreg] for add- on payments under Sec. 412.88, effective for FY 2004. This approval was on the basis of using INFUSE[supreg] for single-level, lumbar spinal fusion, consistent with the FDA's approval and the data presented to us by the applicant. Therefore, we limited the add-on payment to cases using this technology for anterior lumbar fusions in DRGs 497 (Spinal Fusion Except Cervical With CC) and 498 (Spinal Fusion Except Cervical Without CC). Cases involving INFUSE[supreg] that are eligible for the new technology add-on payment are identified by assignment to DRGs 497 and 498 as a lumbar spinal fusion, with the combination of ICD-9-CM procedure codes 84.51 (Insertion of interbody spinal fusion device) and 84.52 (Insertion of recombinant bone morphogenetic protein).

The FDA approved INFUSE[supreg] for use on July 2, 2002. For FY 2005, INFUSE[supreg] was still within the 2-year to 3-year period during which a technology can be considered new under the regulations. Therefore, in the FY 2005 IPPS final rule (69 FR 49007 through 49009), we continued add-on payments for FY 2005 for cases receiving INFUSE[supreg] for spinal fusions in DRGs 497 (Spinal Fusion Except Cervical With CC) and 498 (Spinal Fusion Except Cervical Without CC).

As we discussed in the new technology final rule (66 FR 46915), September 7, 2001 an approval of a new technology for special payment should extend to all technologies that are substantially similar. Otherwise, our payment policy would bestow an advantage to the first applicant to receive approval for a particular new

[[Page 47345]]

technology. In last year's final rule (69 FR 49008), we discussed another product, called OP-1 Putty, manufactured by Stryker Biotech, that promotes natural bone growth by using a closely related bone morphogenetic protein called rhBMP-7. (INFUSE[supreg] is rhBMP-2.) We also stated in last year's final rule that we had determined that the costs associated with the OP-1 Putty are similar to those associated with INFUSE[supreg]. Because the OP-1 Putty became available on the market in May 2004 (when it received FDA approval for spinal fusions) for similar spinal fusion procedures and because this product also eliminates the need for the autograft bone surgery, we extended new technology add-on payments to this technology as well for FY 2005.

As noted above, the period for which technologies are eligible to receive new technology add-on payments is 2 to 3 years after the product becomes available on the market and data reflecting the cost of the technology are reflected in the DRG weights. The FDA approved INFUSE[supreg] bone graft on July 2, 2002. Therefore, data reflecting the cost of the technology are now reflected in the DRG weights. In addition, by the end of FY 2005, the add-on payment will have been made for 2 years. Therefore, as we proposed, we are discontinuing new technology add-on payment for INFUSE[supreg] for FY 2006. Because we apply the same policies in making new technology payment for OP-1 Putty as we do for INFUSE[supreg], we are also discontinuing new technology add-on payment for OP-1 Putty for FY 2006.

Comment: Several commenters agreed with our proposal to terminate add-on payment for INFUSE[supreg] bone graft for spinal fusions.

Response: We are finalizing our proposal to terminate new technology add-on payments for INFUSE[supreg] bone graft for spinal fusions in this final rule. b. InSync[supreg] Defibrillator System (Cardiac Resynchronization Therapy with Defibrillation (CRT-D))

Cardiac Resynchronization Therapy (CRT), also known as bi- ventricular pacing, is a therapy for chronic heart failure. A CRT implantable system provides electrical stimulation to the right atrium, right ventricle, and left ventricle to coordinate or resynchronize ventricular contractions and improve cardiac output.

In the FY 2005 IPPS final rule (69 FR 49016), we determined that cardiac resynchronization therapy with defibrillator (CRT-D) was eligible for add-on payments in FY 2005. Cases involving CRT-D that are eligible for new technology add-on payments are identified by either one of the following two ICD-9-CM procedure codes: 00.51 (Implantation of Cardiac Resynchronization Defibrillator, Total System (CRT-D)) or 00.54 (Implantation or Replacement of Pulse Generator Device Only (CRT- D)). InSync[supreg] Defibrillation System received FDA approval on June 26, 2002. However, another manufacturer, Guidant, received FDA approval for its CRT-D device on May 2, 2002. As we discussed in the new technology final rule (66 FR 46915, September 7, 2001), an approval of a new technology for special payment should extend to all technologies that are substantially similar. Otherwise, our payment policy would bestow an advantage to the first applicant to receive approval for a particular new technology. In the FY 2005 final rule, we also noted that we would extend new technology add-on payments for CRT-D for the entire FY 2005 even though the 2-3 year period of newness ended in May 2005 for CRT-D. Predictability is an important aspect of the prospective payment methodology and, therefore, we believe it is appropriate to apply a consistent payment methodology for new technologies throughout the fiscal year (69 FR 49016).

As noted in the FY 2005 IPPS final rule (69 FR 49014), because CRT- Ds were available upon the initial FDA approval in May 2002, we considered the technology to be new from this date. As a result, for FY 2006, the CRT-D will be beyond the 2-3 year period during which a technology can be considered new. Therefore, as we proposed, we are discontinuing add-on payments for the CRT-D for FY 2006.

Comment: One commenter thanked CMS for approving add-on payments for the CRT-D. The commenter also indicated that add-on payment for this device had contributed significantly to patient access and broader physician adoption of this new treatment. Another commenter requested that CMS continue to make add-on payment for CRT-D to avoid financial problems that hospitals will experience if payment is ceased.

Response: We appreciate the commenter's support of our decision to approve add-on payments for CRT-D. Consistent with section 1886(d)(5)(K)(ii) of the Act, the regulations do not permit us to extend payment for CRT-D beyond the 2-3 year period during which a technology can be considered new. Therefore, we are finalizing our proposal to discontinue add on payments for the CRT-D in FY 2006. c. Kinetra[supreg] Implantable Neurostimulator for Deep Brain Stimulation

Medtronic, Inc. submitted an application for approval of the Kinetra[supreg] implantable neurostimulator device for new technology add-on payments for FY 2005. The Kinetra[supreg] device was approved by the FDA on December 16, 2003. The Kinetra[supreg] implantable neurostimulator is designed to deliver electrical stimulation to the subthalamic nucleus (STN) or internal globus pallidus (GPi) in order to ameliorate symptoms caused by abnormal neurotransmitter levels that lead to abnormal cell-to-cell electrical impulses in Parkinson's Disease and essential tremor. Before the development of Kinetra[supreg], treating bilateral symptoms of patients with these disorders required the implantation of two neurostimulators (in the form of a product called SoletraTM, also manufactured by Medtronic): one for the right side of the brain (to control symptoms on the left side of the body), the other for the left side of the brain (to control symptoms on the right side of the body). Additional procedures were required to create pockets in the chest cavity to place the two generators required to run the individual leads. The Kinetra[supreg] neurostimulator generator, implanted in the pectoral area, is designed to eliminate the need for two devices by accommodating two leads that are placed in both the left and right sides of the brain to deliver the necessary impulses. The manufacturer argued that the development of a single neurostimulator that treats bilateral symptoms provides a less invasive treatment option for patients, and simpler implantation, followup, and programming procedures for physicians.

The FDA approved the device in December 2003. Therefore, for FY 2006, Kinetra[supreg] qualifies under the newness criterion because FDA approval was within the statutory timeframe of 2 to 3 years and its costs are not yet reflected in the DRG weights. Because there were no data available to evaluate costs associated with Kinetra[supreg], in the FY 2005 IPPS final rule, we conducted the cost analysis using SoletraTM, the predecessor technology used to treat this condition, as a proxy for Kinetra[supreg]. The preexisting technology provided the closest means to track cases that have actually used similar technology and served to identify the need and use of the new device. The manufacturer informed us that the cost of the Kinetra[supreg] device is twice the price of a single SoletraTMdevice. Because most patients would receive two SoletraTMdevices if the Kinetra[supreg] device is not implanted,

[[Page 47346]]

we believed data regarding the cost of SoletraTMwould give a good measure of the actual costs that would be incurred. Medtronic submitted data for 104 cases that involved the SoletraTM device (26 cases in DRG 1 (Craniotomy Age > 17 With CC), and 78 cases in DRG 2 (Craniotomy Age > 17 Without CC)). These cases were identified from the FY 2002 MedPAR file using procedure codes 02.93 (Implantation, intracranial neurostimulator) and 86.09 (Other incision of skin and subcutaneous tissue). In the analysis presented by the applicant, the mean standardized charges for cases involving SoletraTMin DRGs 1 and 2 were $69,018 and $44,779, respectively. The mean standardized charge for these SoletraTMcases according to Medtronic's data was $50,839.

Last year, we used the same procedure codes to identify 187 cases involving the SoletraTMdevice in DRGs 1 and 2 in the FY 2003 MedPAR file. Similar to the Medtronic data, 53 of the cases were found in DRG 1, and 134 cases were found in DRG 2. The average standardized charges for these cases in DRGs 1 and 2 were $51,163 and $44,874, respectively. Therefore, the case-weighted average standardized charges for cases that included implantation of the SoletraTMdevice were $46,656. The new cost thresholds established under the revised criteria in Pub. L. 108-173 for DRGs 1 and 2 are $43,245 and $30,129, respectively. Accordingly, the case- weighted threshold to qualify for new technology add-on payment, using the data we identified, was determined to be $33,846. Under this analysis, Kinetra[supreg] met the cost threshold.

We note that an ICD-9-CM code was approved for dual array pulse generator devices, effective October 1, 2004, for IPPS tracking purposes. The new ICD-9-CM code assigned to this device is 86.95 (Insertion or replacement of dual array neurostimulator pulse generator), which includes dual array and dual channel generators for intracranial, spinal, and peripheral neurostimulators. The code does not separately identify cases with the Kinetra[supreg] device and is only used to distinguish single versus dual channel-pulse generator devices. Because the code only became effective on October 1, 2004, we do not have any specific data regarding the costs of cases involving dual array pulse generator devices.

The manufacturer claimed that Kinetra[supreg] provides a range of substantial improvements beyond previously available technology. These include a reduced rate of device-related complications and hospitalizations or physician visits and less surgical trauma because only one generator implantation procedure is required. Kinetra[supreg] has a reed switch disabling function that physicians can use to prevent inadvertent shutoff of the device, as occurs when accidentally tripped by electromagnetic inference (caused by common products such as metal detectors and garage door openers). Kinetra[supreg] also provides significant patient control, allowing patients to monitor whether the device is on or off, to monitor battery life, and to fine-tune the stimulation therapy within clinician-programmed parameters. While Kinetra[supreg] provides the ability for patients to better control their symptoms and reduce the complications associated with the existing technology, it does not eliminate the necessity for two surgeries. Because the patients who receive the device are often frail, the implantation generally occurs in two phases: The brain leads are implanted in one surgery, and the generator is implanted in another surgery, typically on another day. However, implanting Kinetra[supreg] does reduce the number of potential surgeries compared to its predecessor (which requires two surgeries to implant the two single- lead arrays to the brain and an additional surgery for implantation of the second generator). Therefore, the Kinetra[supreg] device reduces the number of surgeries from 3 to 2.

Last year, we solicited comments on (1) the issue of whether the device is sufficiently different from the previously used technology to qualify as a substantially improved treatment for the same patient symptoms; (2) the cost of the device; and (3) the approval of the device for add-on payment, given the uncertainty over the frequency with which the patients receiving the device have the generator implanted in a second hospital stay, and the frequency with which this implantation occurs in an outpatient setting. In response, we received sufficient evidence to demonstrate that Kinetra[supreg] does represent a substantial clinical improvement over the previous SoletraTMdevice. Specifically, the increased patient control, reduced surgery, fewer complications, and elimination of environmental interference significantly improve patient outcomes. Therefore, we approved Kinetra[supreg] for new technology add-on payments for FY 2005.

Cases receiving Kinetra[supreg] for Parkinson's disease or essential tremor on or after October 1, 2004, are eligible to receive an add-on payment of up to $8,285, or half the cost of the device, which is approximately $16,570. These cases are identified by the presence of procedure codes 02.93 (Implantation or replacement of intracranial neurostimulator leads) and 86.95 (Insertion or replacement of dual array neurostimulator pulse generator). If a claim has only the procedure code identifying the implantation of the intracranial leads, or if the claim identifies only insertion of the generator, no add-on payment will be made.

This technology received FDA approval on December 16, 2003, and remains within the 2 to 3 year period during which it can be considered new. Therefore, as we proposed, we are continuing add-on payments for Kinetra[supreg] Implantable Neurostimulator for deep brain stimulation for FY 2006.

Comment: Several commenters supported our decision to continue add- on payments for Kinetra[supreg] Implantable Neurostimulator for deep brain stimulation for FY 2006.

Response: In this final rule, we are finalizing our proposal to continue add-on payments for the Kinetra[supreg] Implantable Neurostimulator for deep brain stimulation for FY 2006. 4. FY 2006 Applications for New Technology Add-On a. INFUSE[supreg] Bone Graft (Bone Morphogenetic Proteins (BMPs) for Tibia Fractures)

Bone Morphogenetic Proteins (BMPs) have been shown to have the capacity to induce new bone formation and, therefore, to enhance the healing of fractures. Using recombinant techniques, some BMPs (also referred to as rhBMPs) can be produced in large quantities. This innovation has cleared the way for the potential use of BMPs in a variety of clinical applications such as in delayed union and nonunion of fractured bones and spinal fusions. One such product, rhBMP-2, is developed as an alternative to bone graft with spinal fusions.

Medtronic Sofamor Danek (Medtronic) resubmitted an application (previously submitted for consideration for FY 2005) for a new technology add-on payment in FY 2006 for the use of INFUSE[supreg] Bone Graft in open tibia fractures. In cases of open tibia fractures, INFUSE[supreg] is applied using an absorbable collagen sponge, which is then applied to the fractured bone to promote new bone formation and improved healing. The manufacturer contends that patient access to this technology is restricted due to the increased costs of treating these cases with INFUSE[supreg]. The FDA approved use

[[Page 47347]]

of INFUSE[supreg] for open tibia fractures on April 30, 2004.

Medtronic's first application for a new technology add-on payment for INFUSE[supreg] Bone Graft in open tibia fractures was denied. As we discussed in the FY 2005 IPPS final rule (69 FR 49010), the FY 2005 application for INFUSE[supreg] for open tibia fractures was denied because a similar product, OP-1, was approved in 2001 for the treatment of nonunion of tibia fractures.

Comment: In comments presented at the February 2005 new technology town hall meeting, Medtronic contended that there was no opportunity for public comment on our decision that INFUSE[supreg] for open tibia fractures was substantially similar to OP-1 Implant for recalcitrant long bone non-unions. Medtronic stated that ``the public had no opportunity to comment on whether the follow-on products were `substantially similar' to the primary technologies under consideration. The absence of such provisions led to unpredictability and confusion about the new-technology add-on program.''

Response: In the FY 2005 IPPS final rule, we noted that a commenter brought the existence of the Stryker Biotech OP-1 product to our attention during the comment period on the IPPS proposed rule for FY 2005. The commenter noted OP-1's clinical similarity to INFUSE[supreg] and contended that the products should be treated the same with respect to new technology payments when the product is used for tibia fractures. At that time, we determined that, despite the differences in indications under the respective FDA approvals, the two products were in use for many of the same kinds of cases. Specifically, clinical studies on the safety of OP-1 included patients with complicated fractures of the tibia, and those cases were similar to the cases described in the clinical trials for INFUSE[supreg] for open tibia fractures. In addition, cases involving the use of OP-1 for long bone union and open tibia fractures are assigned to the same DRGs (DRGs 218 and 219 (Lower Extremity Procedures With and Without CC, respectively)) as cases involving INFUSE[supreg]. Therefore, we denied new technology add-on payments for INFUSE[supreg] for open tibia fractures for FY 2005 on the grounds that technology using bone morphogenetic proteins to treat severe long bone fractures (including open tibia fractures) and recalcitrant long bone fractures had been in use for more than 3 years.

We note that Medtronic had ample opportunity, prior to the issuance of the FY 2005 IPPS final rule, to bring to our attention the fact that there was a similar product on the market that was being used in long bone fractures and to explain why this product should not affect our consideration of the application for new technology add-on payments for INFUSE[supreg]. We based our decision for FY 2005 on the record that was placed at our disposal by the applicant and by commenters during the comment period. Nevertheless, we have considered the issues raised by these two products again in the course of evaluating Medtronic's new application for approval of INFUSE[supreg] for open tibia fractures for new technology add-on payments in FY 2006.

As part of its FY 2006 application, Medtronic advanced several arguments designed to demonstrate that OP-1 and INFUSE[supreg] are substantially different. The application cites data from several studies as evidence of the clinical superiority of INFUSE[supreg] over OP-1. Medtronic presented studies at the February 2005 new technology town hall meeting to provide evidence that INFUSE[supreg] is superior to OP-1 in the time it takes for critical-sized defects to heal: in radiographic assessment and mechanical testing of the repaired bone; and in histology of the union for trial subjects receiving INFUSE[supreg] compared with OP-1. (Study subjects were canines whose ulnas had 2.5 cm each of bone removed and then equal amounts of OP-1 and INFUSE[supreg] were put into the front legs in a head to head trial.) Medtronic has also argued that these studies demonstrate that OP-1 has been shown to be less effective than using the patient's own bone or the current standard of care (nail fixation with soft tissue medical management). Medtronic argued that the INFUSE[supreg] product is not only superior to OP-1 for patients with open tibia fractures, but also that it is superior to any other treatment for these serious injuries.

Medtronic also pointed out that the FDA approved OP-1 for Humanitarian Device Exemption (HDE) status, whereas INFUSE[supreg] received a Pre-Market Approval (PMA). To receive HDE approval, a product only needs to meet a safety standard, while standards of both safety and efficacy have to be met for a PMA approval. Medtronic argued that, because the only point the manufacturer of OP-1 was able to prove was that it did not harm those individuals that received it, the efficacy of OP-1 not only has not been demonstrated for the general population, but also more specifically, it has not been proven in the Medicare population. Medtronic presented arguments that INFUSE[supreg] is a superior product to OP-1 because the INFUSE[supreg] product has demonstrated safety and efficacy, while the OP-1 product has merely demonstrated that it is safe to use in humans. Medtronic pointed to the labeled indications and package inserts provided with the two products, stating that only INFUSE[supreg] provides a substantial clinical improvement to patients receiving a BMP product.

We do not believe that the different types of FDA approvals for the two products are relevant to distinguish between the two products in determining whether either product should be considered for new technology add-on payments under the IPPS. Manufacturers seek different types of FDA approval for many different reasons, including timing, the availability of adequate studies, the availability of resources to pursue research studies, and the size of the patient population that may be affected. The FDA has stated that the HDE approval process was established to address cases involving devices used in the treatment or diagnosis of diseases affecting fewer than 4,000 individuals in the United States per year: ``A device manufacturer's research and development costs could exceed its market returns for diseases or conditions affecting small patient populations. FDA, therefore, developed and published [the regulation establishing the HDE process] to provide an incentive for the development of devices for use in the treatment or diagnosis of diseases affecting these populations.'' (http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfHDE/HDEInformation.cfm ) The fact that two products received different types

of approval does not demonstrate either that they are substantially different for purposes of new technology add-on payments, or that one is new and the other is not. Nor do the different types of FDA approval imply that one product could meet our substantial clinical improvement criterion and the other could not. Neither type of FDA approval requires that products establish substantial clinical improvement over existing technologies, as is required for approval of new technology add-on payments. Theoretically, a product that receives an FDA HDE approval could subsequently meet our substantial clinical improvement criterion, while a product that receives an FDA PMA approval could fail to do so. We base our substantial clinical improvement determinations on the evidence presented in the course of the application process, and not on the type of FDA approval.

For purposes of determining whether the use of rhBMPs for open tibia fracture

[[Page 47348]]

represents a new technology, the crucial consideration is whether the costs of this technology are represented in the weights of the relevant DRGs. Cases that involve treatment of non-healed and acute tibia fractures fall into the same DRGs. We have identified 10,047 cases involving the use of rhBMPs in the FY 2004 MedPAR data file. This use includes the approved indications for INFUSE[supreg] in spinal fusions (6,712 cases) and tibia DRGs (77 cases). However, we note that an additional 3,258 cases involving the off-label use of rhBMPs were found in 47 DRGs in the FY 2004 MedPAR data. We also note that, in our analysis of the FY 2003 MedPAR data, an additional 890 cases of off- label use (identified by the presence of ICD-9-CM code 84.52) were found in 36 DRGs. Therefore, we note that the use of rhBMPs, made by Medtronic or otherwise, has penetrated the cost data that were used to set the FY 2005 and FY 2006 DRG weights. Even if it were possible to differentiate between patients who would be eligible to receive the OP- 1 Implant for nonunions or the INFUSE[supreg] bone graft for open tibia fractures, the patient populations both fall into the same DRGs. In addition, as we stated in last year's final rule in connection with our decision to make add-on payments for both products when used for spinal fusions, we have determined that the costs associated with the two products are comparable (69 FR 49009). Therefore, because BMP products have been used in treating both types of fractures included in the same DRGs since 2001, we continue to believe that the hospital charge data used in developing the relative weights of the relevant DRGs reflect the costs of these products.

Prior to the publication of the FY 2006 IPPS proposed rule, we received the following public comments on the application for add-on payments for FY 2006.

Comment: In our Federal Register announcement of the February 23, 2005 new technology town hall meeting, held on February 23, 2005, we solicited comments on the issue of when products should be considered substantially similar. As a result, Medtronic recommended several criteria for determining whether two or more products are substantially similar and requested that we apply these criteria in determining whether OP-1 and INFUSE[supreg] are similar for new technology add-on payment purposes. The three criteria recommended by Medtronic are:

The technologies or services in question use the same, or a similar, mechanism of action to achieve the therapeutic outcome.

The technologies or services are indicated for use in the same population for the same condition.

The technologies or services achieve the same level of substantial improvement.

Medtronic also argued that, according to its proposed criteria, OP- 1 would fail on two of the three proposed tests for substantial similarity:

According to Medtronic, the OP-1 implant ``arguably'' uses the same or a similar mechanism of action to achieve the therapeutic outcome.

OP-1 and INFUSE[supreg] are indicated for use in different populations and different conditions. According to Medtronic, INFUSE[supreg] Bone Graft has an indication for acute, open tibia fractures only, used within 14 days, and is to be used with an intramedullary (IM) nail as part of the primary procedure. There is no limitation on the number of patients that can receive the technology. OP-1 Implant is indicated only for recalcitrant long-bone non-unions that have failed to heal. The HDE approval also specifies that use of OP-1 is limited to secondary procedures (as would be expected with nonunions). The number of patients able to receive the device is limited to 4,000 patients per year and there is oversight from an Institutional Review Board.

Medtronic argues the products do not achieve the same level of substantial improvement (as discussed above).

Response: We agree with Medtronic that its first proposed criterion has some relevance in determining whether products are substantially similar. In evaluating the application for new technology add-on payments for INFUSE[supreg] for open tibia fractures last year, we made the determination that, while these products are not identical chemically, the products do use the same mechanism of action to achieve the therapeutic outcome. However, we do not agree that the other two criteria recommended by Medtronic should be controlling considerations for this purpose. As we have discussed above, we believe that whether cases involving different products are assigned to the same DRGs is a more relevant consideration than whether the products have the same specific indications. In addition, as we have already stated, we continue to believe that the hospital charge data used in developing the relative weights of the relevant DRGs reflect the costs of both of these products. Furthermore, we do not necessarily agree that considerations about the degrees of clinical improvements offered by different products should enter into decisions about whether products are new. We have always based our decisions about new technology add-on payments on a logical sequence of determinations, moving from the newness criterion to the cost criterion and finally to the substantial clinical improvement criterion. Specifically, we do not make determinations about substantial improvement unless a product has already been determined to be new and to meet the cost criterion. Therefore, we are reluctant to import substantial clinical improvement considerations into the logically prior decision about whether technologies are new. Furthermore, while we may sometimes need to make separate determinations about whether similar products meet the substantial clinical improvement criterion, we do not believe that it would be appropriate to make determinations about whether one product or another is clinically superior.

Comment: In response to our request for comments on the issue of substantial similarity in the Federal Register announcement of the new technology town hall meeting, Medtronic also suggested revisions to the application process that are designed to assist in identifying substantially similar products and provide the public with opportunity for comment on specific instances in which substantial similarity is an issue. The suggested proposed revisions are:

After receipt of all new applications for a fiscal year, CMS should publish a Federal Register notice specifically asking manufacturers to identify if they wish to receive consideration for products that may be substantially similar to applications received. Such notice would probably occur in January. Responses would be required by a date certain in advance of the new technology town hall meeting, and would include justification of how the products meet the ``substantial similarity'' criteria.

The new technology town hall meeting should include a discussion of products identified by manufacturers as ``substantially similar'' to other approved products or pending applications.

CMS should publish initial findings about ``substantial similarity'' in the proposed hospital inpatient rule, with opportunity for public comment.

CMS should publish ultimate findings in the inpatient final rule.

Alternatively, Medtronic suggested that, if a manufacturer identifies a product that may be substantially similar to a technology with an approved add-on payment, the manufacturer may choose to submit an

[[Page 47349]]

application under the normal deadlines for the add-on payment program.

Response: We appreciate Medtronic's suggestions for evaluating similar technologies for new technology add-on payment. We have stated on several occasions that we wish to avoid creating situations in which similar products receive different treatment because only one manufacturer has submitted an application for new technology add-on payments. As we discussed in the new technology final rule (66 FR 46915), an approval of a new technology for special payment should extend to all technologies that are substantially similar. Otherwise, our payment policy would bestow an advantage to the first applicant to receive approval for a particular new technology.

In addition, we note that commenters on the FY 2005 proposed rule placed a great deal of emphasis on the fact that many manufacturers developing new technologies are not aware of the existence of the add- on payment provision or lack the resources to apply for add-on payment. Therefore, commenters on that proposed rule argued that the regulations we have established are already too stringent and cumbersome, especially for small manufacturers to access the new technology add-on payment process. The proposal by Medtronic would place further burden on these small manufacturers, both to know that an application has been made for a similar product and to make representations on a product that may or may not be on the market. Therefore, we are reluctant to adopt a process that places the formal burden on a competitor to seek equal treatment. However, in the FY 2006 IPPS proposed rule, we solicited comments on the use of substantial similarity to determine whether products qualify for new technology add-on payments while we continued to consider these issues. The comments we received in response to this request are addressed below in our discussion of substantial similarity.

We note that, in support of its application for add-on payments for FY 2006, Medtronic submitted data on 236 cases using INFUSE[supreg] for open tibia fractures in the FY 2003 MedPAR data file, as identified by procedure code 79.36 (Reduction, fracture, open, internal fixation, tibia and fibula) and diagnosis codes of either 823.30 (Fracture of tibia alone, shaft, open) or 823.32 (Fracture of fibula and tibia, shaft, open). Medtronic also noted that the patients in clinical trials with malunion fractures (diagnosis code 733.81) or nonunion fractures (diagnosis code 733.82) would also be likely candidates to receive INFUSE[supreg]. Based on the data submitted by the applicant, INFUSE[supreg] would be used primarily in two different DRGs: 218 and 219 (Lower Extremity and Humerus Procedures Except Hip, Foot, Femur Age > 17, With and Without CC, respectively). The analysis performed by the applicant resulted in a case-weighted cost threshold of $24,461 for these DRGs. The average case-weighted standardized charge for cases using INFUSE[supreg] in these DRGs would be $39,537. Therefore, the applicant maintains that INFUSE[supreg] for open tibia fractures meets the cost criterion.

However, because the costs of INFUSE[supreg] and OP-1 are already reflected in the relevant DRGs, these products cannot be considered new. Therefore, in the FY 2006 IPPS proposed rule we proposed to deny new technology add-on payments for INFUSE[supreg] bone graft for open tibia fractures for FY 2006.

During the 60-day comment period on the FY 2006 IPPS proposed rule, we received the following comments on this application:

Comment: Several commenters wrote to support the application for INFUSE[supreg] bone graft for open tibia fractures for new technology add-on payments. These commenters disagreed with our assertion that the costs for this technology are adequately reflected in the DRG weights. The commenters argued that the data include few claims for OP-1 and do not justify denying add-on payments to INFUSE[supreg]. Further, commenters argued that the different types of FDA approval are relevant to the discussion of newness and substantial clinical improvement of the BMP products. Commenters pointed to the limited number of cases that would have been eligible to receive OP-1 due to its limited FDA humanitarian device exemption (HDE) approval. Commenters noted that an HDE approval limits the number of patients that can receive the product to 4,000 patients, and therefore the costs of the cases are not adequately reflected in the DRG weights. According to the commenters, CMS' own analysis supports this point because there were only 77 cases in the FY 2004 MedPAR data, indicating that a patient received a BMP product with no mention as to whether there were any cases in the relevant DRGs for FY 2003. Therefore, commenters argued, the technology is not used frequently enough to be adequately reflected in the DRG weights. In addition, commenters argued that OP-1 is only indicated for non-union fractures while INFUSE[supreg] is for open tibia fractures.

Response: We appreciate the commenters' input on this technology. However, we continue to believe that INFUSE[supreg] is not a new product because of its substantial similarity to OP-1. These products are both designed to promote healing of broken bones even though they are FDA approved for somewhat different indications. Furthermore, treatment of open tibia fractures and non-unions of tibia fractures will be paid using the same DRGs. Because the OP-1 Implant received FDA approval in 2001 and INFUSE[supreg] is a similar product that will be included in the same DRG, we do not believe that the product can be considered new for the purposes of new technology add-on payments. While the commenters argue that the MedPAR data do not include a sufficient number of cases for CMS to argue that payment for BMP products are included in the DRG weights, we do not believe that case volume is a relevant consideration for making the determination as to whether a product is new. Consistent with the statute, a technology no longer qualifies as new once it is more than 2 to 3 years old irrespective of how frequently it has been used in the Medicare population. Thus, if a product is more than 2 to 3 years old, we consider its costs to be included in the DRG relative weights whether its use in the Medicare population has been frequent or infrequent. We also recognize that, without financial incentive to code BMPs, it is possible that hospitals may not have included procedure code 84.52 on hospital bills for all instances when a BMP product was used. Therefore, the incidence of actual use of BMPs for this period may be higher than shown in the Medicare data. Nevertheless, even though hospitals may not have coded all uses of procedure code 84.52, hospital bills would still include charges for all items and services furnished to a Medicare patient including use of a BMP product. Therefore, even though we may be not be able to identify all uses of a BMP product in the Medicare charge data, hospital charges for the DRG would continue to reflect use of these products. In addition, we note that open tibia fractures are not common among the elderly population, and we would therefore not expect to find a high incidence of these cases in the MedPAR data. Also, given the penetration that BMPs have made in DRGs 219 and 220, in addition to many other DRGs, we believe that the BMP technology is adequately reflected in our MedPAR data that were used to recalibrate the DRG weights for FY 2006. Therefore, the

[[Page 47350]]

technology can no longer be considered new for the purposes of new technology add-on payments. In this final rule, we are finalizing our proposal to deny add-on payments for INFUSE[supreg] bone graft for tibia fractures.

Comment: As discussed above, prior to publication of the FY 2006 IPPS proposed rule, we received a comment offering suggestions for how to define when products are ``substantially similar.'' We responded to this comment in the proposed rule (70 FR 23359), and indicated that we welcomed further comments on this issue. Several commenters raised concerns about CMS' responses to this comment.

One commenter indicated that CMS ``is using the determination of `substantial similarity' as a basis to support a preliminary determination that these technologies are `not new' * * * when no such criter[ion] exists in the threshold criteria.'' Another commenter indicated that the discussion of substantial similarity creates confusion between the issue of substantial similarity and the three add-on payment criteria. This commenter indicated that the discussion of this issue in the proposed rule implies that substantial similarity is a subfactor of the newness criterion, while prior rules have implied that it is a subfactor of the substantial clinical improvement criterion or a replacement for all three criteria. To support this point, the commenter stated that the new technology final rule (66 FR 46915) indicates that a substantially similar technology would still be required to submit data showing that the technology was inadequately paid and meets the criterion for being new, thus implying that substantial similarity is a subfactor of the substantial clinical improvement criterion. The commenter referenced the discussion in the FY 2005 IPPS final rule (69 FR 49008-49009) indicating that new technology add-on payments would be extended to OP-1 putty without the submission of an application for add-on payments as evidence that substantial similarity has replaced all three criteria. Commenters further expressed concern over the detrimental effects that this standard could have, denying patient access to therapies ``merely because the therapy has the same mechanism of action as an existing treatment.'' These commenters recommended that CMS eliminate substantial similarity from our new technology add-on payment deliberations, and grant add-on payments based solely on whether a product satisfies the newness, cost, and substantial clinical improvement criteria specified in the regulations.

Other commenters noted that CMS has no way to distinguish between manufacturers when similar products use the same ICD-9-CM codes. Therefore, the commenters argued, there is no need for competitors to apply for their own new technology add-on payment if a product has already been approved for add-on payments, despite the language contained in the new technology final rule stating that the manufacturers of substantially similar products would be required to file a separate application for add-on payment (66 FR 46915).

Response: With respect to the discussion of substantial similarity in the new technology final rule, we did indicate that a manufacturer of a substantially similar product would have to submit an application to be awarded add-on payments. However, we note that this statement was made without any actual experience with the implementation of section 1886(d)(5)(K) of the Act. After reviewing and approving technologies for add-on payment for several years, we have found that our original policy did not adequately reflect the fact that substantially similar products will use the same ICD-9-CM codes and that it would be impractical to create manufacturer-specific codes and also require each manufacturer to submit separate applications for products that are essentially the same. Moreover, given that we cannot distinguish one manufacturer from another when substantially similar technologies use the same ICD-9-CM code, there is no practical purpose for manufacturers of substantially similar products to apply separately for new technology add-on payments. Therefore, we have not required that an application for add-on payments be submitted for a substantially similar product that uses the same ICD-9-CM code as a product that has previously been approved for add-on payments. In addition, we have made an effort to identify competitors that might be eligible to receive new technology add-on payments for their devices. In fact, we note that we have discussed several such technologies in this year's and previous years' rules and have allowed for add-on payments for particular, new classes of technologies that fall within the same ICD-9-CM code (for example, CRT-D).

We believe that these commenters raise interesting and complex policy issues regarding the application of the new technology add-on payment policy to products that are substantially similar. While the commenters generally appear to agree with our policy when we have extended new technology add-on payments to substantially similar products, they appear to disagree with our application of the concept of substantial similarity when we have denied add-on payments. (We note that one commenter disagreed with both the decision to extend new technology add-on payments to OP-1 for spinal fusions and the decision to deny them to INFUSE[supreg] for tibia fractures on the basis of substantial similarity. Nevertheless, this same commenter has also asked us to use the concept of substantial similarity to extend new technology add-on payments to the Talent Endovascular Stent Graft.

This apparent policy contradiction is illustrated with the example of INFUSE[supreg] and OP-1. We extended new technology add-on payments to OP-1 for spinal fusions without a separate application because of its substantial similarity to INFUSE[supreg] and without specifically finding that the product met all three criteria for add-on payments. We determined that OP-1 putty was substantially similar to another product that had been approved for new technology add-on payments. OP-1 putty was clearly new given the date it was approved by the FDA and was substantially similar to another new product that had been approved for new technology add-on payments. However, because the technology of using BMPs for spinal fusions had already been found to meet the newness, cost and substantial clinical improvement criteria, we did not separately address these criteria. Rather, after determining that the two products were substantially similar, we extended the approval of add-on payments to OP-1. The commenters appear to agree with this decision and the concept of extending new technology add-on payments to substantially similar products so that our payment policy does not bestow an advantage to the first applicant representing a particular new technology to receive approval. However, the commenters appear to disagree with our denial of new technology add-on payments to INFUSE[supreg] for tibia fractures on the basis of its substantial similarity to OP-1. Because OP-1 Implant for recalcitrant long bone unions had been in use for 3 years and the costs for this technology had been included in the weights for the DRGs where cases involving INFUSE[supreg] for tibia fractures are assigned, in the final rule for FY 2005, we determined that INFUSE[supreg] could not longer be considered ``new.'' (69 FR 49012).

We believe that the concept of substantial similarity needs to be applied consistently both in the context of extending and denying new

[[Page 47351]]

technology add-on payments. Thus, we believe it is important to clarify whether a finding of substantial similarity among products constitutes only a decision about the newness criterion or about all three criteria. One commenter indicated that our decision to extend new technology add-on payments to OP-1 for spinal fusions because of its similarity to INFUSE[supreg] implies that our determination on substantial similarity replaced consideration of the three criteria. This commenter and others believed, however, that our determination on substantial similarity between OP-1 and INFUSE[supreg] for tibia fractures implies that we are applying the concept as a subfactor of newness.

In both cases, we only made a determination about the similarity of the products and did not specifically make a finding as to whether all three criteria for add-on payments were met. When we denied new technology add-on payments to INFUSE[supreg] for open tibia fractures, we effectively made a logical prior determination about newness based on our finding of substantial similarity and, as a result, we did not need to evaluate either the cost or substantial clinical improvement criteria. Similarly, when we extended new technology add-on payments to OP-1 for spinal fusions on the grounds that it is substantially similar to INFUSE[supreg], we effectively indicated that both products were new but did not make a specific finding about cost and substantial clinical improvement with respect to OP-1. Rather, we extended the existing approval of add-on payments for the new technology of using BMPs in spinal fusions to a substantially similar product in order to avoid bestowing an advantage to the first product to receive an approval of add-on payments for this particular new technology.

We see two policy options to address this issue. Under the first option, we continue our current practice. That is, if we make a finding of substantial similarity among two products, we will extend new technology add-on payment without a further application from the manufacturer of the competing product or a specific finding on cost and clinical improvement. Also, we will deny new technology add-on payments to substantially similar products if one of the products no longer qualifies as a new medical technology without a specific finding on the remaining two criteria. Under the second option, we would depart from our current practice and only extend new technology add-on payment to an applicant's product after making a determination that it meets the newness, cost, and substantial clinical improvement criteria. As we have indicated in the past, we believe that continuing our current practice is the better policy because we avoid:

Creating manufacturer-specific codes for substantially similar products.

Requiring different manufacturers of substantially similar products from having to submit separate new technology applications.

Having to compare the merits of competing technologies on the basis of substantial clinical improvement.

Bestowing an advantage to the first applicant representing a particular new technology to receive approval.

The commenters also argued that the concept of substantial similarity is being applied without having been defined in the regulations. We do not believe that it would be appropriate at this time to adopt rigid criteria to define substantial similarity. Such criteria would restrict unduly our ability to make appropriate determinations regarding whether a product should qualify for new technology add-on payments. For example, if we were to use the Medtronic definition of substantial similarity described above, each manufacturer of a competing technology would have to submit a separate application for an add-on payment and, potentially, we would have to create separate codes for each manufacturer's product if we found that one product met all of the criteria for an add-on payment while the other did not. For instance, Medtronic supported the application of W. L. Gore & Associates, Inc. for its Endovascular Graft Repair of the Thoracic Aorta (GORE TAG). If this device were to be approved for new technology add-on payments, Medtronic recommended that we extend these payments to its Talent Endovascular Stent Graft once it is approved by the FDA. As indicated below, we are approving for the GORE TAG device for new technology add-on payments. If we were to use Medtronic's criteria for defining substantial similarity, for us to extend new technology add-on payments to its device for an endovascular thoracic aortic aneurysm repair, we would have to make a determination that the products: (1) Use the same or a similar mechanism of action to achieve the therapeutic outcome; (2) are indicated for use in the same population for the same condition; and (3) achieve the same level of substantial clinical improvement. While it may be possible to make a determination on the first of these two criteria based on a description of the products and their FDA approved indications, we believe it would not be possible to make a decision on the third criterion without a new technology application and specific review in order to determine whether the two products achieve the same level of substantial clinical improvement. Applying Medtronic's criteria, we do not believe that new technology add-on payments could be extended to a substantially similar product in the middle of a fiscal year. Thus, for example, add-on payments for Medtronic's Talent Endovascular Stent Graft, which has not yet received FDA approval, could not begin until at least FY 2007. Further, in the absence of a finding that the products achieve the same level of substantial clinical improvement, we would need to establish a specific code for the GORE TAG device that other manufacturers of similar products could not use unless they also made a new technology application and we made a finding on the three criteria for determining substantial similarity suggested by Medtronic. Thus, in this circumstance, application of Medtronic's suggested criteria for defining substantial similarity would bestow an advantage to GORE TAG until we could make a specific finding on the Talent Endovascular Stent Graft.

In the proposed rule, we indicated that whether a product uses the same or a similar mechanism of action to achieve the therapeutic outcome has some relevance for determining substantial similarity. We also indicated that the whether the products are assigned to the same or a different DRG is also relevant for determining substantial similarity and assessing if the hospital charge data used in developing the relative weights of the relevant DRGs reflects the costs of these products. In making a determination of substantial similarity, we believe both of these criteria should be met. If only one of the criteria is met, we do not believe the products should be considered substantially similar and new technology add-on payments should not be extended or denied on this basis. In the case of OP-1 and INFUSE[supreg], both are bone morphogenetic products that are used to induce bone growth (``use the same or similar mechanism of action to achieve the therapeutic outcome'') assigned to the same DRGs (DRGs 497 and 498 for spinal fusions and DRGs 218 and 219 for tibia fractures). Furthermore, both of these products can be described by the same ICD-9- CM code (code 84.52, Insertion of recombinant bone morphogenetic protein). Thus, our decisions to extend new technology add-on payments to OP-1 for spinal fusions and deny them to INFUSE[supreg] for tibia fractures on the

[[Page 47352]]

basis of substantial similarity, applied, the two above described criteria consistently.

We believe the above discussion indicates that these are complex issues. While the application of the above two criteria worked well in the context of OP-1 and INFUSE[supreg] (as well as the GORE TAG and Talent Endovascular Stent Graft), it is possible that we should have the flexibility to consider these or other factors in some contexts but not in others. For these reasons, we will continue to analyze the question of substantial similarity, and welcome further public input on this issue.

In this final rule, we are finalizing our proposal to deny add-on payments for INFUSE[supreg] bone graft for open tibia fractures for the reasons discussed above. b. AquadexTMSystem 100 Fluid Removal System (System 100)

CHF Solutions, Inc. resubmitted an application (previously submitted for consideration for FY 2005) for the approval of the System 100 for new technology add-on payments for FY 2006. The System 100 is designed to remove excess fluid (primarily excess water) from patients suffering from severe fluid overload through the process of ultrafiltration. Fluid retention, sometimes to an extreme degree, is a common problem for patients with chronic congestive heart failure. This technology removes excess fluid without causing hemodynamic instability. It also avoids the inherent nephrotoxicity and tachyphylaxis associated with aggressive diuretic therapy, the mainstay of current therapy for fluid overload in congestive heart failure.

The System 100 consists of: (1) An S-100 console; (2) a UF 500 blood circuit; (3) an extended length catheter (ELC); and (4) a catheter extension tubing. The System 100 is designed to monitor the extracorporeal blood circuit and to alert the user to abnormal conditions. Vascular access is established via the peripheral venous system, and up to 4 liters of excess fluid can be removed in an 8-hour period.

On June 3, 2002, FDA approved the System 100 for use with peripheral venous access. On November 20, 2003, FDA approved the System 100 for expanded use with central venous access and catheter extension use for infusion or withdrawal circuit line with other commercially applicable venous catheters. According to the applicant, although the FDA first approved System 100 in June 2002, it was not used by hospitals until August 2002 because of the substantial amount of time necessary to market and sell the device to hospitals. The applicant presented data and evidence demonstrating that the System 100 was not marketed until August 2002.

We note the applicant submitted an application for FY 2005 and was denied new technology add-on payments. Our review indicated that the applicant did not present sufficient objective clinical evidence to determine that the System 100 meets the substantial clinical improvement criterion (such as a large prospective, randomized clinical trial) even though it is indicated for use in patients with congestive heart failure, a common condition in the Medicare population. However, for FY 2006, we proposed to deny System 100 new technology add-on payments on the basis of our determination that it is no longer new. Technology is no longer considered new 2 to 3 years after data reflecting its costs begin to become available. Because data on the costs of the System 100 first became available in 2002, the costs are currently reflected in the DRG weights and the device is no longer new.

The applicant also submitted information for the cost and substantial clinical improvement criteria. As stated last year, it is important to note at the outset of the cost analysis that the console is reusable and is, therefore, a capital cost. Only the circuits and catheters are components that represent operating expenses. Section 1886(d)(5)(K)(i) of the Act requires that the Secretary establish a mechanism to recognize the costs of new medical services or technologies under the payment system established under subsection (d) of section 1886, which establishes the system for paying for the operating costs of inpatient hospital services. The system of payment for capital costs is established under section 1886(g) of the Act, which makes no mention of any add-on payments for a new medical service or technology. Therefore, it is not appropriate to include capital costs in the add-on payments for a new medical service or technology and these costs should also not be considered in evaluating whether a technology meets the cost criterion. The applicant has applied for add- on payments for only the circuits and catheter, which represent the operating expenses of the device. However, as stated in the FY 2005 IPPS final rule, we believe that the catheters cannot be considered new technology for this device. As a result, we considered only the UF 500 disposable blood circuit as relevant to the evaluation of the cost criterion.

The applicant submitted data from the FY 2003 MedPAR file in support of its application for new technology add-on payments for FY 2006. The applicant used a combination of diagnosis codes to determine which cases could potentially use the System 100. The applicant found 28,155 cases with the following combination of ICD-9-CM diagnosis codes: 428.0 through 428.9 (Heart Failure), 402.91 (Unspecified with Heart Failure), or 402.11 (Hypertensive Heart Disease with Heart Failure), in combination with 276.6 (Fluid Overload) and 782.3 (Edema). The 28,155 cases were found among 148 DRGs with 50.1 percent of cases mapped across DRGs 88, 89, 127, 277 and 316. The applicant eliminated those DRGs with less than 150 cases, which resulted in a total of 22,620 cases that could potentially use the System 100. The case- weighted average standardized charge across all DRGs was $13,619.32. The case-weighted threshold across all DRGs was $16,125.42. Although the case-weighted threshold is greater than the case-weighted standardized charge, it is necessary to include the standardized charge for the circuits used in each case. In order to establish the charge per circuit, the applicant submitted data regarding 76 actual cases that used the System 100. Based on these 76 cases, the standardized charge per circuit was $2,591. The applicant also stated that an average of two circuits is used per case. Therefore, adding $5,182 for the charge of the two circuits to the case-weighted average standardized charge of $13,619.32 results in a total case-weighted standardized charge of $18,801.32. This amount is greater than the case-weighed threshold of $16,125.42.

The applicant contended that the System 100 represents a substantial clinical improvement for the following reasons: It removes excess fluid without the use of diuretics; it does not lead to electrolyte imbalance, hemodynamic instability or worsening renal function; it can restore diuretic responsiveness; it does not adversely affect the renin-angiotensin system; it reduces hospital length of stay for the treatment of congestive heart failure, and it requires only peripheral venous access. The applicant also noted that there are some clinical trials that have demonstrated the clinical safety and effectiveness as well as cost effectiveness of the System 100 in treating patients with fluid overload.

However, as stated above, we proposed to deny new technology add-on payments for the System 100 because it does not meet the newness criterion

We received no public comments regarding this application for add- on payments prior to publication of the FY 2006 IPPS proposed rule. During the 60-

[[Page 47353]]

day comment period for the FY 2006 IPPS proposed rule, we also received no comments. Therefore, we are finalizing our proposal to deny new technology add-on payments for the System 100 because it does not meet the newness criterion. c. CHARIT[Eacute]TMArtificial Disc (CHARIT[Eacute]TM)

DePuy SpineTMsubmitted an application for new technology add-on payments for the CHARIT[Eacute]TM Artificial Disc for FY 2006. This device is a prosthetic intervertebral disc. DePuy SpineTMstated that the CHARIT[Eacute]TMArtificial Disc is the first artificial disc approved for use in the United States. It is a 3-piece articulating medical device consisting of a sliding core that is placed between two metal endplates. The sliding core is made from a medical grade plastic and the endplates are made from medical grade cobalt chromium alloy. The endplates support the core and have small teeth that are secured to the vertebrae above and below the disc space. The sliding core fits in between the endplates.

On October 26, 2004, the FDA approved the CHARIT[Eacute]TMArtificial Disc for single level spinal arthroplasty in skeletally mature patients with degenerative disc disease (DDD) between L4 and S1. The FDA further stated that DDD is defined as discogenic back pain with degeneration of the disc confirmed by patient history and radiographic studies. These DDD patients should have no more than 3 mm of spondylolisthesis at an involved level. Patients receiving the CHARIT[Eacute]TMArtificial Disc should have failed at least 6 months of conservative treatment prior to implantation of the CHARIT[Eacute]TMArtificial Disc. Because the device is within the statutory timeframe of 2 to 3 years and data is not yet reflected within the DRGs, we consider the CHARIT[Eacute]TMArtificial Disc to meet the newness criterion.

We note that an ICD-9-CM code was effective October 1, 2004, for IPPS tracking purposes. The code assigned to the CHARIT[Eacute]TMwas 84.65 (Insertion of total spinal disc prosthesis, lumbosacral).

For analysis of the cost criterion, the applicant submitted two sets of data: one that used actual cases and one that used FY 2003 MedPAR cases. The cases using CHARIT[Eacute]TMmap to DRGs 499 and 500. The applicant submitted 68 actual cases from 35 hospitals that used the CHARIT[Eacute]TM. Of these 68 cases, only 3 were Medicare patients; the remaining cases were privately insured patients or patients for whom the payer was unknown. Using data from the 68 actual cases, the average standardized charge was $40,722. The applicant maintained that this figure is well in excess of the thresholds for DRGs 499 and 500 (regardless of a case weighted threshold) of $24,828 and $17,299 respectively. Based on this analysis, the applicant maintained that the CHARIT[Eacute]TMmeets the cost criterion because the average standardized charge exceeds the charge thresholds for DRGs 499 and 500.

In addition, as stated above, the applicant submitted cases from the FY 2003 MedPAR file. The applicant searched the MedPAR file for ICD-9-CM procedure codes 81.06, 81.07, and 81.08 in combination with diagnosis codes 722.10, 722.2, 722.5, 722.52, 722.6, 722.7, 722.73 and 756.12, to identify a patient population that could be eligible for the CHARIT[Eacute]TMArtificial Disc and found a total of 12,680 cases. However, these cases are from the FY 2003 MedPAR file and precede the effective date of ICD-9-CM code 84.65 that is currently used to track the device. Of these 12,680 cases, 55.5 percent were reported in DRG 497, and 44.5 percent were reported in DRG 498. As noted above, cases using the CHARIT[Eacute]TMdevice group to the DRGs for back and neck procedures that exclude spinal fusions (DRGs 499 and 500). However, the applicant argues that the CHARIT[Eacute]TMcould be a substitute for spinal fusion procedures found in DRGs 497 and 498 and, therefore, used cases from these DRGs to evaluate whether the CHARIT[Eacute]TMmeets the cost criterion and to argue that procedures using the technology should be grouped to the spinal fusion DRGs. The average standardized charge per case was $50,098 for DRG 497 and $41,290 for DRG 498. Using revenue codes 272 and 278 from the MedPAR file, the applicant then subtracted the charges for surgical and medical supplies used in connection with spinal fusion procedures, which resulted in a standardized charge of all other charges of $24,333 for DRG 497 and $22,183 for DRG 498. Based on the actual cases above, the applicant then estimated the average standardized charge for surgical and medical supplies per case for the CHARIT[Eacute]TMwas $20,033. The applicant estimated that charges have grown by 15 percent from FY 2003 to FY 2005 and, therefore, deflated the average standardized charge for surgical and medical supplies of the CHARIT[Eacute]TMby 15 percent to $17,420. The applicant then added the average standardized charge for surgical and medical supplies of the CHARIT[Eacute]TMto the standardized charge of the remaining charges for DRG 497 and 498 and also inflated the charges by 15 percent in order to update the data to FY 2005 charge levels. This computation resulted in a case-weighted average standardized charge of $46,256. Although the analysis was completed with DRGs 497 and 498, it is necessary to compare the average standardized charge to the thresholds of DRGs 499 and 500 where these cases are grouped. As a result, the case-weighted threshold was $21,480. Similar to the analysis above, the applicant stated that the case-weighted average standardized charge is greater than the case-weighted threshold and, as a result, the applicant maintained that the CHARIT[Eacute]TMmeets the cost criterion.

Comment: The applicant commissioned two independent consultants to conduct separate data analyses demonstrating with actual cases of CHARIT[Eacute]TMthat the device meets the cost criterion. The consultants found 308 cases using CHARIT[Eacute]TM including 9 Medicare cases. One consultant found 94 cases with average standardized charges of $43,065, and the other consultant found 214 cases with average standardized charges of $45,791. As in the proposed rule, the commenter noted that the average standardized charges per case are well in excess of the threshold for DRG 499.

Response: We appreciate the commenter's submission of additional data in support of its application. Based on these data, it appears that the technology meets the cost criterion.

The applicant also contended that the CHARIT[Eacute]TM represents a substantial clinical improvement over existing technology. Use of the CHARIT[Eacute]TMmay eliminate the need for spinal fusion and the use of autogenous bone, and the applicant stated that, based on the Investigational Device Exemption (IDE) study, ``A Prospective Randomized Multicenter Comparison of Artificial Disc vs. Fusion for Single Level Lumbar Degenerative Disc Disease'' (Blumenthal, S, et al, National American Spine Society 2004 Abstract) that patients who received the CHARIT[Eacute]TMArtificial Disc were discharged from the hospital after an average of 3.7 days compared to 4.2 days in the fusion group. Furthermore, the applicant stated that patients who received the CHARIT[Eacute]TMArtificial Disc had a statistically greater improvement in Oswetry Disability Index scores and Visual Analog Scale Pain scores compared to the fusion group at 6 weeks and 3, 6 and 12 months. The study also showed greater improvement from baseline compared to the fusion group on the Physical Component Score at 3, 6, and 23 months. In addition, the applicant states that patients receiving

[[Page 47354]]

the CHARIT[Eacute]TMArtificial Disc returned to normal activities in half the time, compared to patients who underwent fusion, and at the 2 year follow up, 15 percent of patients who underwent a fusion were dissatisfied with the postoperative improvements compared to 2 percent who received the CHARIT[Eacute]TMArtificial Disc. Also, patients who received the CHARIT[Eacute]TM Artificial Disc returned to work on average of 12.3 weeks after surgery compared to 16.3 weeks after circumferential fusion and 14.4 weeks with Bagby and Kuslich cages. The applicant finally stated that the motion preserving technology of the CHARIT[Eacute]TMArtificial Disc may reduce the risk of increase of degenerative disc disease (DDD). The applicant explained that degeneration of adjacent discs due to increased stress has been strongly associated with spinal fusion utilizing instrumentation. In a follow up of 100 patients (minimum 10 years) who received the CHARIT[Eacute]TMArtificial Disc, the incidence of adjacent level DDD was 2 percent.

In the FY 2006 IPPS proposed rule, we indicated that we were continuing to review the information on whether the CHARIT[Eacute]TMArtificial Disc would appear to represent a substantial clinical improvement over existing technology for certain patient populations. Based on the studies submitted to the FDA and CMS, we remain concerned that the information presented may not definitively substantiate whether the CHARIT[Eacute]TMArtificial Disc is a substantial clinical improvement over spinal fusion. In addition, we are concerned that the cited IDE study enrolled no patients over 60 years of age, which excludes much of the Medicare population. We also are concerned about the prevalence of osteoporosis within the Medicare population, because it is a contraindication for this device. In the FY 2006 IPPS proposed rule, we invited comment on both of these points and on the more general question of whether the device satisfies the substantial clinical improvement criterion.

Despite the issues mentioned above, we noted in the FY 2006 IPPS proposed rule that we were still considering whether it is appropriate to approve new technology add-on payment status for the CHARIT[Eacute]TMArtificial Disc for FY 2006. If approved for add-on payments, hospitals would be reimbursed for up to half of the costs for the device. Because the manufacturer has stated that the cost for the CHARIT[Eacute]TMArtificial Disc would be $11,500, the maximum add-on payment for the device would be $5,750.

We finally noted that the applicant requested a DRG reassignment for cases of the CHARIT[Eacute]TMArtificial Disc from DRGs 499 (Back and Neck Procedures Except Spinal Fusion With CC) and 500 (Back and Neck Procedures Except Spinal Fusion Without CC) to DRGs 497 (Spinal Fusion Except Cervical With CC) and 498 (Spinal Fusion Except Cervical Without CC). The applicant argued that the costs associated with an artificial disc surgery are similar to spinal fusion and inclusion in DRGs 497 and 498 would obviate the need to make a new technology add-on payment. On October 1, 2004, we created new codes for the insertion of spinal disc prostheses (codes 84.60 through 84.69). In the FY 2005 IPPS proposed rule and the final rule, we described the new DRG assignments for these new codes in Table 6B of the Addendum to the rules. We received a number of comments recommending that we change the DRG assignments from DRGs 499 and 500 in MDC 8 to the DRGs for spinal fusion (DRGs 497 and 498). In the FY 2005 IPPS final rule (69 FR 48938), we indicated that DRGs 497 and 498 are limited to spinal fusion procedures. Because the surgery involving the CHARIT[Eacute]TMis not a spinal fusion, we decided not to include this procedure in these DRGs. However, in the FY 2006 IPPS proposed rule, we indicated that we would continue to analyze this issue and solicited public comments on both the new technology application for the CHARIT[Eacute]TMand the DRG assignment for spinal disc prostheses.

We received no public comments regarding this application for new technology add-on payments prior to the publication of the FY 2005 IPPS proposed rule. However, we received the following comments during the 60-day comment period on the proposed rule.

Comment: The applicant noted that on July 15, 2005, two new articles were published in the journal ``Spine.'' \3\ The applicant maintained that the studies demonstrate the following conclusions:

\3\ A. Blumenthal et al., ``A Prospective, Randomized, Multi Center FDA IDE Study of Lumbar Total Disc Replacement with the CHARIT[Eacute]TMArtificial Disc vs. Lumbar Fusion: Part I--Evaluation of Clinical Outcomes.''

B. McAfee et al., ``A Prospective, Randomized, Multi Center FDA IDE Study of Lumbar Total Disc Replacement with the CHARIT[Eacute]TMArtificial Disc vs. Lumbar Fusion: Part II--Evaluation of Radiographic Outcomes and Correlation of Surgical Technique Accuracy with Clinical Outcomes.''

The CHARIT[Eacute]TMobviates the iliac crest bone graft donor site morbidity.

The CHARIT[Eacute]TMpreserves segmental motion in flexion/extension through 24 months post implantation.

The CHARIT[Eacute]TMprovided maintenance of post operative disc height through 24 months compared to anterior interbody fusion; disc space height was maintained in greater than 99 percent of CHARIT[Eacute]TMsubjects through 24 month followup.

The CHARIT[Eacute]TMhas the potential to reduce second surgical procedures for adjacent disc disease by maintaining motion (the manufacturer intends to investigate this).

The CHARIT[Eacute]TMprovides early improvement in pain and function as measured by the Oswestry Disability Index compared to anterior interbody fusion at 6 weeks, 3 months, 6 months, and 12 months.

The CHARIT[Eacute]TMprovides improvement in pain reduction as measured by the Visual Analog Scale compared to anterior interbody fusion at 6 weeks, 3 months, 6 months, and 12 months.

The CHARIT[Eacute]TMprovides improvement in quality of life on the physical component score of the SF-36 outcomes tool at 3 months, 6 months, and 24 months.

CMS requested comments on whether or not the results from the IDE study can be generalized to the Medicare population. The commenter commissioned a consultant to conduct a survey to capture clinical information for the Medicare population 65 years or older and the Medicare population that had been implanted with the CHARIT[Eacute]TM, noting that the under 65 Medicare disabled population represents 14 percent of all Medicare beneficiaries or approximately 5 million people. The consultant found data for 18 Medicare beneficiaries and submitted the following results: Surgeons reported that 94.4 percent of the patients demonstrated improvement in overall outcome, pain, and function after the CHARIT[Eacute]TMhad been implanted. Surgeons also noted the following: 100 percent of the patients reported an improved level of activity; 50 percent of the patients achieved full recovery, the other 50 percent had an improved level of activity compared to their preoperative status; and 100 percent of the surgeons recommended the CHARIT[Eacute]TMfor other Medicare patients who meet the clinical indications. The commenter believed that the above studies and the IDE trial demonstrate that CHARIT[Eacute]TMoffers a substantial clinical improvement over fusion for Medicare beneficiaries.

The commenter also stated that Medicare beneficiaries with disabilities make up 21.8 percent of all discharges in DRGs 496, 497, and 498. It is likely that a significant number of these

[[Page 47355]]

patients could benefit from the CHARIT[Eacute]TM. In response to CMS' concern that CHARIT[Eacute]TMis contraindicated in patients with osteoporosis, the commenter noted that spinal fusion surgery is also not indicated in this patient population. Nevertheless, the commenter noted that the Medicare charge data included nearly 98,000 spinal fusions in FY 2004.

The commenter further stated that, although many patients above the age of 65 do have osteoporosis, implanting surgeons report seeing many patients over the age of 65 who are extremely active and do not have signs of osteoporosis, as validated by a Dexascan.

The commenter also requested that CMS apply the substantial clinical improvement criteria consistently to CHARIT[Eacute]TMand INFUSE[supreg] bone graft for spinal fusions. The commenter noted that in the FY 2004 IPPS final rule (68 FR 45388, August 1, 2003), CMS approved INFUSE[supreg] for new technology add-on payment even though evidence was submitted for a small number of Medicare aged patients treated with the product. CMS acknowledged that there was some positive, though limited, evidence for generalized application for the Medicare population, leading CMS to conclude that based on ``[t]hese results, combined with the benefits of the elimination of the need to harvest bone graft from the iliac crest (and associated complications), INFUSE[supreg] does meet the substantial clinical improvement criteri[on].'' The commenter added that, in addition to eliminating the need for harvesting bone from the iliac crest, the CHARIT[Eacute]TMprovides other significant clinical improvements, including maintaining a more normal range of motion, restoration of disc height, potential to reduce adjacent level disc disease, earlier and sustained improvement in pain and function and earlier return to normal activity and improvement in qualify of life.

Based on the comments above, the commenter noted that the CHARIT[Eacute]TMmeets all the criteria and should be approved for new technology add-on payments.

Response: There have been a number of clinical studies conducted on the CHARIT[Eacute]TM(some of the studies referenced below were also submitted by the applicant). One study showed unsatisfactory long term results. Three studies \4\ \5\ \6\ demonstrated excellent or good results, but did not explicitly compare the surgery to spinal fusion. One study \7\ showed promising short-term results, but had no long-term data and indicated the need for further study. After reviewing all the information supplied by the applicant and in these clinical studies discussed above, CMS acknowledges that the CHARIT[Eacute]TMmay have potential benefit for certain carefully selected Medicare beneficiaries. However, our medical officers could not find sufficient evidence to support a finding that this device meets the criteria for being a substantial clinical improvement. Specifically, we are concerned about the lack of comparative data beyond 24 months in the materials that were submitted for review. While the clinical studies above cited by the manufacturer suggest positive outcomes with the device for up to 24 months, other studies cast doubt on both its short-term and long-term performance, and raise troubling questions regarding longer term adverse outcomes. Specifically, as mentioned above, one study \8\ included 27 patients who received the device between 1989 and 2001. Of these patients, 12 reported some short-term benefit, while 14 others reported no benefit at all. The study found that patients in this study had ``recurrent or persistent back and leg pain [that] was caused mainly by disc degeneration on neighboring levels, hyperlordosis of the operated segment, subsidence and migration.'' In addition, the study indicated that removal of the prosthesis is dangerous, and posterior fusion without removing the prosthesis will give suboptimal results. The study further suggested that the CHARIT[Eacute]TMshould be considered experimental until long term results by unbiased observers can indicate to the orthopedic community if the device is an acceptable orthopedic procedure. We also are concerned about the very low number of Medicare beneficiaries who have received the device (18). In addition, aside from a lack of long-term clinical evidence that demonstrates the effectiveness of the device, we also note significant controversy within the orthopedic and spine surgery community regarding the overall effectiveness and safety of this device regardless of a patient's age, primarily based on the lack of long term data to support its use. Therefore, due to the lack of good evidence of long-term clinical benefit and safety, and because of the degree of controversy surrounding the device within the orthopedic and spine surgery community, we do not believe it meets the criterion for substantial clinical improvement and we are denying the application for new technology add-on payments for FY 2006.

\4\ David TJ. ``Lumbar disc prosthesis; Five years follow-up study on 96 patients [abstract]'' Presented at the 15th Annual Meeting of the North American Spine Society (NASS), New Orleans, LA, 2000.

\5\ Lemaire JP., ``SB Charite III intervertebral disc prosthesis: Biomechanical, clinical, and radiological correlations with a series of 100 cases over a follow-up of more than 10 years'', Rachis [Fr]. 2002;14:271-285, cited in DePuy Spine, Inc. Charit[eacute] Artificial Disc. Technical Monograph. SA01-030-000. JC/AG. Raynham, MA: DePuy; November 2004.

\6\ Caspi I, Levinkopf M, Nerubay J., ``Results of lumbar disk prosthesis after a follow-up period of 48 months'', Israel Medical Association Journal. Volume 5, Issue 1, Pages 9-11, 2003.

\7\ A. Geisler FH, Blumenthal SL, Guyer Rd, et al., ``Neurological complications of lumbar artificial disc replacement and comparison of clinical results with those related to lumbar arthrodesis in the literature: Results of a multicenter, prospective, randomized investigational device exemption study of Charit[eacute] intervertebral disc'', Journal of Neurosurgery (Spine 2) Volume 1 Number 2, Pages 143-154, 2004.

\8\ Van Ooij, Oner, ``Complications of Artificial Disc Replacement.'', Verbout Journal of Spinal Disorders and Techniques, Vol. 16 No. 4, p. 369-383, 2003.

We finally note that we believe we have applied a consistent standard of evidence. While the applicant stated there may be similarities between this device and INFUSE[supreg], as noted above, we believe there are still many unanswered questions regarding CHARIT[Eacute]TM, including the lack of long-term clinical evidence and the overall effectiveness of the device, which preclude us from determining that it meets the substantial clinical improvement criterion.

Comment: One commenter who had the CHARIT[Eacute]TM implanted supported approving the CHARIT[Eacute]TMfor new technology add-on payments. The commenter explained that the device has offered clinical benefits, such as pain relief, that other procedures or surgeries were unable to achieve. Other commenters also supported approval of the CHARIT[Eacute]TM, indicating that the FDA prospective study showed a reduction in length of stay of a half day and patients also returned to normal activities in half of the time of spinal fusion patients.

Response: As noted above, we acknowledge that the CHARIT[Eacute]TMmay have potential benefit for certain carefully selected Medicare beneficiaries. However, we do not believe that one patient's experience is sufficient to show the substantial clinical improvement criterion has been met. Further, while the patient's experience indicates that there may be short-term benefits from receiving treatment with CHARIT[Eacute]TM, we remain

[[Page 47356]]

concerned that the data supplied by the applicant did not demonstrate substantial clinical improvement long term, despite the product being available on the European market since 1987. We are also concerned about the degree of controversy surrounding the device within the orthopedic and spine surgery community. Therefore, we are denying this application for new technology add-on payments because we did not find enough evidence that the product meets the substantial clinical improvement criterion.

Comment: One commenter noted that CMS did not acknowledge that section 1886(d)(5)(K) of the Act states:

``Before establishing any add-on payment * * * with respect to a new technology, the Secretary shall seek to identify one or more diagnosis-related groups associated with such technology, based on similar clinical or anatomical characteristics and the cost of the technology.''

The commenter explained that, in the proposed rule, CMS solicited comment on whether to reassign ICD-9-CM code 84.65 and on the new technology application for the CHARIT[Eacute]TM. The commenter added that, instead of considering these as two distinct issues, CMS should consider a DRG change within the new technology application as mandated by the statute.

Another commenter indicated that the purpose of the new technology add-on program is to provide a cost-based bridge to compensate hospitals for additional costs related to new technology. Consistent with CMS' position not to consider DRG changes until sufficient data became available in MedPAR to support it, the commenter believed it would be premature to reassign spinal disc prostheses to DRGs 497 and 498 until further data become publicly available. The commenter added that DRGs 497 and 498 are well established and any changes to these DRGs, such as including cases of disc prosthesis in these DRGs without more complete data could result in an inappropriate reduction to the weight of these DRGs.

Response: We agree with the comments regarding section 1886(d)(5)(K) of the Act. If a product meets all of the criteria for Medicare to pay for a product as a new technology, there is a clear preference expressed in the statute for us to assign the technology to a DRG based on similar clinical or anatomical characteristics and costs. However, as stated above, we are denying new technology add on payments for CHARIT[Eacute]TMbecause we could not establish that it meets the substantial clinical improvement criterion. Nevertheless, we did evaluate whether to make a DRG change for CHARIT[Eacute]TMoutside of the context of the new technology process. We are providing a full analysis of this issue in section II.B.6.d. of the preamble to this final rule. d. Endovascular Graft Repair of the Thoracic Aorta

Endovascular stent-grafting of the descending thoracic aorta (TA) provides a less invasive alternative to the traditional open surgical approach required for the management of descending thoracic aortic aneurysms. W. L. Gore & Associates, Inc. submitted an application for consideration of its Endovascular Graft Repair of the Thoracic Aorta (GORE TAG) for new technology add-on payments for FY 2006. The GORE TAG device is a tubular stent-graft mounted on a catheter-based delivery system, and it replaces the synthetic graft normally sutured in place during open surgery. The device is identified using ICD-9-CM procedure code 39.79 (Other endovascular repair (of aneurysm) of other vessels). The applicant has requested a unique ICD-9-CM procedure code. (We refer readers to Tables 6A through 6H in the Addendum to this final rule for information regarding ICD-9-CM codes.)

At the time of the initial application, the FDA had not yet approved this technology for general use. Subsequently, however, we were notified that FDA approval was granted on March 23, 2005. Therefore, GORE TAG meets the newness criterion. Although we discussed some of the data submitted with the application for new technology add- on payments, we were unable to include a detailed analysis of cost data and substantial clinical improvement data in the FY 2006 IPPS proposed rule because FDA approval occurred too late for us to conduct a complete analysis.

The applicant submitted cost threshold information for the GORE TAG device. According to the manufacturer, cases using the GORE TAG device would fall into DRGs 110 and 111 (Major Cardiovascular Procedures With and Without CC, respectively). The applicant identified 185 cases in the FY 2003 MedPAR using procedure code 39.79 (Other endovascular repair (of aneurysm) of other vessels) and primary diagnosis codes 441.2 (Thoracic aneurysm, without mention of rupture), 441.1 (Thoracic aneurysm, ruptured), or 441.01 (Dissection of aorta, thoracic). The case-weighted standardized charge for 177 of these cases was $60,905. According to the manufacturer, the case-weighted cost threshold for these DRGs is $49,817. Based on this analysis, the manufacturer maintained that the technology meets our cost threshold.

The manufacturer argued that the GORE TAG represents a substantial clinical improvement over existing technology, primarily by avoiding the traditional open aneurysm repair procedure with its associated high morbidity and mortality. The applicant argued that a descending thoracic aorta aneurysm is a potentially life threatening condition that currently requires a major operative procedure for its treatment. The mortality and complication rates associated with this surgery are very high, and the surgery is frequently performed under urgent or emergent conditions. The applicant noted that such complications can increase the length of the hospital stay and can include neurological damage, paralysis, renal failure, pulmonary emboli, hemorrhage, and sepsis. The average time for patients undergoing surgical repair to return to normal activity is 3 to 4 months, but can be significantly longer.

In comparison, the applicant argued that endovascular stent- grafting done with the GORE TAG thoracic endoprosthesis is minimally invasive. The manufacturer noted that patients treated with the endovascular technique experience far less aneurysm-related mortality and morbidity, compared to those patients that receive the open procedure, resulting in reduced overall length-of-stay, less intensive care unit days and less operative complications.

We received the following public comments, in accordance with section 503(b)(2) of Pub. L. 108-173, regarding this application for add-on payments prior to publication of the FY 2006 IPPS proposed rule.

Comment: Several commenters expressed support for approval of new technology add-on payments for the GORE TAG device. These commenters noted that the data presented to the FDA advisory panel for consideration for FDA approval of the device clearly demonstrate the safety and efficacy of the GORE TAG device. They also noted that nearly 200 patients have been treated with the endografts, with a highly significant difference in both postoperative mortality and a reduction in the incidence of spinal cord ischemic complications, with some commenters noting the trial results, which showed a reduction in the rate of paraplegia from 14 percent to 3 percent, compared to open surgery. The commenters also stressed the rigorous nature of the open surgery, which requires a left lateral

[[Page 47357]]

thoracotomy, resulting in significant morbidity. The commenters further argued that, since many of the patients with degenerative aneurysm of the thoracic aorta are elderly or present with significant comorbidities, or both, it is ``a common circumstance in clinical practice to deny repair to such patients because of the magnitude of the conventional open surgery.'' Other commenters stated that the 5- year mortality in all patients diagnosed with thoracic aortic aneurysm is as high as 80 percent in some groups of patients. Therefore, the commenters argued, the GORE TAG device for thoracic aortic aneurysm satisfies the criteria for substantial clinical improvement.

Response: We appreciate the commenters' input on this criterion. In the FY 2006 proposed rule, we indicated that we would consider these comments regarding the substantial clinical improvement criterion in the final rule if we determined that the technology meets the other two criteria.

Comment: A representative of another device manufacturer stated at the town hall meeting that the manufacturer has a similar product awaiting FDA approval.

Response: In the proposed rule, we responded that as we discussed in the new technology final rule (66 FR 46915), an approval of a new technology for special payment should extend to all technologies that are substantially similar. Otherwise, our payment policy would bestow an advantage to the first applicant to receive approval for a particular new technology. In this case, we will determine whether the GORE TAG device qualifies for new technology add-on payments in this final rule. In the event that this technology satisfies all the criteria, as we indicated in the FY 2006 IPPS proposed rule, we would extend new technology payments to any substantially similar technology that also receives FDA approval prior to publication of the FY 2006 final rule. In the FY 2006 IPPS proposed rule, we solicited comments regarding this technology in light of its recent FDA approval, particularly with regard to the cost threshold and the substantial clinical improvement criteria.

During the 60-day comment period for the FY 2006 IPPS proposed rule, we received the following comments:

Comment: The applicant submitted an additional validation sample of cases to confirm the costs associated with this technology. In this sample, charges for the device ranged from approximately $7,000.00 to $11,000.00 per device.

Response: We have reviewed the evidence presented above and have determined that the manufacturer has demonstrated that this device meets the cost threshold for the DRGs to which these cases will be assigned. However, we note that we would expect there to be significantly fewer hospital resources required to care for a patient undergoing the endovascular procedure compared to an open thoracotomy. Thus we are concerned that the cost of cases using this device is unnecessarily high. We will continue to monitor the data associated with the endovascular repair of a thoracic aortic aneurysm in the future to obtain further information about this issue.

Comment: Several commenters encouraged CMS to approve the GORE TAG device for new technology add-on payment approval. These commenters indicated that this device is a significant advance in the treatment of thoracic aortic aneurysms, particularly for elderly, frail patients who are not candidates for the open procedure to correct life-threatening aneurysms. They added that physicians pointed to the mortality and comorbidity rates associated with the open procedure, stating ``even in centers of excellence, the risk of either mortality or paraplegia complicating surgery runs up to the 10 percent range.''

Response: We appreciate the commenters' input on the substantial clinical improvement criterion, and we have determined that the GORE TAG device meets the substantial clinical improvement criterion. In our view, the GORE TAG device meets a number of the standards that we use to evaluate whether a new technology is a substantial clinical improvement. For instance, GORE TAG offers a treatment option for patients with thoracic aortic aneurysms that are not candidates for open surgery. Prior to endovascular treatment with this device, there were no treatment options available for patients who were not candidates for open repair of a thoracic aortic aneurysm. We also believe that, relative to the open repair procedure, endovascular aneurysm repair improves clinical outcomes through lower mortality and complication rates, reduced overall length-of-stay, less intensive care unit days and less operative complications. For the reasons stated above, we find that the GORE TAG device meets the substantial clinical improvement criterion.

As indicated earlier, GORE TAG meets both the newness and cost criteria. Therefore, in this final rule, we are approving the GORE TAG device for new technology add-on payment for FY 2006. These cases generally are in DRGs 110 and 111. Cases involving the device should code for the device using the newly created ICD-9-CM procedure code 39.73 (Endovascular implantation of graft in thoracic aorta). The cost of a single device is $12,798. Because the average patient receives 1.8 endovascular prostheses, we estimate the cost of the device to be $21,198 per patient. Therefore, beginning October 1, 2005, cases that include code 39.73 will be eligible to receive new technology add-on payments up to $10,599, or half the cost of the device.

Comment: In the proposed rule, we stated that ``we would extend new technology payments to any substantially similar technology that also receives FDA approval prior to publication of the FY 2006 final rule.'' Commenters argued that, CMS should not require, FDA approval to be granted to substantially similar devices prior to the publication of the final rule for CMS to extend new technology payments to these products.

Response: We agree with the commenters. Any substantially similar device that is FDA-approved after the publication of the final rule that uses the same ICD-9-CM procedure code as GORE TAG and falls into the same DRGs as those approved for new technology add-on payments should also receive the new technology add-on payment associated with this technology in FY 2006. The discussion of this issue in the preamble to the proposed rule was intended to communicate that we would extend new technology payments to any substantially similar product that is assigned to the same ICD-9-CM code, as long as the applicant's product received FDA approval prior to the final rule. For the reason stated above, we have changed our position on this issue and will extend add-on payment to any substantially similar products that are assigned to the same ICD-9-CM code and that receive FDA approval either before or during FY 2006. e. Restore[supreg] Rechargeable Implantable Neurostimulator

Medtronic Neurological submitted an application for new technology add-on payments for its Restore[supreg] Rechargeable Implantable Neurostimulator. The Restore[supreg] Rechargeable Implantable Neurostimulator is designed to deliver electrical stimulation to the spinal cord for treatment of chronic, intractable pain.

Neurostimulation is designed to deliver electrical stimulation to the spinal cord to block the sensation of pain. The current technology standard for neurostimulators utilizes internal sealed batteries as the power source to

[[Page 47358]]

generate the electrical current. These internal batteries have finite lives, and require replacement when their power has been completely discharged. According to the manufacturer, the Restore[supreg] Rechargeable Implantable Neurostimulator ``represents the next generation of neurostimulator technology, allowing the physician to set the voltage parameters in such a way that fully meets the patient's requirements to achieve adequate pain relief without fear of premature depletion of the battery.'' The applicant stated that the expected life of the Restore[supreg] rechargeable battery is 9 years, compared to an average life of 3 years for conventional neurostimulator batteries. The applicant stated that this represents a significant clinical improvement because patients can use any power settings that are necessary to achieve pain relief with less concern for battery depletion and subsequent battery replacement.

At the time of the FY 2006 proposed rule, this device had not yet received approval for use by the FDA; however, another manufacturer had received approval for a similar device. (Advanced Bionics' Precision[supreg] Rechargeable Neurostimulator was approved by the FDA on April 27, 2004.)

Medtronic Neurological also provided data to determine whether the Restore[supreg] Rechargeable Implantable Neurostimulator meets the cost criterion. Medtronic Neurological stated that the cases involving use of the device would primarily fall into DRGs 499, 500, 531 and 532, which have a case-weighted threshold of $24,090. The manufacturer stated that the anticipated average standardized charge per case involving the Restore[supreg] technology is $59,265. The manufacturer derived this estimate by identifying cases in the FY 2003 MedPAR that reported procedure code 03.93 (Insertion or replacement of spinal nerostimulators). The manufacturer then added the total cost of the Restore[supreg] Rechargeable Implantable Neurostimulator to the average standardized charges for those cases. Of the applicable charges for the Restore[supreg] Rechargeable Implantable Neurostimulator, only the components that the applicant identified as new would be eligible for new technology add-on payments. Medtronic Neurological submitted information that distinguished the old and new components of the device and submitted data indicating that the neurostimulator itself is $17,995 and the patient recharger, antenna, and belt are $3,140. Thus, the total cost for new components would be $21,135, with a maximum add- on amount of $10,568 if the product were to be approved for new technology payments.

We note that we reviewed a technology for add-on payments for FY 2003 called RenewTMRadio Frequency Spinal Cord Stimulation (SCS) Therapy, made by Advanced Neuromodulation Systems (ANS). In the FY 2003 final rule, we discussed and subsequently denied an application for new technology add-on payment for RenewTMSCS because ``RenewTMSCS was introduced in July 1999 as a device for the treatment of chronic intractable pain of the trunk and limbs'' and could no longer be considered a new product (67 FR 50019). We also noted, ``[t]his system only requires one surgical placement and does not require additional surgeries to replace batteries as do other internal SCS systems.''

The applicant also stated in its application for Restore[supreg] that cases where it is used will be identified by ICD-9-CM procedure code 03.93 (Insertion or replacement of spinal neurostimulators), and this code was also used to identify the predecessor technology in order to perform the cost threshold analysis. As we discussed in the FY 2003 final rule (67 FR 50019), the RenewTMSCS is identified by the same ICD-9-CM procedure code. As discussed in the proposed rule, the applicant applied for and was assigned a new ICD-9-CM code for rechargeable neurostimulator pulse generator. (We refer readers to Tables 6A through 6H in the Addendum to this final rule for information regarding ICD-9-CM codes.) Because the RenewTMSCS and Restore[supreg] technologies appear similar, we asked Medtronic to provide information that would demonstrate how the products were substantially different. The applicant noted that the RenewTMSCS, while programmable and rechargeable, is not a good option for those patients who have high energy requirements because of chronic intractable pain that will result in more battery wear and subsequent surgery to replace the device. Both systems rely on rechargeable batteries, and in the case of RenewTMSCS the energy is transmitted through the skin from a radiofrequency source for the purpose of recharging. Medtronic contends that the Restore[supreg] device is superior to the RenewTMdevice because RenewTMrequires an external component that uses a skin adhesive that is uncomfortable and inconvenient (causes skin irritation, is affected by moisture that will come from bathing, sweating, swimming, etc.), leading to patient noncompliance.

Because FDA approval had not yet been received for this device, in the proposed rule, we indicated that we were making no decision concerning the Restore[supreg] application. We indicated that we would make a formal determination if FDA approval occurs in sufficient time for full consideration in this final FY 2006 rule. However, we noted that we had reservations about whether this technology is new for purposes of the new technology add-on payments because of its similarity to other products that are also used to treat the same conditions. Although we recognized the benefits of a more easily rechargeable neurostimulator system, we believed that the Restore[supreg] device might not be sufficiently different from predecessor devices to meet the newness criterion for the new technology add-on payment. As we discussed above, similar products have been on the market since 1999. Therefore, these technologies are already represented in the DRG weights and are not considered new for the purposes of the new technology add-on payment provision. We received no public comments regarding this application for add-on payments prior to the publication of the FY 2006 IPPS proposed rule. In the proposed rule, we solicited comments on this application for add-on payments, specifically regarding how the Restore[supreg] device may or may not be significantly different from previous devices. We also sought comments on whether the product meets the cost and significant improvement criteria.

During the 60-day comment period for the proposed rule, we received the following comments:

Comment: Several commenters supported the application for the rechargeable implantable neurostimulator for add-on payment. Commenters noted that there is a large difference between the radio frequency (RF) devices and the rechargeable implantable neurostimulators. They argued that there is little relief with the RF systems, because once the transmitter/power source is removed, the therapy immediately ends. Further, commenters argued that due to these restrictions and the difficulty of ensuring patient compliance with this device, the pain relief the RF system is intended to provide is not possible. As such, the commenters argued that the rechargeable implantable device is a much better option for many patients with high power needs than previously available neurostimulators.

Commenters argued that the new, rechargeable, implantable neurostimulators meet the substantial clinical improvement criterion by eliminating surgeries to replace the

[[Page 47359]]

batteries, reducing the infection rate associated with greater frequency of replacement surgeries, and providing more treatment options for those patients that require high energy stimulation. In addition, commenters noted the clinical improvement associated with the ability to use two 16-electrode leads instead of the 8-channel leads that are used in older neurostimulators. They pointed out that, by using leads with more electrodes, the physician can place the leads so that more coverage is provided to the spinal nerves, and the physician is provided an option to reprogram the neurostimulator without further invasive surgery if a lead migrates after the unit is installed. Further, commenters argued that, by paying the higher up-front expenses associated with these technologies, CMS will ultimately save money on reduced surgical and physician encounters, while improving the care that Medicare beneficiaries receive. Finally, the manufacturer submitted an updated price for the Restore[supreg] rechargeable implantable neurostimulator that reflects a decrease in total costs for the new components associated with the device. Based on this change, the manufacturer calculated the new maximum add-on payment amount to be $9,320 if the application is approved.

Response: We appreciate these commenters' input regarding this device. While the comments were submitted in support of a finding that this device meets the substantial clinical improvement criterion, they have also convinced us that the device is significantly different from predecessor devices. Therefore, we are reversing our preliminary determination that the Restore[supreg] Rechargeable Implantable Neurostimulator is likely not new, and we have determined that it can be considered new for the purposes of the new technology add-on payment for this reason. The manufacturer also provided data from its device registry demonstrating that nearly 34 percent of patients aged 65 and older, who receive non-rechargeable devices, require a replacement surgery within the first 10 years of implantation. In addition, of those patients that require replacement surgeries, more than half of those patients have high energy needs that deplete the battery within the first 3 years. By avoiding the need for a battery replacement surgery, we believe these data demonstrate that this device is a substantial clinical improvement for a large proportion of the patients who receive implantable neurostimulators. In addition, we agree that the patient compliance issues with the predecessor devices that use of RF as the recharging source are significant. The applicant has demonstrated that there will not be the same patient compliance issues with its product. Because of the elimination of the need for serial battery replacement surgeries and in light of the information provided by the manufacturer and commenters further clarifying the distinctions and improvements of the Restore[supreg] technology when compared to other devices, we believe that the device is a substantial clinical improvement over prior technologies.

As stated in the proposed rule, we had previously determined that Restore[supreg] in combination with the other devices that already received FDA approval in 2004 and 2005, meets the newness and cost threshold criteria. Therefore, we are approving new technology add-on payments for rechargeable, implantable neurostimulators for FY 2006. Cases involving these devices will be identified by the presence of newly created ICD-9-CM code 86.98 (Insertion or replacement of dual array rechargeable neurostimulator pulse generator). These cases are generally included in the following DRGs: 7, 8, 499, 500, 531, or 532. In the proposed rule, we stated that the maximum add-on payment for the new components of the device would be $10,568, or half of $21,135. The applicant reported a reduction in the price to $18,640 after publication of the proposed rule, making the maximum add-on payment for the device $9,320. Therefore, we are finalizing a maximum add-on payment of $9,320 for cases that involve this technology. f. Safe-Cross(r) Radio Frequency Total Occlusion Crossing System (Safe- Cross[supreg])

Intraluminal Therapeutics submitted an application for the Safe Cross[supreg] Radio Frequency (RF) Total Occlusion Crossing System. This device performs the function of a guidewire during percutaneous transluminal angioplasty of chronic total occlusions of peripheral and coronary arteries. Using fiberoptic guidance and radiofrequency ablation, it is able to cross lesions where a standard guidewire is unsuccessful. On November 21, 2003, the FDA approved the Safe Cross[supreg] for use in iliac and superficial femoral arteries. In January 2004, the FDA approved the Safe Cross[supreg] for coronary arteries. The device was also approved by the FDA for all native peripheral arteries except carotids in August 2004. Because the device is within the statutory timeframe of 2 to 3 years for all approved uses and data regarding the cost of this device are not yet reflected within the DRG weights, we consider the Safe Cross[supreg] to meet the newness criterion.

We note that the applicant submitted an application for a distinctive ICD-9-CM code. The applicant noted in its application that the device is currently coded with ICD-9-CM procedure codes 36.09 (Other removal of coronary artery obstruction) and 39.50 (Angioplasty or atherectomy of other noncoronary vessels).

As we stated in last year's final rule, section 1886(d)(5)(K)(i) of the Act requires that the Secretary establish a mechanism to recognize the costs of new medical services or technologies under the payment system established under subsection (d) of section 1886, which establishes the system for paying for the operating costs of inpatient hospital services. The system of payment for capital costs is established under section 1886(g) of the Act, which makes no mention of any add-on payments for a new medical service or technology. Therefore, it is not appropriate to include capital costs in the add-on payments for a new medical service or technology, and these costs should not be considered in evaluating whether a technology meets the cost criterion. As a result, we consider only the Safe Cross[supreg] crossing wire, ground pad, and accessories to be operating equipment that is relevant to the evaluation of the cost criterion.

The applicant submitted the following two analyses on the cost criterion. The first analysis contained 27 actual cases from two hospitals. Of these 27 cases, 25.1 percent of the cases were reported in DRGs 24 (Seizure and Headache Age >17 With CC), 107 (Coronary Bypass With Cardiac Catheterization), 125 (Circulatory Disorders Except AMI, With Cardiac Catheterization and Without Complex Diagnosis), 518 (Percutaneous Cardiovascular Procedure Without Coronary Artery Stent or AMI), and 526 (Percutaneous Cardiovascular Procedure With Drug-Eluting Stent With AMI); and 74.9 percent were reported in DRG 527 (Percutaneous Cardiovascular Procedure With Drug-Eluting Stent Without AMI). This resulted in a case-weighted threshold of $37,304 and a case- weighted average standardized charge of $40,705. (We have updated the case weighted threshold and case weighted average standardized charge from the proposed rule due to an inadvertent clerical error in reporting these figures in the proposed rule.) Because the case- weighted average standardized charge is greater than the case-weighted threshold, the applicant maintained that the Safe Cross[supreg] meets the cost criterion.

[[Page 47360]]

The applicant also submitted cases from the FY 2003 MedPAR. The applicant found a total of 1,274,535 cases that could be eligible for the Safe Cross[supreg] using diagnosis codes 411 through 411.89 (Other acute and subacute forms of ischemic heart disease) or 414 through 414.19 (Other forms of chronic ischemic heart disease) in combination with any of the following procedure codes: 36.01 (Single vessel percutaneous transluminal coronary angioplasty (PTCA) or coronary atherectomy without mention of thrombolytic agent), 36.02 (Single vessel PTCA or coronary atherectomy with mention of thrombolytic agent), 36.05 (Multiple vessel PTCA or coronary atherectomy performed during the same operation with or without mention of thrombolytic agent), 36.06 (Insertion of nondrug-eluting coronary artery stent(s)), 36.07 (Insertion of drug-eluting coronary artery stent(s)) and 36.09 (Other removal of coronary artery obstruction). A total of 59.40 percent of these cases fell into DRG 517 (Percutaneous Cardiovascular Procedure With Nondrug-Eluting Stent Without AMI), 16.4 percent of cases into DRG 516 (Percutaneous Cardiovascular Procedure With AMI), and 16.2 percent of cases into DRG 527, while the rest of the cases fell into the remaining DRGs 124, 518, and 526. The average case- weighted standardized charge per case was $40,318. This amount included an extra $6,000 for the charges related to the Safe Cross[supreg]. The case-weighed threshold across the DRGs mentioned above was $35,955. Similar to the analysis above, because the case-weighted average standardized charge is greater than the case-weighted threshold, the applicant maintained that the Safe Cross[supreg] meets the cost criterion.

The applicant maintained that the device meets the substantial clinical improvement criterion. The applicant explained that many traditional guidewires fail to cross a total arterial occlusion due to difficulty in navigating the vessel and to the fibrotic nature of the obstructing plaque. By using fiberoptic guidance and radiofrequency ablation, the Safe Cross[supreg] succeeds where standard guidewires fail. The applicant further maintained that in clinical trials where traditional guidewires failed, the Safe Cross[supreg] succeeded in 54 percent of cases of coronary artery chronic total occlusions (CTOs), and in 76 percent of cases of peripheral artery CTOs.

However, in the FY 2006 IPPS proposed rule, we noted that we use similar standards to evaluate substantial clinical improvement in the IPPS and OPPS. The IPPS regulations provide that technology may be approved for add-on payments when it ``represents an advance in medical technology that substantially improves, relative to technologies previously available, the diagnosis or treatment of Medicare beneficiaries'' (66 FR 46912). Under the OPPS, the standard for approval of new devices is ``a substantial improvement in medical benefits for Medicare beneficiaries compared to the benefits obtained by devices in previously established (that is, existing or previously existing) categories or other available treatments'' (67 FR 66782). Furthermore, the OPPS and IPPS employ identical language (for IPPS, see 66 FR 46914, and for OPPS, see 67 FR 66782) to explain and elaborate on the kinds of considerations that are taken into account in determining whether a new technology represents substantial improvement. In both systems, we employ the following kinds of considerations in evaluating particular requests for special payment for new technology:

The device offers a treatment option for a patient population unresponsive to, or ineligible for, currently available treatments.

The device offers the ability to diagnose a medical condition in a patient population where that medical condition is currently undetectable or offers the ability to diagnose a medical condition earlier in a patient population than allowed by currently available methods. There must also be evidence that use of the device to make a diagnosis affects the management of the patient.

Use of the device significantly improves clinical outcomes for a patient population as compared to currently available treatments. Some examples of outcomes that are frequently evaluated in studies of medical devices are the following:

--Reduced mortality rate with use of the device. --Reduced rate of device-related complications. --Decreased rate of subsequent diagnostic or therapeutic interventions (for example, due to reduced rate of recurrence of the disease process). --Decreased number of future hospitalizations or physician visits. --More rapid beneficial resolution of the disease process treatment because of the use of the device. --Decreased pain, bleeding, or other quantifiable symptom. --Reduced recovery time.

In a letter to the applicant dated October 25, 2004, we denied approval of the Safe Cross[supreg] for pass-through payments for the OPPS on the basis that the technology did not meet the substantial clinical improvement criterion. In particular, we found that studies failed to show long-term or intermediate-term results, and the device had a relatively low rate of successfully opening occlusions. Since that initial determination, the applicant has requested reconsideration for pass-through payments under the IPPS. However, on the basis of the original findings under the OPPS, we do not now believe that the technology can qualify for new technology add-on payments under the IPPS. Therefore, in the FY 2006 IPPS proposed rule, we proposed to deny new technology add-on payment for FY 2006 for Safe Cross[supreg] on the grounds that it does not appear to be a substantial clinical improvement over existing technologies. We sought further information on whether this device meets the substantial clinical improvement criterion, and indicated that we would consider any further information prior to making our final determination in this final rule.

We received no public comments regarding this application for add- on payments prior to the publication of the FY 2006 IPPS proposed rule. During the 60-day comment period on the FY 2006 IPPS proposed rule, we received the following comment:

Comment: One commenter expressed support for the Safe Cross[supreg], explaining that the increased chance of crossing a CTO enables the placement of drug-eluting stents and represents a substantial clinical improvement for treating the most challenging clinical subgroup with these conditions. Using the device also raises the cost per case and, therefore, the commenter recommended that CMS pay new technology add-on payments for this device.

Response: In a letter dated June 3, 2003 to the applicant, CMS denied pass-through payments under the OPPS for the Safe Cross[supreg] because it did not demonstrate a substantial clinical improvement. The letter explained that the company has not yet provided intermediate to long-term results regarding reocclusion of previously occluded vessels after angioplasty with substantially improved patient outcomes, which could demonstrate that the Safe Cross[supreg] technology leads to significant clinical improvement for patients in comparison with other available treatments. Given the similar criteria for making pass- through payments under the OPPS and new technology add-on payments under the IPPS, a finding that Safe Cross[supreg] does not

[[Page 47361]]

meet the OPPS criteria means that, in the absence of relevant new information, it cannot qualify for new technology add-on payments under the IPPS. Therefore, we are finalizing our proposal to deny new technology add-on payments for the Safe Cross[supreg] in FY 2006 because it does not meet the substantial clinical improvement criterion. g. Trident[supreg] Ceramic Acetabular System

Stryker Orthopaedics submitted an application for new technology add-on payments for the Trident[supreg] Ceramic Acetabular System. This system is used to replace the ``ball and socket'' joint of a hip when a total hip replacement is performed for patients suffering from arthritis or related conditions. The applicant stated that, unlike conventional hip replacement systems, the Trident[supreg] system utilizes alumina ceramic-on-ceramic bearing surfaces rather than metal- on-plastic or metal-on-metal. Alumina ceramic is the hardest material next to diamond. The Trident[supreg] System is a patented design that captures the ceramic insert in a titanium sleeve. This design increases the strength of the ceramic insert by 50 percent over other designs. The manufacturer stated that the alumina ceramic bearing of the device is a substantial clinical improvement because it is extremely hard and scratch resistant, has a low coefficient of friction and excellent wear resistance, has improved lubrication over metal or polyethylene, has no potential for metal ion release, and has less alumina particle debris. The manufacturer also stated that fewer hip revisions are needed when this product is used (2.7 percent of ceramic versus 7.5 percent for polyethylene). Stryker stated that the ceramic implant also causes less osteolysis (or bone loss from particulate debris). Due to these improvements over traditional hip implants, the manufacturer stated the Trident[supreg] Ceramic Acetabular System has demonstrated significantly lower wear versus the conventional plastic/metal system in the laboratory; therefore, it is anticipated that these improved wear characteristics will extend the life of the implant.

In addition, we note that the Trident[supreg] Ceramic Acetabular System received FDA approval in February 3, 2003. However, this product was not available on the market until April 2003. The period that technologies are eligible to receive new technology add-on payment is no less than 2 years but not more than 3 years from the point the product comes on the market. At this point, we begin to collect charges reflecting the cost of the device in the MedPAR data. Because the device became available on the market in April 2003, charges reflecting the cost of the device may have been included in the data used to calculate the DRG weights in FY 2005 and the final DRG weights for FY 2006. Therefore, the technology may no longer be considered new for the purposes of new technology add-on payments. For this reason, in the FY 2006 IPPS proposed rule, we proposed to deny add-on payments for the Trident[supreg] Ceramic Acetabular System for FY 2006.

The applicant submitted cost threshold information for the Trident[supreg] Ceramic Acetabular System, stating that cases using the system would be included in DRG 209 (Major Joint and Limb Reattachment Procedures of Lower Extremity). The manufacturer indicated that there is not an ICD-9-CM code specific to ceramic hip arthroplasty, but it is currently reported using code 81.51 (Total hip replacement). Of the applicable charges for the Trident[supreg] Ceramic Acetabular System, only the components that the applicant identified as new would be eligible for new technology add-on payments. The estimated cost of the new portions of the device, according to the information provided in the application, is $6,009. The charge threshold for DRG 209 is $34,195. The data submitted by Stryker Orthopaedics showed an average standardized charge, assuming a 28 percent implant markup, of $34,230.

Regarding the issue of substantial clinical improvement, we recognize that the Trident[supreg] Ceramic Acetabular System represents an incremental advance in prosthetic hip technology. However, we also recognize that there are a number of other new prostheses available that utilize a variety of bearing surface materials that also offer increased longevity and decreased wear. For this reason, we do not believe that the Trident[supreg] system has demonstrated itself to be a clearly superior new technology.

We received the following public comments, in accordance with section 503(b)(2) of Pub. L. 108-173, regarding this application for add-on payments prior to publication of the FY 2006 IPPS proposed rule.

Comment: One commenter noted that clinical outcomes for the Trident[supreg] Ceramic Acetabular System are not a significant clinical improvement over similar devices on the market. A member of the orthopedic community noted at the new technology town hall meeting that this system is not the only new product that promises significantly improved results because of enhancements to materials and design. This commenter suggested that it may be inappropriate to recognize only one of these new hip replacement products for new technology add-on payments.

Response: We appreciate the commenter's input on this criterion. In the proposed rule, we indicated that we would consider these comments regarding the substantial clinical improvement criterion. However, based on the observations provided at the town hall meeting, we noted that we are considering alternative methods of recognizing technological improvements in this area other than approving only one of these new technologies for add-on payments. For example, as discussed in section II.B.6.a. of the preamble to the proposed rule, we proposed to split DRG 209 to create a new DRG for revisions of hip and knee replacements. We would leave all other replacements and attachment procedures in a separate, new DRG. We also stated that we would review these DRGs based on new procedure codes that will provide more detailed data on the specific nature of the revision procedures performed. In addition, we are creating new procedure codes that will identify the type of bearing surface of a hip replacement. As we obtain data from these new codes, we stated that we would consider additional DRG revisions to better capture the various types of joint procedures. We also stated that we may consider a future restructuring of the joint replacement and revision DRGs that would better capture the higher costs of products that offer greater durability, extended life, and improved outcomes. In doing so, of course, we may need to create additional, more precise ICD-9-CM codes. In the FY 2006 IPPS proposed rule, we sought comments on this issue, and generally on whether the Trident[supreg] Ceramic Acetabular System meets the criteria to qualify for new technology add-on payments and received the following comments during the 60-day comment period.

During the 60-day comment period on the FY 2006 proposed rule, we received the following comments:

Comment: Several commenters supported new technology add-on payments for the Trident[supreg] ceramic on ceramic hip. Many of these comments reiterated the comment from the device manufacturer, disagreeing with our assertion that the technology represents only an incremental improvement over other technologies. The commenters emphasized that the Trident[supreg] Ceramic Acetabular System had been evaluated in an extensive prospective, randomized, controlled clinical study, and that the FDA Panel reviewing the

[[Page 47362]]

study commended it for its design, statistical report, and patient followup. Therefore, the commenters argued, the product had shown clinical superiority where other devices and improved designs had not shown clinical superiority to the metal on polyethelene hip implants. The commenter also cited a post-market study of a subset of the original study patients that demonstrates continued good patient outcomes at a mean of 5.2 years followup, as presented at the 2005 American Academy of Orthopedic Surgeons Annual Meeting.

Response: The Trident[supreg] Ceramic Acetabular System is used to replace the ``ball and socket'' joint of a hip when a total hip replacement is performed for patients suffering from arthritis or related conditions. Prosthetic hip joints have been used to treat these conditions for many years. The Trident[supreg] Ceramic Acetabular System differs from its predecessor prosthetic hips only in the materials that are used in the joint. Thus, the Trident[supreg] Ceramic Acetabular System uses the same or a similar mechanism of action to achieve a therapeutic outcome (that is, it replaces the joint to address pain and related conditions for patients suffering from arthritis or related conditions). Further, we note that the cases using the Trident[supreg] Ceramic Acetabular System will go into new DRGs 544 or 545 (Major Joint Replacement, Revision of Hip or Knee Replacement), the same DRGs as the patients that receive the older prosthetic hip replacements. Therefore, because the Trident product appears to offer only an incremental advance in the treatment of patients requiring a total hip replacement, we find that it does not meet the substantial clinical improvement criterion. We also note that in this final rule, as proposed, CMS is splitting DRG 209 into two separate DRGs (544 and 545) in order to better reflect the higher costs of revising hip and knee replacements.

Comment: Several commenters objected to our interpretation of the period of new with regard to this technology. Several commenters noted that there appeared to be inconsistency in the method CMS has used to determine the period of ``newness'' for each technology, noting in particular that both the CRT-D device and INFUSE[supreg] bone graft for spinal fusion received new technology add-on payment beyond the 2-3 year period that the devices could be considered new. As noted in the proposed rule, commenters argued that, by CMS' own rationale, payment beyond this period was designed to provide payment predictability and consistency for the entire fiscal year, rather than terminating payments part way through the year. Commenters urged us to reconsider whether this technology meets the newness criterion because it will not be 3 years old until more than 6 months into FY 2006.

Response: We believe the commenters make a good point about application of the newness criteria to the Trident[supreg] product. The commenters are correct that we have generally followed a guideline that uses a 6-month window before and after the start of the fiscal year to determine whether to extend the add-on payment for an additional year. In general, we extend add-on payments for an additional year if the 3 year anniversary date of the product's entry on the market occurs in the latter half of the fiscal year.

In the case of the Trident[supreg] ceramic acetabular system, the device was not available on the market until April, 2003. Thus, the product will not have been available on the market for 3 years until the second half of FY 2006. Thus, under policy, the Trident[supreg] ceramic acetabular system could potentially qualify as new for FY 2006. However, the device is very similar to existing products, only differing in the composite material used in manufacturing. It is also used in the same DRGs as these other similar technologies, and we question whether it would be appropriate to deem this technology new and substantially different from previous hip prosthetics. Thus, as noted above, we continue to find that the device does not meet our substantial clinical improvement criterion. Therefore, in this final rule, we are finalizing our decision to deny new technology add-on payments for this device for FY 2006. h. WingspanTMStent System With GatewayTMPTA Balloon Catheter

Boston Scientific submitted an application for the WingspanTMStent System with GatewayTMPTA Balloon Catheter for new technology add-on payments. The device is designed for the treatment of patients with intracranial atherosclerotic disease who suffer from recurrent stroke despite medical management. The device consists of the following: A self- expanding nitinol stent, a multilumen over the wire delivery catheter, and a Gateway PTA Balloon Catheter. The device is used to treat stenoses that occur in the intracranial vessels. Prior to stent placement, the Gateway PTA Balloon is inflated to dilate the target lesion, and then the stent is deployed across the lesion to restore and maintain luminal patency. Effective October 1, 2004, two new ICD-9-CM procedure codes were created to code intracranial angioplasty and intracranial stenting procedures: Procedure codes 00.62 (Percutaneous angioplasty or atherectomy of intracranial vessels) and 00.65 (Percutaneous insertion of intracranial vascular stents).

On January 9, 2004, the FDA designated the WingspanTMas a Humanitarian Use Designation (HUD). The manufacturer has also applied for Humanitarian Device Exemption (HDE) status and expects approval from the FDA in July 2005. It is important to note that currently CMS has a noncoverage policy for percutaneous transluminal angioplasty to treat lesions of intracranial vessels. The applicant is working closely with CMS to review this decision upon FDA approval. Because the device is neither FDA-approved nor Medicare-covered, we did not believe it was appropriate to present our full analysis on whether the technology meets the individual criteria for the new technology add-on payment in the proposed rule. However, we note that the applicant did submit the following information below on the cost criterion and substantial clinical improvement criterion.

The manufacturer submitted data from MedPAR and non-MedPAR databases. The non-MedPAR data was from the 2003 patient discharge data from California's Office of Statewide Health Planning and Development database for hospitals in California and from the 2003 patient data from Florida's Agency for Health Care Administration for hospitals in Florida. The applicant identified cases that had a diagnosis code of 437.0 (Cerebral atherosclerosis), 437.1 (Other generalized ischemic cerebrovascular disease) or 437.9 (Unspecified) or any diagnosis code that begins with the prefix of 434 (Occlusion of cerebral arteries) in combination with procedure code 39.50 (Angioplasty or atherectomy of noncoronary vessel) or procedure code 39.90 (Insertion of nondrug- eluting, noncoronary artery stents). The applicant used procedure codes 39.50 and 39.90 because procedure codes 00.62 and 00.65 were not available until FY 2005. The applicant found cases in DRG 5 (Extracranial Vascular Procedures) (which previously existed under the Medicare IPPS DRG system prior to a DRG split) and in DRGs 533 (Extracranial Procedure with CC) and 534 (Extracranial Procedure Without CC). Even though DRG 5 was split into DRGs 533 and 534 in FY 2003, some hospitals continued to use DRG 5 for non-Medicare cases. The applicant found 22 cases that had an intracranial

[[Page 47363]]

PTA with a stent. The average (nonstandardized) charge per case was $78,363.

The applicant also submitted data from the FY 2002 and FY 2003 MedPAR files. Using the latest data from the FY 2003 MedPAR and the same combination of diagnosis and procedure codes mentioned above to identify cases of intracranial PTA with stenting, the applicant found 116 cases in DRG 533 and 20 cases in DRG 534. The case-weighted average standardized charge per case was $51,173. The average case-weighted threshold was $25,394. Based on this analysis, the applicant maintained that the technology meets the cost criteria since the average case- weighted standardized charge per case is greater than the average case- weighted threshold.

The applicant also maintained that the technology meets the substantial clinical improvement criterion. Currently, there is no available surgical or medical treatment for recurrent stroke that occurs despite optimal medical management. The WingspanTMis the first commercially available PTA/stent system designed specifically for the intracranial vasculature. However, because the WingspanTMdoes not have FDA approval or Medicare coverage, as stated above, in the FY 2006 IPPS proposed rule, we proposed to deny add-on payment for this new technology.

We received no public comments regarding this application for add- on payments prior to the publication of the FY 2006 IPPS proposed rule.

During the 60-day comment period for the FY 2006 IPPS proposed rule, we received the following comment:

Comment: One commenter, the applicant, commented that the WingspanTMrepresents a substantial clinical improvement over what is currently available to treat patients with intracranial atherosclerotic disease, and who suffer from recurring stroke and recommended that, upon FDA approval of the WingspanTM, CMS determine the most appropriate payment for this new therapy.

Response: We thank the commenter for its comments and upon FDA approval we encourage the applicant to reapply for new technology add- on payments. However, because the WingspanTMdoes not have FDA approval or Medicare coverage, we are finalizing our proposal to deny add-on payment for this new technology.

III. Changes to the Hospital Wage Index

A. Background

Section 1886(d)(3)(E) of the Act requires that, as part of the methodology for determining prospective payments to hospitals, the Secretary must adjust the standardized amounts ``for area differences in hospital wage levels by a factor (established by the Secretary) reflecting the relative hospital wage level in the geographic area of the hospital compared to the national average hospital wage level.'' In accordance with the broad discretion conferred under the Act, we currently define hospital labor market areas based on the definitions of statistical areas established by the Office of Management and Budget (OMB). A discussion of the FY 2006 hospital wage index based on the statistical areas, including OMB's revised definitions of Metropolitan Areas, appears under section III.B. of this preamble.

Beginning October 1, 1993, section 1886(d)(3)(E) of the Act requires that we update the wage index annually. Furthermore, this section provides that the Secretary base the update on a survey of wages and wage-related costs of short-term, acute care hospitals. The survey should measure the earnings and paid hours of employment by occupational category, and must exclude the wages and wage-related costs incurred in furnishing skilled nursing services. This provision also requires us to make any updates or adjustments to the wage index in a manner that ensures that aggregate payments to hospitals are not affected by the change in the wage index. The adjustment for FY 2006 is discussed in section II.B. of the Addendum to this final rule.

As discussed below in section III.H. of this preamble, we also take into account the geographic reclassification of hospitals in accordance with sections 1886(d)(8)(B) and 1886(d)(10) of the Act when calculating the wage index. Under section 1886(d)(8)(D) of the Act, the Secretary is required to adjust the standardized amounts so as to ensure that aggregate payments under the IPPS after implementation of the provisions of sections 1886(d)(8)(B) and (C) and 1886(d)(10) of the Act are equal to the aggregate prospective payments that would have been made absent these provisions. The budget neutrality adjustment for FY 2006 is discussed in section II.B. of the Addendum to this final rule.

Section 1886(d)(3)(E) of the Act also provides for the collection of data every 3 years on the occupational mix of employees for short- term, acute care hospitals participating in the Medicare program, in order to construct an occupational mix adjustment to the wage index. A discussion of the occupational mix adjustment that we are applying beginning October 1, 2005 (the FY 2006 wage index) appears under section III.C. of this preamble.

B. Core-Based Statistical Areas Used for the Proposed Hospital Wage Index

The wage index is calculated and assigned to hospitals on the basis of the labor market area in which the hospital is located. In accordance with the broad discretion under section 1886(d)(3)(E) of the Act, beginning with FY 2005, we define hospital labor market areas based on the Core-Based Statistical Areas (CBSAs) established by OMB and announced in December 2003 (69 FR 49027). OMB defines a CBSA, beginning in 2003, as ``a geographic entity associated with at least one core of 10,000 or more population, plus adjacent territory that has a high degree of social and economic integration with the core as measured by commuting ties.'' The standards designate and define two categories of CBSAs: Metropolitan Statistical Areas (MSAs) and Micropolitan Statistical Areas (65 FR 82235).

According to OMB, MSAs are based on urbanized areas of 50,000 or more population, and Micropolitan Statistical Areas (referred to in this discussion as Micropolitan Areas) are based on urban clusters with a population of at least 10,000 but less than 50,000. Counties that do not fall within CBSAs are deemed ``Outside CBSAs.'' In the past, OMB defined MSAs around areas with a minimum core population of 50,000, and smaller areas were ``Outside MSAs.''

The general concept of the CBSAs is that of an area containing a recognized population nucleus and adjacent communities that have a high degree of integration with that nucleus. The purpose of the standards is to provide nationally consistent definitions for collecting, tabulating, and publishing Federal statistics for a set of geographic areas. CBSAs include adjacent counties that have a minimum of 25 percent commuting to the central counties of the area. (This is an increase over the minimum commuting threshold of 15 percent for outlying counties applied in the previous MSA definition.)

The new CBSAs established by OMB comprised MSAs and the new Micropolitan Areas based on Census 2000 data. (A copy of the announcement may be obtained at the following Internet address: http://www.whitehouse.gov/omb/bulletins/

[[Page 47364]]

fy04/b04-03.html.) The definitions recognize 49 new MSAs and 565 new Micropolitan Areas, and extensively revised the composition of many of the existing MSAs.

The new area designations resulted in a higher wage index for some areas and lower wage index for others. Further, some hospitals that were previously classified as urban are now in rural areas. Given the significant payment impacts upon some hospitals because of these changes, we provided a transition period to the new labor market areas in the FY 2005 IPPS final rule (69 FR 49027 through 49034). As part of that transition, we allowed urban hospitals that became rural under the new definitions to maintain their assignment to the Metropolitan Statistical Area (MSA) where they were previously located for the 3- year period of FY 2005, FY 2006, and FY 2007. Specifically, these hospitals were assigned the wage index of the urban area to which they previously belonged. (For purposes of wage index computation, the wage data of these hospitals remained assigned to the statewide rural area in which they are located.) The hospitals receiving this transition will not be considered urban hospitals; rather they will maintain their status as rural hospitals. Thus, the hospital would not be eligible, for example, for a large urban add-on payment under the capital PPS. In other words, it is the wage index, but not the urban or rural status, of these hospitals that is being affected by this transition. The higher wage indices that these hospitals are receiving are also being taken into consideration in determining whether they qualify for the out-commuting adjustment discussed in section III.I. of this preamble and the amount of any adjustment.

FY 2006 will be the second year of this transition period. We will continue to assign the wage index for the urban area in which the hospital was previously located through FY 2007. In order to ensure this provision remains budget neutral, we will continue to adjust the standardized amount by a transition budget neutrality factor to account for these hospitals. Doing so is consistent with the requirement of section 1886(d)(3)(E) of the Act that any ``adjustments or updates [to the adjustment for different area wage levels] * * * shall be made in a manner that assures that aggregate payments * * * are not greater or less than those that would have been made in the year without such adjustment.''

Beginning in FY 2008, these hospitals will receive their statewide rural wage index, although they will be eligible to apply for reclassification by the MGCRB, both during this transition period as well as in subsequent years. These hospitals will be considered rural for reclassification purposes.

In addition, in the FY 2005 IPPS final rule (69 FR 49032 and 49033), we provided a 1-year transition blend for hospitals that, due solely to the changes in the labor market definitions, experienced a decrease in their FY 2005 wage index compared to the wage index they would have received using the labor market areas included in calculating their FY 2004 wage index. Hospitals that experienced a decrease in their wage index as a result of adoption of the new labor market area changes received a wage index based on 50 percent of the CBSA labor market area definitions and 50 percent of the wage index that the provider would have received under the FY 2004 MSA boundaries (in both cases using the FY 2001 wage data). This blend applied to any provider experiencing a decrease due to adoption of the new definitions, including providers who were reclassifying under MGCRB requirements, section 1886(d)(8)(B) of the Act, or section 508 of Pub. L. 108-173. In the FY 2005 IPPS final rule (69 FR 49027 through 49033), we described the determination of this blend in detail. We noted that this blend does not prevent a decrease in wage index due to any reason other than adoption of CBSAs, nor does it apply to hospitals that benefited from a higher wage index due to the new labor market definitions.

Consistent with the FY 2005 IPPS final rule, beginning in FY 2006, we are providing that hospitals receive 100 percent of their wage index based upon the new CBSA configurations. Specifically, we have determined for each hospital a new wage index for FY 2006 employing wage index data from FY 2002 hospital cost reports and using the CBSA labor market definitions.

Comment: Commenters asked CMS to defer 100 percent adoption of the new labor market area definitions to allow hospitals more time to adjust to the significant reimbursement impact. Most commenters urged CMS to maintain the current 50 percent CBSA/50 percent MSA blend. One commenter proposed using a 75 percent CBSA/25 percent MSA blend.

Response: We have decided not to provide for a longer transition because we have already, in effect, provided 1 year at a higher wage index for hospitals by delaying full implementation of the new Census designations. Given that the new designations are based on the most recent Census data, whereas the prior labor market areas are based on 1990 Census data, we believe it is both reasonable and appropriate to adopt the new designations for FY 2006.

Comment: One commenter noted that, while CMS provided urban hospitals that became rural under the new definitions hold harmless protection for 3 years, urban hospitals that remained in MSAs that experienced large wage index reductions did not receive that same protection. The commenter stated that, although all hospitals that experienced a decrease in their wage index from the effects of the labor market area changes received a 1-year blended transition, this transition expires on September 30, 2005. The commenter urged CMS to provide hold harmless protection to all hospitals that experienced a wage index decrease of more than 10 percent as a result of the new labor market areas, regardless of whether the hospital remained urban or rural.

Response: We refer readers to the FY 2005 IPPS final rule for a full discussion of our rationale for limiting hold harmless protection to a particular group of hospitals (69 FR 49032).

Comment: A few commenters addressed the use of Micropolitan Areas as geographic areas. They stated that because CMS assigns Micropolitan Areas to the statewide rural area for purposes of the IPPS, several hospitals, by virtue of now being in a Micropolitan county, are reclassified as rural despite their previous designation as an urban hospital. They noted that, although CMS provided a 3-year transition period to help alleviate the decreased wage index payments for hospitals that were previously classified as urban and are now in rural areas based on the new definitions, this transition did not ameliorate any reductions in DSH payments, because the transition did not affect a hospital's urban/rural status. They emphasized that, while urban hospitals of 100 or more beds have no cap on DSH payments, rural hospitals of all sizes are capped at 12 percent for DSH payments. Commenters offered various recommendations about how to protect these hospitals from the changes in the labor market area definitions. Most commenters advocated allowing counties that are reclassified as Micropolitan Areas despite their previous urban designation to be grandfathered into their previously urban MSA. Other commenters recommended that CMS provide an exception to these hospitals under section 1886(d)(5)(I)(i) of the Act. Further, commenters suggested that CMS adopt OMB's new standards for use in defining labor market areas, but lower the commuting threshold used by OMB to define CBSAs.

[[Page 47365]]

Response: We disagree with the commenters that hospitals that changed status from urban to rural received no amelioration with respect to DSH. As stated in the FY 2005 IPPS final rule (69 FR 49033), the provisions of Sec. 412.102 provide special protections for hospitals against abrupt reductions in DSH payments resulting from transitions from urban to rural status. Specifically, as described in Sec. 412.102, in the first year after a hospital loses urban status, the hospital will receive an additional payment that equals two-thirds of the difference between the urban disproportionate share payments applicable to the hospital before its redesignation from urban to rural and the rural disproportionate share payments applicable to its redesignation from urban to rural. In the second year after the hospital loses urban status, the hospital will receive an additional payment that equals one-third of the difference between the urban disproportionate share payments applicable to the hospital before its redesignation from urban to rural and the rural disproportionate share payments applicable to its redesignation from urban to rural. Because hospitals are already receiving adequate relief with respect to DSH payments, we do not believe it is necessary to address the commenters' recommendations regarding grandfathering, exceptions, or use of lower commuting thresholds. We refer readers to the explanation in the FY 2005 IPPS final rule for our adoption of the new Census designations as well as the treatment of Micropolitan areas as rural (69 FR 49027).

C. Occupational Mix Adjustment to FY 2006 Index

As stated earlier, section 1886(d)(3)(E) of the Act provides for the collection of data every 3 years on the occupational mix of employees for each short-term, acute care hospital participating in the Medicare program, in order to construct an occupational mix adjustment to the wage index, for application beginning October 1, 2004 (the FY 2005 wage index). The purpose of the occupational mix adjustment is to control for the effect of hospitals' employment choices on the wage index. For example, hospitals may choose to employ different combinations of registered nurses, licensed practical nurses, nursing aides, and medical assistants for the purpose of providing nursing care to their patients. The varying labor costs associated with these choices reflect hospital management decisions rather than geographic differences in the costs of labor. 1. Development of Data for the Occupational Mix Adjustment

In the FY 2005 IPPS final rule (69 FR 49034), we discussed in detail the data we used to calculate the occupational mix adjustment to the FY 2005 wage index. For the final FY 2006 wage index, as proposed, we are using the same CMS Wage Index Occupational Mix Survey and Bureau of Labor Statistics (BLS) data that we used for the FY 2005 wage index, with two exceptions. The CMS survey requires hospitals to report the number of total paid hours for directly hired and contract employees in occupations that provide the following services: Nursing, physical therapy, occupational therapy, respiratory therapy, medical and clinical laboratory, dietary, and pharmacy. These services each include several standard occupational classifications (SOCs), as defined by the BLS' Occupational Employment Statistics (OES) survey. For the FY 2006 wage index, we used revised survey data for 20 hospitals that took advantage of the opportunity we afforded hospitals to submit changes to their occupational mix data during the FY 2006 wage index data collection process (see discussion of wage data corrections process under section III.J. of this preamble). We also excluded survey data for hospitals that became designated as CAHs since the original survey data were collected and hospitals for which there are no corresponding cost report data for the FY 2006 wage index. The FY 2006 wage index includes occupational mix data from 3,541 out of 3,742 hospitals (94.6 percent response rate). The results of the occupational mix survey are included in the chart below.

BILLING CODE 4120-01-P

[[Page 47366]]

[GRAPHIC] [TIFF OMITTED] TR12AU05.000

[[Page 47367]]

[GRAPHIC] [TIFF OMITTED] TR12AU05.001

BILLING CODE 4120-01-C

Comment: Two commenters noted that the ``Medicare Occupational Mix Survey Results'' table in the FY 2006 proposed rule (70 FR 23369) did not include data pertaining to medical and clinical laboratory services, all other occupations, and total hospital employees. The commenters requested that CMS publish the complete table in the final rule.

Response: We apologize for any inconveniences caused by the misprint of the table in the proposed rule. The above table includes the complete set of occupational mix survey results for the final FY 2006 wage index.

Comment: As a mechanism to achieve a higher response rate, one commenter recommended that CMS reward hospitals that submit occupational mix survey data. The commenter suggested that, for hospitals that submit occupational mix data, CMS should apply a higher percentage of the occupational mix adjustment if the adjustment results in a positive impact, and a lower percentage if the adjustment results in a negative impact.

Response: Although the commenter's suggestion pertaining to a procedural mechanism by which CMS conducts the occupational mix survey is not a subject of the final policies included in this final rule, we note that we disagree with the suggestion. We do not believe that hospitals should receive a special reward for completing and submitting the occupational mix survey. Rather, a hospital should deem the submission of occupational mix data as a necessary part of its responsibility to provide complete and accurate data for the wage index. We also note that implementing an occupational mix adjustment so that it applies to reporting hospitals only when it is beneficial to such hospitals would defeat the purpose of the occupational mix adjustment. 2. Calculation of the FY 2006 Occupational Mix Adjustment Factor and the FY 2006 Occupational Mix Adjusted Wage Index

For the final FY 2006 wage index, we used the same methodology that we used to calculate the occupational mix adjustment to the FY 2005 wage index (69 FR 49042). We used the following steps for calculating the FY 2006 occupational mix adjustment factor and the occupational mix adjusted wage index:

Step 1--For each hospital, the percentage of the general service category attributable to an SOC is determined by dividing the SOC hours by the general service category's total hours. Repeat this calculation for each of the 19 SOCs.

Step 2--For each hospital, the weighted average hourly rate for an SOC is determined by multiplying the percentage of the general service category (from Step 1) by the national average hourly rate for that SOC from the 2001 BLS OES survey, which was used in calculating the occupational mix adjustment for the FY 2005 wage index. The 2001 OES survey is BLS' latest available hospital-specific survey. (See Chart 4 in the FY 2005 IPPS final rule, 69 FR 49038.) Repeat this calculation for each of the 19 SOCs.

Step 3--For each hospital, the hospital's adjusted average hourly rate for a general service category is computed by summing the weighted hourly rate for each SOC within the general category. Repeat this calculation for each of the seven general service categories.

Step 4--For each hospital, the occupational mix adjustment factor for a general service category is calculated by dividing the national adjusted average hourly rate for the category by the hospital's adjusted average hourly rate for the category. (The national adjusted average hourly rate is computed in the same manner as Steps 1 through 3, using instead, the total SOC and general service category hours for all hospitals in the occupational mix survey database.) Repeat this calculation for each of the seven general service categories. If the hospital's adjusted rate

[[Page 47368]]

is less than the national adjusted rate (indicating the hospital employs a less costly mix of employees within the category), the occupational mix adjustment factor will be greater than 1.0000. If the hospital's adjusted rate is greater than the national adjusted rate, the occupational mix adjustment factor will be less than 1.0000.

Step 5--For each hospital, the occupational mix adjusted salaries and wage-related costs for a general service category are calculated by multiplying the hospital's total salaries and wage-related costs (from Step 5 of the unadjusted wage index calculation in section III.F. of this preamble) by the percentage of the hospital's total workers attributable to the general service category and by the general service category's occupational mix adjustment factor (from Step 4 above). Repeat this calculation for each of the seven general service categories. The remaining portion of the hospital's total salaries and wage-related costs that is attributable to all other employees of the hospital is not adjusted for occupational mix.

Step 6--For each hospital, the total occupational mix adjusted salaries and wage-related costs for a hospital are calculated by summing the occupational mix adjusted salaries and wage-related costs for the seven general service categories (from Step 5) and the unadjusted portion of the hospital's salaries and wage-related costs for all other employees. To compute a hospital's occupational mix adjusted average hourly wage, divide the hospital's total occupational mix adjusted salaries and wage-related costs by the hospital's total hours (from Step 4 of the unadjusted wage index calculation in section III.F. of this preamble).

Step 7--To compute the occupational mix adjusted average hourly wage for an urban or rural area, sum the total occupational mix adjusted salaries and wage-related costs for all hospitals in the area, then sum the total hours for all hospitals in the area. Next, divide the area's occupational mix adjusted salaries and wage-related costs by the area's hours.

Step 8--To compute the national occupational mix adjusted average hourly wage, sum the total occupational mix adjusted salaries and wage- related costs for all hospitals in the Nation, then sum the total hours for all hospitals in the Nation. Next, divide the national occupational mix adjusted salaries and wage-related costs by the national hours. The national occupational mix adjusted average hourly wage for FY 2006 is $28.0272.

Step 9--To compute the occupational mix adjusted wage index, divide each area's occupational mix adjusted average hourly wage (Step 7) by the national occupational mix adjusted average hourly wage (Step 8).

Step 10--To compute the Puerto Rico specific occupational mix adjusted wage index, follow Steps 1 through 9 above. The Puerto Rico occupational mix adjusted average hourly wage for FY 2006 is $12.7985.

An example of the occupational mix adjustment was included in the FY 2005 IPPS final rule (69 FR 49043).

For the FY 2005 final wage index, we used the unadjusted wage data for hospitals that did not submit occupational mix survey data. For calculation purposes, this equates to applying the national SOC mix to the wage data for these hospitals, because hospitals having the same mix as the Nation would have an occupational mix adjustment factor equaling 1.0000. In the FY 2005 IPPS final rule (69 FR 49035), we noted that we would revisit this matter with subsequent collections of the occupational mix data. Because we are using essentially the same survey data for the FY 2006 occupational mix adjustment that we used for FY 2005, with the only exceptions as stated in section III.C.1. of this preamble, we are treating the wage data for hospitals that did not respond to the survey in this same manner for the FY 2006 wage index.

In implementing an occupational mix adjusted wage index based on the above calculation, the wage index values for 14 rural areas (29.8 percent) and 206 urban areas (53.4 percent) would decrease as a result of the adjustment. Seven (7) rural areas (14.9 percent) and 111 urban areas (28.8 percent) would experience a decrease of 1 percent or greater in their wage index values. The largest negative impact for a rural area would be 1.9 percent and for an urban area, 4.2 percent. Meanwhile, 32 rural areas (68.1 percent) and 179 urban areas (46.4 percent) would experience an increase in their wage index values. Although these results show that rural hospitals would gain the most from an occupational mix adjustment to the wage index, their gains may not be as great as might have been expected. Further, it might not have been anticipated that almost one-third of rural hospitals would actually fare worse under the adjustment. Overall, a fully implemented occupational mix adjusted wage index would have a redistributive effect on Medicare payments to hospitals.

In the FY 2005 IPPS proposed rule, we indicated that, for future data collections, we would revise the occupational mix survey to allow hospitals to provide both salaries and hours data for each of the employment categories that are included on the survey. We also indicated that we would assess whether future occupational mix surveys should be based on the calendar year or if the data should be collected on a fiscal year basis as part of the Medicare cost report. (One logistical problem is that cost report data are collected yearly, but occupational mix survey data are collected only every 3 years.) We are currently reviewing options for revising the occupational mix survey and improving the data collection process.

Comment: Several commenters provided recommendations for the design and release of a revised occupational mix survey.

Response: We plan to release a revised occupational mix survey in an upcoming Federal Register notice. We will address the design and data collection issues, including the commenters' recommendations, as part of that notice.

In our continuing efforts to meet the information needs of the public, we provided via the Internet three additional public use files for the proposed occupational mix adjusted wage index concurrently with the publication of the FY 2006 IPPS proposed rule: (1) A file including each hospital's unadjusted and adjusted average hourly wage (FY 2006 Proposed Rule Occupational Mix Adjusted and Unadjusted Average Hourly Wage by Provider); (2) a file including each CBSA's adjusted and unadjusted average hourly wage (FY 2006 Proposed Rule Occupational Mix Adjusted and Unadjusted Average Hourly Wage and Pre-Reclassified Wage Index by CBSA); and (3) a file including each hospital's occupational mix adjustment factors by occupational category (Provider Occupational Mix Adjustment Factors for Each Occupational Category). We also plan to post these files via the Internet with future applications of the occupational mix adjustment.

D. Worksheet S-3 Wage Data for the FY 2006 Wage Index Update

The FY 2006 wage index values (effective for hospital discharges occurring on or after October 1, 2005 and before October 1, 2006) in section VI. of the Addendum to this final rule are based on the data collected from the Medicare cost reports submitted by hospitals for cost reporting periods beginning in FY 2002 (the FY 2005 wage index was based on FY 2001 wage data).

[[Page 47369]]

The FY 2006 wage index includes the following categories of data associated with costs paid under the IPPS (as well as outpatient costs):

Salaries and hours from short-term, acute care hospitals (including paid lunch hours and hours associated with military leave and jury duty).

Home office costs and hours.

Certain contract labor costs and hours (which includes direct patient care, certain top management, pharmacy, laboratory, and nonteaching physician Part A services).

Wage-related costs, including pensions and other deferred compensation costs.

The September 1, 1994 Federal Register (59 FR 45356) included a list of core wage-related costs that are included in the wage index, and discussed criteria for including other wage-related costs. In that discussion, we instructed hospitals to use generally accepted accounting principles (GAAPs) in developing wage-related costs for the wage index for cost reporting periods beginning on or after October 1, 1994. We discussed our rationale that ``the application of GAAPs for purposes of compiling data on wage-related costs used to construct the wage index will more accurately reflect relative labor costs, because certain wage-related costs (such as pension costs), as recorded under GAAPs, tend to be more static from year to year.''

Since publication of the September 1, 1994 rule, we have periodically received inquiries for more specific guidance on developing wage-related costs for the wage index. In response, we have provided clarifications in the IPPS rules (for example, health insurance costs (66 FR 39859)) and in the cost report instructions (Provider Reimbursement Manual (PRM), Part II, Section 3605.2). Due to recent questions and concerns we received regarding inconsistent reporting and overreporting of pension and other deferred compensation plan costs, as a result of an ongoing Office of Inspector General review, we are clarifying in this final rule that hospitals must comply with the requirements in 42 CFR 413.100, the PRM, Part I, sections 2140, 2141, and 2142, and related Medicare program instructions for developing pension and other deferred compensation plan costs as wage- related costs for the wage index. The Medicare instructions for pension costs and other deferred compensation costs combine GAAPs, Medicare payment principles, and Department of Labor and Internal Revenue Service requirements. We believe that the Medicare instructions allow for both consistent reporting among hospitals and for the development of reasonable deferred compensation plan costs for purposes of the wage index.

With the FY 2007 wage index, hospitals and fiscal intermediaries must ensure that pension, post-retirement health benefits, and other deferred compensation plan costs for the wage index are developed according to the above terms.

Comment: A few commenters addressed our discussion regarding the treatment of pension, post-retirement health benefits, and other deferred compensation costs for purposes of the wage index. Two commenters expressed concern that the instructions are a significant change from our original instructions published September 1, 1994. The commenters asserted that CMS provided no rationale for moving away from using GAAP for developing these costs for the wage index, and requested an additional opportunity for public comment. One commenter suggested that using GAAP provides a more consistent methodology for capturing these costs than Medicare reasonable cost principles. A fourth commenter requested a more specific description of the treatment of pension, post-retirement health benefits, and other deferred compensation costs if there are other ``related Medicare program instructions'' as we stated above.

Response: For cost reporting periods beginning prior to October 1, 1994, hospitals were required to include in the wage index only the amount of actual payments that the hospital made to retirees in the reporting year. For periods beginning on or after October 1, 1994, CMS instructed hospitals to use GAAPs, an accrual method of accounting, for developing pension, deferred compensation, and other wage-related costs for wage index purposes. All other wage costs on Worksheet S-3 must reflect costs that are actually expended by the hospital during the cost reporting period. We believed then and continue to believe that the use of accrual accounting allows hospitals to be more inconsistent in their reporting of wage-related costs from year to year so that the wage index could be more static.

Section 413.24 of the regulations also provides for the accrual basis of accounting for developing costs under Medicare's cost finding principles. However, a major difference between GAAP and Medicare principles for recognizable pension and other deferred compensation plan costs is an issue of funding. In Sec. 413.100 (and as discussed in 60 FR 33126, June 27, 2005), we clarified and codified CMS' longstanding requirement that, for purposes of program payment, providers must timely liquidate their liabilities. GAAP does not specify a time requirement for recognizing accrued costs.

In 2003, we updated the cost report instructions in section 3605.2 of the PRM, Part II, to also clarify the September 1, 1994 instructions for the wage index. At the instructions for wage-related costs, lines 13 through 20, we noted that, ``Although hospitals must use GAAP in developing wage-related costs, the amount reported for wage index purposes must meet the reasonable cost provisions of Medicare.'' The clarification was to ensure that a hospital includes in the wage index only those pension and other deferred compensation plan costs that meet the timely liquidation requirements for Medicare reasonable cost principles. When CMS issued the September 1, 1994 instructions, CMS did not anticipate nor intend for hospitals to include costs in the wage index that have not been funded and may never be funded. Including unfunded deferred compensation costs in the wage index can significantly misrepresent an area's average hourly wage, especially if the plan is never funded. In a May 4, 2005 Early Alert to CMS's Administrator, the OIG stated that ``While some hospitals include millions of dollars in unfunded pension and other postretirement benefit costs in the annual wage data shown on their Medicare cost reports, others include only funded amounts. As a result, the wage indexes for the hospitals that include unfunded amounts are inflated, which leads to an inadequate distribution of Medicare payments among hospitals.'' In addition, the OIG warned that ``* * * the hospitals' inclusion of costs related to unfunded liabilities could compromise the reliability of the wage data that CMS uses to develop the market basket * * *. Thus, the inclusion of costs related to unfunded liabilities in hospitals' wage data could produce an inaccurate market basket index for use in updating payments to hospitals.''

Regarding the comment requesting a specific description of the treatment of pension, post-retirement health benefits, and other deferred compensation costs if there are other ``related Medicare program instructions,'' we included this phrase to set forth that hospitals must also comply with any future instructions related to these costs that may be initially issued through rulemaking or a one- time notice before being included in the above PRM sections.

[[Page 47370]]

We believe that our discussion in the proposed rule was sufficient notification for this policy clarification. Therefore, we do not agree that CMS should provide another comment period for this matter. In addition, we believe that hospitals and intermediaries should be able to ensure that pension and other deferred compensation costs are developed according to the above terms by the FY 2007 wage index, as hospitals have been required, since cost reporting periods beginning during FY 1995, to complete Form 339, a reconciliation worksheet between GAAP and Medicare principles.

Consistent with the wage index methodology for FY 2005, the wage index for FY 2006 also excludes the direct and overhead salaries and hours for services not subject to IPPS payment, such as SNF services, home health services, costs related to GME (teaching physicians and residents) and certified registered nurse anesthetists (CRNAs), and other subprovider components that are not paid under the IPPS. The FY 2006 wage index also excludes the salaries, hours, and wage-related costs of hospital-based rural health clinics (RHCs), and Federally qualified health centers (FQHCs) because Medicare pays for these costs outside of the IPPS (68 FR 45395). In addition, salaries, hours and wage-related costs of CAHs are excluded from the wage index, for the reasons explained in the FY 2004 IPPS final rule (68 FR 45397).

Comment: Two commenters recommended that CAHs be included in the wage index. One commenter suggested that CMS should exclude the wage data for a CAH only if it is designated a CAH during the base year for the wage index calculation. MedPAC suggested that CMS should include the wage data for all CAHs, even if the hospital is a CAH in the base year that is used for calculating the wage index. In addition, MedPAC stated the following:

The wage index should ideally reflect the data for all providers that are similar in services and occupations to hospitals receiving payment under Medicare's IPPS and OPPS. CAHs are similar to other small rural hospitals and in many cases are located close enough to IPPS hospitals to compete for the same workers.

About 500 hospitals converted to CAH status over the past 3 years. Since CAHs now dominate the rural areas for some states, the data for CAHs may become critical for an accurate representation of rural area wage levels. It is important to note that this representation affects payment for not only the IPPS hospitals but also for other providers that are paid under a Medicare prospective payment system, such as SNFs, HHAs, and LTCHs.

MedPAC recommended that CMS begin collecting wage data from CAHs this year.

Response: In the FY 2004 final rule (68 FR 45397), we provided a complete discussion, rationale, and analysis of our policy for excluding CAHs from the wage index. In that rule, we stated that CAHs are not paid under the IPPS, and, like other non-IPPS providers such as SNFs, HHAs, LTCHs, and children's hospitals, we have always excluded non-IPPS providers from the wage index calculation. We also stated that, due to their remote location and more limited services, CAHs ``are unique compared to other short-term acute care hospitals.'' Using data collected from cost reporting periods beginning during FY 2000, we further noted that, in most labor market areas with hospitals that converted to CAH status some time after FY 2000, the average hourly wage for CAHs was significantly lower than the average hourly wage for other short-term hospitals in the area. As a result, with the FY 2004 wage index, we began excluding the data for any CAH, even if it was an IPPS provider during the wage index base year.

We agree with MedPAC that CAHs have recently become more similar in composition, services, and proximity to other rural hospitals, largely due to the Pub. L. 108-173 (MMA). Section 405 of Pub. L. 108-173 allows for more hospitals to now qualify and more seamlessly convert to CAH status. However, because Pub. L. 108-173 was enacted in calendar year 2003, it would not affect the FY 2002 base year for the FY 2006 IPPS wage index. In addition, our analysis of the FY 2006 wage index shows that rural areas are not harmed by the exclusion of CAHs. For FY 2006, we removed the wage data for 162 hospitals in 39 rural areas because they became CAHs after they filed their FY 2002 cost reports as IPPS hospitals. In all 39 rural areas, the average hourly wages for FY 2006 increased over those for FY 2005. For 76.9 percent of the rural areas, the average hourly wage increase is 5 percent or greater.

Therefore, we continue to believe that it is prudent policy to remove the data from CAHs from the wage index. As such, we have excluded from the FY 2006 wage index in this final rule the wages and hours for all hospitals that are currently designated as a CAH, even if the hospital was paid under the IPPS during FY 2002, the cost reporting period used in calculating the FY 2006 wage index. We will reconsider our policy when we can collect and analyze wage data for a base year that could be impacted by Pub. L. 108-173 changes for CAHs.

Data collected for the IPPS wage index are also currently used to calculate wage indices applicable to other providers, such as SNFs, home health agencies, and hospices. In addition, they are used for prospective payments to rehabilitation, psychiatric, and long-term care hospitals, and for hospital outpatient services. We note that in the IPPS rules, we do not address comments pertaining to the wage indices for non-PPS providers. Such comments should be made in response to separate proposed rules for those providers.

In the FY 2005 IPPS final rule, we stated that a commenter had asked CMS to designate provider-based clinics as IPPS-excluded areas in order to remove the costs from the wage index (69 FR 49049). The commenter noted that provider-based clinics are like physician private offices, which are excluded from the wage index calculation, and that services provided in the provider-based clinics are paid for not through the IPPS, but rather under the hospital outpatient PPS. In response to the comment, we stated that we were not prepared to grant the commenter's request without first studying the issue, and that we would explore the matter of salaries related to provider-based clinics in a future rule.

Regulations at 42 CFR 413.65 describe the criteria and procedures for determining whether a facility or organization is provider-based. Historically, under the Medicare program, some providers, referred to as ``main providers,'' have functioned as single entities while owning and operating multiple provider-based departments, locations, and facilities that are treated as part of the main provider for Medicare purposes. Section 413.65(a)(2) defines various types of provider-based facilities, including ``department of a provider.'' A ``department of a provider'' means a facility or organization that is either created by, or acquired by, a main provider for the purposes of furnishing health care services of the same type as those furnished by the main provider under the name, ownership, and financial and administrative control of the main provider * * * a department of a provider may not itself be qualified to participate in Medicare as a provider under Sec. 489.2 * * * the term `department of a provider' does not include an RHC or * * * an FQHC.'' Thus, if a facility offers services that are similar to those provided in a freestanding physician's office, and the

[[Page 47371]]

facility meets the criteria to become provider-based under Sec. 413.65, the facility would be considered a ``department of a provider.'' More specifically, the hospital would integrate the facility into the main provider's outpatient department, since the facility offers health care services of the same type as those furnished by the main provider. In addition, because a physician's office would not receive its own provider agreement or receive a Medicare provider number under Sec. 489.2 unlike an FQHC or an RHC, it cannot be considered a ``provider-based entity,'' rather it would be considered a department of a provider. (We note that a provider-based RHC or FQHC may, by itself, be qualified to participate in Medicare as a provider under Sec. 489.2 and, thus, would be classified not as a ``department of a provider'' but as a ``provider-based entity,'' as defined at Sec. 413.65(a)(2).) This provider-based facility, or provider-based clinic, as the commenter referred to it, would be reported on the main provider's Medicare cost report as an outpatient service cost center, on Worksheet A, line 60. With the exception of RHC and FQHC salaries that have been excluded from the wage index beginning with FY 2004 (68 FR 45395), the salaries attributable to employees working in these outpatient service cost centers, including emergency departments, are included in the main provider's total salaries on Worksheet S-3, Part II, line 1, and accordingly, are included in the wage index calculation. We have historically included the salaries and wages of hospital employees working in the outpatient departments in the calculation of the hospital wage index since these employees often work in both the IPPS and in the outpatient areas of the hospital. Consistent with this longstanding treatment of outpatient salary costs in the wage index calculation, we believe it is appropriate to continue to include the salaries and wages of employees working in outpatient departments, including provider-based clinics, in the wage index calculation.

Comment: Two commenters objected to our clarification of historical policy that the salaries of employees working in provider-based clinics should continue to be included in the wage index calculation. The commenters referred to these facilities as ``hospital-owned provider- based physician practices'' that may be qualified to participate in Medicare as providers under Sec. 489.2 of the regulations, and therefore, by definition, are not ``departments of a provider.'' They argued that CMS should exclude ``hospital-owned provider-based physician practices'' from the wage index because, similar to RHCs and FQHCs, the services provided by these facilities are also not paid for under the IPPS. The commenters alluded to the OIG 2004 Red Book (October 22, 2004), which proposed that CMS should eliminate provider- based designations for ``hospital-owned physician practices,'' since hospitals treat these facilities as provider-based without CMS' approval. The commenters questioned whether it would be ``more accurate and practical'' to exclude all ``hospital-owned provider-based physician practices'' from the wage index, in light of the OIG's proposal. Lastly, the commenters asserted that CMS' statement that the salaries and wages of hospital employees working in the outpatient areas of the hospital have historically been included in the wage index since those employees often work both in the inpatient and outpatient areas of the hospital, is inaccurate with respect to ``hospital-owned provider-based physician practices'' because the facilities do not provide services to IPPS areas of the hospital.

Response: In the FY 2006 IPPS proposed rule (70 FR 23371), we discussed whether to include the costs of provider-based clinics in the wage index because we stated in a response to a comment in the FY 2005 IPPS final rule (69 FR 49049) that we would explore the matter in a future rule. Thus, we considered the issue and concluded that it is appropriate to include the salaries and hours of employees working in provider-based clinics in the wage index calculation. We came to this conclusion because provider-based clinics cannot qualify by themselves to participate in Medicare as a provider under Sec. 489.2 of the regulations and are, therefore, categorized as ``departments of a provider'' under the provider-based regulations at Sec. 413.65. Accordingly, they would be reported as part of the main provider's outpatient department on line 60 of Worksheet C of the Medicare cost report. In making this conclusion, we distinguished provider-based clinics that are part of the hospital outpatient department and included in the IPPS wage index from ``provider-based entities'' (such as SNFs, RHCs, and FQHCs) that are excluded from the IPPS wage index because, under the regulations at Sec. 413.65, they participate in Medicare under their own provider agreements. Commenters are incorrect when they asserted that RHCs and FQHCs would be included in the wage index except for the fact that these entities are not paid under the IPPS. Rather, wage data from RHCs and FQHCs are also not included in the wage index because, although they may be provider-based, these entities are providers in their own right and may, by themselves, qualify to participate in Medicare as a provider under Sec. 489.2. As a general rule, we do not include the wage data of facilities that are providers in their own right in the IPPS wage index. Thus, the commenters are also incorrect that ``hospital-owned provider-based physician practices'' may, by themselves, be qualified to participate in Medicare as a provider under Sec. 489.2 of the regulations, and therefore, by definition, are not ``departments of a provider.'' We note that Sec. 489.2 does not list ``hospital-owned provider-based physician practice'' as one of the facilities that may participate in Medicare as a provider. Further, while Sec. 489.2 does list ``clinics'' as a type of facility that can participate in Medicare as a provider, Sec. 489.2(c) specifies that only clinics that furnish outpatient physical therapy and speech pathology services may qualify as providers. Therefore, if a hospital wishes that a physician practice be considered provider-based, the physician practice, by definition, must be categorized as part of hospital outpatient departments. As such, the services provided in these provider-based clinics are paid for by Medicare under the OPPS. Accordingly, it is appropriate that the salaries and hours attributable to the provider-based clinics are included in the IPPS wage index.

In response to the commenters' speculation as to whether it would be ``more accurate and practical'' to exclude all ``hospital-owned provider-based physician practices'' from the wage index, in light of the OIG's proposal in the OIG 2004 Red Book (October 22, 2004), we believe the commenter is confusing the policies regarding (a) who should be considered provider-based and (b) whether salaries and hours associated with provider-based clinics should be included in the wage index. These are two different policy matters. On the first policy, we agree that firm oversight and consistent audit procedures for determining and monitoring provider-based status are necessary, since our existing payment systems provide for more generous payment to hospital outpatient departments than similar freestanding facilities. However, the proposed rule discussed the second matter, not the first. The purpose of the discussion in the proposed rule was not to debate

[[Page 47372]]

whether physician practices should ever be considered for provider- based status. Certainly, we agree that freestanding physician offices, or facilities that have been denied provider-based status by the CMS Regional Office, should not be included in the wage index. Rather, the purpose of the discussion in the proposed rule was to clarify our longstanding policy that as long as a hospital reports, and the CMS Regional Office approves, that a facility which might formerly have been a freestanding physician office is provider-based, the proper categorization of such a facility is as an outpatient department and the wages and hours attributable to that outpatient department are included in the IPPS wage index. Thus, we believe the commenters' reference to the OIG Red Book is misplaced.

Further, the commenters' provide no support for their assertion that workers in ``hospital-owned provider-based physician practices'' do not provide services to IPPS areas of the hospital. We have not seen any evidence suggesting that the employees working in provider-based clinics work exclusively there, or in other outpatient areas of the hospital. Furthermore, we believe it would be extremely complicated and unnecessary to attempt to distinguish between the salaries and hours of employees that work in the various outpatient areas of hospitals, for purposes of computing the IPPS wage index. Hospitals often maintain provider-based facilities since the Medicare payment for services provided in a hospital (provider-based) setting is typically more than the payment would be for the same service provided in a freestanding setting. Hospitals should not be permitted to treat these facilities as part of the hospital for one purpose, and separate from the hospital for purposes of the wage index. If hospitals wish to exclude certain facilities from the wage index, they have the option to do so by converting them to freestanding facilities. Therefore, as stated in the FY 2006 proposed rule, consistent with our longstanding policy, we continue to believe that it is appropriate to include the salaries and hours of employees working in the outpatient departments, including provider-based clinics, in the wage index calculation.

Comment: Two commenters expressed concern that the data used in calculating the wage index are developed inconsistently across the Nation. One of the commenters stressed the need for consistent interpretation and application of all wage index policies by all fiscal intermediaries. The commenters did not provide any specific examples of cases where wage index data is developed inconsistently, or where intermediaries are interpreting wage index policies inconsistently.

Response: We are equally concerned about consistent interpretation and application of wage index policies by both intermediaries and hospitals, as the wage index is a relative measure of area wage differences. Throughout the years, we have revised and refined our policy statements and cost reporting instructions in order to achieve more accurate reporting of wage and hours data among hospitals and intermediaries. In addition, we seek to close any loopholes in our policies that may result in varied applications among hospitals. Our work to ensure accuracy and consistency in the wage index is a continuous effort. We encourage hospitals and intermediaries to bring to our attention any instances of perceived inconsistencies. Also, we remind hospitals that the wage data correction process is another mechanism that is available for hospitals that require CMS' intervention to settle disputes with intermediaries over wage index policy interpretations (see section III.J. of this preamble).

E. Verification of Worksheet S-3 Wage Data

The wage data for the proposed FY 2006 wage index were obtained from Worksheet S-3, Parts II and III of the FY 2002 Medicare cost reports. Instructions for completing the Worksheet S-3, Parts II and III are in the Provider Reimbursement Manual, Part I, sections 3605.2 and 3605.3. The data file used to construct the wage index includes FY 2002 data as of June 30, 2005. As in past years, we performed an intensive review of the wage data, mostly through the use of edits designed to identify aberrant data.

We asked our fiscal intermediaries to revise or verify data elements that resulted in specific edit failures. While most of the edit failures were resolved, we did remove the wage data of some hospitals from the final FY 2006 wage index. For the final FY 2006 wage index in this final rule, we removed the data for 235 hospitals from our database: 201 hospitals became CAHs between February 20, 2004, the cutoff date for exclusion of CAHs from the FY 2005 wage index, and February 18, 2005, this year's cutoff date for the exclusion of CAHS from the FY 2006 wage index, and 27 hospitals were low Medicare utilization hospitals or failed edits that could not be corrected because the hospitals terminated the program or changed ownership. In addition, we removed the wage data for 7 hospitals with incomplete or inaccurate data resulting in zero or negative, or otherwise aberrant, average hourly wages. As a result, the final FY 2006 wage index is calculated based on FY 2002 wage data from 3,742 hospitals.

In constructing the FY 2006 wage index, we include the wage data for facilities that were IPPS hospitals in FY 2002, even for those facilities that have since terminated their participation in the program as hospitals, as long as those data do not fail any of our edits for reasonableness. We believe that including the wage data for these hospitals is, in general, appropriate to reflect the economic conditions in the various labor market areas during the relevant past period. However, we exclude the wage data for CAHs (as discussed in 68 FR 45397). The wage index in this final rule excludes hospitals that are designated as CAHs by February 1, 2005, the date of the latest available Medicare CAH listing at the time we released the proposed wage index public use file (PUF) on February 25, 2005.

Comment: Two commenters expressed concern that the wage data for two CAHs would not be removed from the final FY 2006 wage index. The commenters explained that the effective date for conversion to CAH status for both providers was in December 2004, but because of the timing of the notification of the CAH status, the providers' wage data was included in the February 25, 2005 PUF, and in Tables 2 and 4A that accompanied publication of the proposed rule. The commenters noted that, although CMS subsequently removed these providers' wage data from the May 6, 2005 PUF, their wage data continued to be included in the revised Table 2 that was posted June 1, 2005 on the CMS Web site. The commenters asked for assurance that the wage data for these two CAHs would not be included in the final FY 2006 wage index.

Response: As stated in the FY 2004 IPPS final rule (68 FR 45398), we exclude providers from the wage index that were designated as CAHs by 7 or more days prior to the posting of the preliminary PUF. This year, since the preliminary PUF was posted on February 25, 2005, we excluded providers that were designated as CAHs by February 18, 2005. These hospitals were both designated as CAHs prior to February 18, 2005, and should not be included in the FY 2006 wage index calculations. The commenters are correct that, initially, we did not receive notification of the providers' CAH status in time to remove their wage data from the February 25, 2005 PUF. We did not

[[Page 47373]]

include their wage data in the May 6, 2005 PUF. However, these hospitals continued to be included in the updated Table 2 posted on the CMS Web site on June 1, 2005 because these revisions to the wage data were based on the February 25, 2005 PUF. However, the data for these two CAHs are not included in the FY 2006 final wage index calculations. We note that these two providers will continue to appear on Table 2 published along with the FY 2006 final rule because, although no average hourly wage will be listed next to these providers for FY 2006, they did have wage data that contributed to the wage index for their CBSA in FY 2004 and FY 2005.

F. Computation of the FY 2006 Unadjusted Wage Index

The method used to compute the FY 2006 wage index without an occupational mix adjustment follows:

Step 1--As noted above, we based the FY 2006 wage index on wage data reported on the FY 2002 Medicare cost reports. We gathered data from each of the non-Federal, short-term, acute care hospitals for which data were reported on the Worksheet S-3, Parts II and III of the Medicare cost report for the hospital's cost reporting period beginning on or after October 1, 2001 and before October 1, 2002. In addition, we included data from some hospitals that had cost reporting periods beginning before October 2001 and reported a cost reporting period covering all of FY 2002. These data were included because no other data from these hospitals would be available for the cost reporting period described above, and because particular labor market areas might be affected due to the omission of these hospitals. However, we generally describe these wage data as FY 2002 data. We note that, if a hospital had more than one cost reporting period beginning during FY 2002 (for example, a hospital had two short cost reporting periods beginning on or after October 1, 2001 and before October 1, 2002), we included wage data from only one of the cost reporting periods, the longer, in the wage index calculation. If there was more than one cost reporting period and the periods were equal in length, we included the wage data from the later period in the wage index calculation.

Step 2--Salaries--The method used to compute a hospital's average hourly wage excludes certain costs that are not paid under the IPPS. In calculating a hospital's average salaries plus wage-related costs, we subtracted from Line 1 (total salaries) the GME and CRNA costs reported on Lines 2, 4.01, 6, and 6.01, the Part B salaries reported on Lines 3, 5 and 5.01, home office salaries reported on Line 7, and excluded salaries reported on Lines 8 and 8.01 (that is, direct salaries attributable to SNF services, home health services, and other subprovider components not subject to the IPPS). We also subtracted from Line 1 the salaries for which no hours were reported. To determine total salaries plus wage-related costs, we added to the net hospital salaries the costs of contract labor for direct patient care, certain top management, pharmacy, laboratory, and nonteaching physician Part A services (Lines 9 and 10), home office salaries and wage-related costs reported by the hospital on Lines 11 and 12, and nonexcluded area wage- related costs (Lines 13, 14, and 18).

We note that contract labor and home office salaries for which no corresponding hours are reported were not included. In addition, wage- related costs for nonteaching physician Part A employees (Line 18) are excluded if no corresponding salaries are reported for those employees on Line 4.

Step 3--Hours--With the exception of wage-related costs, for which there are no associated hours, we computed total hours using the same methods as described for salaries in Step 2.

Step 4--For each hospital reporting both total overhead salaries and total overhead hours greater than zero, we then allocated overhead costs to areas of the hospital excluded from the wage index calculation. First, we determined the ratio of excluded area hours (sum of Lines 8 and 8.01 of Worksheet S-3, Part II) to revised total hours (Line 1 minus the sum of Part II, Lines 2, 3, 4.01, 5, 5.01, 6, 6.01, 7, and Part III, Line 13 of Worksheet S-3). We then computed the amounts of overhead salaries and hours to be allocated to excluded areas by multiplying the above ratio by the total overhead salaries and hours reported on Line 13 of Worksheet S-3, Part III. Next, we computed the amounts of overhead wage-related costs to be allocated to excluded areas using three steps: (1) We determined the ratio of overhead hours (Part III, Line 13) to revised hours (Line 1 minus the sum of Lines 2, 3, 4.01, 5, 5.01, 6, 6.01, 7, 8, and 8.01); (2) we computed overhead wage-related costs by multiplying the overhead hours ratio by wage- related costs reported on Part II, Lines 13, 14, and 18; and (3) we multiplied the computed overhead wage-related costs by the above excluded area hours ratio. Finally, we subtracted the computed overhead salaries, wage-related costs, and hours associated with excluded areas from the total salaries (plus wage-related costs) and hours derived in Steps 2 and 3.

Step 5--For each hospital, we adjusted the total salaries plus wage-related costs to a common period to determine total adjusted salaries plus wage-related costs. To make the wage adjustment, we estimated the percentage change in the employment cost index (ECI) for compensation for each 30-day increment from October 14, 2001 through April 15, 2003 for private industry hospital workers from the Bureau of Labor Statistics' Compensation and Working Conditions. We use the ECI because it reflects the price increase associated with total compensation (salaries plus fringes) rather than just the increase in salaries. In addition, the ECI includes managers as well as other hospital workers. This methodology to compute the monthly update factors uses actual quarterly ECI data and assures that the update factors match the actual quarterly and annual percent changes. The factors used to adjust the hospital's data were based on the midpoint of the cost reporting period, as indicated below.

Midpoint of Cost Reporting Period

Adjustment After

Before factor

10/14/2001..................................... 11/15/2001 1.06469 11/14/2001..................................... 12/15/2001 1.06007 12/14/2001..................................... 1/15/2002 1.05566 01/14/2002..................................... 02/15/2002 1.05139 02/14/2002..................................... 03/15/2002 1.04725 03/14/2002..................................... 04/15/2002 1.04317 04/14/2002..................................... 05/15/2002 1.03907 05/14/2002..................................... 06/15/2002 1.03496 06/14/2002..................................... 07/15/2002 1.03083 07/14/2002..................................... 08/15/2002 1.02672 08/14/2002..................................... 09/15/2002 1.02261 09/14/2002..................................... 10/15/2002 1.01860 10/14/2002..................................... 11/15/2002 1.01478 11/14/2002..................................... 12/15/2002 1.01116 12/14/2002..................................... 01/15/2003 1.00757 01/14/2003..................................... 02/15/2003 1.00385 02/14/2003..................................... 03/15/2003 1.00000 03/14/2003..................................... 04/15/2003 0.99613

For example, the midpoint of a cost reporting period beginning January 1, 2002 and ending December 31, 2002 is June 30, 2002. An adjustment factor of 1.03083 would be applied to the wages of a hospital with such a cost reporting period. In addition, for the data for any cost reporting period that began in FY 2002 and covered a period of less than 360 days or more than 370 days, we annualized the data to reflect a 1-year cost report. Dividing the data by the number of days in the cost report and then multiplying the results by 365 accomplishes annualization.

[[Page 47374]]

Step 6--Each hospital was assigned to its appropriate urban or rural labor market area before any reclassifications under section 1886(d)(8)(B), section 1886(d)(8)(E), or section 1886(d)(10) of the Act. Within each urban or rural labor market area, we added the total adjusted salaries plus wage-related costs obtained in Step 5 for all hospitals in that area to determine the total adjusted salaries plus wage-related costs for the labor market area.

Step 7--We divided the total adjusted salaries plus wage-related costs obtained under both methods in Step 6 by the sum of the corresponding total hours (from Step 4) for all hospitals in each labor market area to determine an average hourly wage for the area.

Step 8--We added the total adjusted salaries plus wage-related costs obtained in Step 5 for all hospitals in the nation and then divided the sum by the national sum of total hours from Step 4 to arrive at a national average hourly wage. Using the data as described above, the national average hourly wage is $28.0011.

Step 9--For each urban or rural labor market area, we calculated the hospital wage index value by dividing the area average hourly wage obtained in Step 7 by the national average hourly wage computed in Step 8.

Step 10--Following the process set forth above, we developed a separate Puerto Rico-specific wage index for purposes of adjusting the Puerto Rico standardized amounts. (The national Puerto Rico standardized amount is adjusted by a wage index calculated for all Puerto Rico labor market areas based on the national average hourly wage as described above.) We added the total adjusted salaries plus wage-related costs (as calculated in Step 5) for all hospitals in Puerto Rico and divided the sum by the total hours for Puerto Rico (as calculated in Step 4) to arrive at an overall average hourly wage of $12.8063 for Puerto Rico. For each labor market area in Puerto Rico, we calculated the Puerto Rico-specific wage index value by dividing the area average hourly wage (as calculated in Step 7) by the overall Puerto Rico average hourly wage.

Step 11--Section 4410 of Pub. L. 105-33 provides that, for discharges on or after October 1, 1997, the area wage index applicable to any hospital that is located in an urban area of a State may not be less than the area wage index applicable to hospitals located in rural areas in that State. (For all-urban States, we established an imputed floor (69 FR 49109). Furthermore, this wage index floor is to be implemented in such a manner as to ensure that aggregate IPPS payments are not greater or less than those that would have been made in the year if this section did not apply. For FY 2006, this change affects 174 hospitals in 63 urban areas. The areas affected by this provision are identified by a footnote in Table 4A in the Addendum of this final rule.

Comment: Numerous commenters were concerned with the proposed change in step 4 of the wage index calculation in the FY 2006 IPPS proposed rule (70 FR 23373). In order to allocate overhead wage-related costs to areas of a hospital that are excluded from the IPPS, CMS uses three steps: (1) Determine the ratio of overhead hours to revised (that is, allowable) hours; (2) compute overhead wage-related costs by multiplying the overhead hours ratio from Step 1 by wage-related costs; and (3) multiply the overhead wage-related costs from Step 2 by the excluded hours ratio (see Step 4 for more detail). The commenters noted that, for FY 2006, the calculation of the overhead hours ratio in Step 1 will be modified to subtract hours attributable to excluded areas (from line 8 for SNFs and line 8.01 for excluded areas of Worksheet S- 3, Part II of the Medicare cost report). The commenters observed that this change results in a higher overhead hours ratio, which, in turn, results in a greater amount of overhead cost being allocated to excluded areas. The commenters believed that, because more costs are being allocated to excluded areas, a hospital's average hourly wage would decrease as a result of the proposal. One commenter added that the proposed methodology is flawed, but did not indicate why. Other commenters stated that the excluded area overhead hours ratio computed with CMS' proposed methodology is ``dramatically high'' and does not accurately reflect the hospital's overhead costs attributable to its employee benefit amounts, but they did not offer an explanation or an alternative for accurately identifying excluded overhead costs.

In general, the commenters, including the national hospital association, were concerned that the proposed rule did not discuss the impact of the change, and did not include a lengthy discussion of the changes. These commenters believed that CMS should postpone the change until a lengthy discussion of the proposal can be included in a future proposed rule. The commenters further recommended that, because the change in the wage index calculation caused confusion among hospitals as to the correct average hourly wages, hospitals should be given an opportunity to withdraw or reinstate their requests for geographic reclassification within 30 days of the publication of the final rule.

Response: We have carefully considered the comments we received regarding the proposed change in the FY 2006 IPPS proposed rule to the methodology for removing overhead wage-related costs attributable to areas of the hospital excluded from the IPPS. Overall, commenters seemed to be more concerned that the proposed rule did not contain a detailed discussion of the modification, rather than disagreeing in principle with our modification. Therefore, we are adopting our proposal without modifications because we believe the proposal most accurately calculates the overhead wage-related costs that are attributable to excluded areas. Historically, the wage index used to adjust a hospital's payment under the IPPS has only reflected costs of services that are provided in areas of the hospital that are covered under the IPPS. That is, because certain areas of a hospital are specifically excluded from the IPPS, such as hospital-based SNFs, or distinct part rehabilitation and psychiatric units, the proportion of the salaries paid to and the hours worked by employees in areas of the hospital excluded from the IPPS are identified and removed from the hospital's total salaries and hours. The remaining allowable salaries and hours are used to compute the hospital's average hourly wage, which, in turn, is used to calculate the wage index for the labor market area in which the hospital is located.

In addition to removing salaries and hours that are directly attributable to employees working in excluded areas, for each hospital reporting both total overhead salaries and total overhead hours greater than zero, we also remove any overhead (administrative and general) costs and hours attributable to excluded areas by allocating overhead costs and hours between the IPPS areas of the hospital and the areas of the hospital excluded from the IPPS. We do this by determining the ``excluded rate'' for each hospital, which is the ratio of excluded area hours to total hours (see Step 4 of the wage index calculation). The ``excluded rate'' reflects the percentage of hours worked by hospital employees in areas of the hospital excluded from the IPPS. For example, an ``excluded rate'' of 0.15 means that approximately 15 percent of total employee hours was spent in excluded areas (and therefore, about 85 percent of the employees' time worked was spent in the IPPS areas of the hospital). We then determine the amount of overhead salaries and hours to be allocated to the excluded areas by taking the ``excluded rate'' and multiplying it by the

[[Page 47375]]

hospital's total salaries and hours attributable to overhead.

Next, because wage-related costs are separate from salaries, we perform a similar calculation to determine the percentage of wage- related costs attributable to overhead that should be allocated to the excluded areas of the hospital. We do this by computing the ``overhead rate,'' which is the percentage of allowable (that is, does not include excluded area) overhead hours to total hours. The ``overhead rate'' is multiplied by total wage-related costs to determine the amount of wage- related cost attributable to overhead. Finally, the amount of wage- related costs attributable to overhead is multiplied by the ``excluded rate'' to determine the amount of overhead wage-related costs that are associated with excluded areas, and, therefore, should be subtracted from the total allowable wages used in the wage index. Obviously, the larger the ``overhead rate,'' the greater the amount of overhead wage- related costs to be allocated across the hospital, and the greater the excluded area, the greater the amount of overhead wage-related cost that is identified as being associated with excluded areas and that should be subtracted from allowable wages.

Through FY 2005, in determining the ``overhead rate,'' we divided the allowable overhead hours by the hospital's total hours, including hours attributable to excluded areas, even though the latter hours are excluded from the wage index. Last year, after publication of the FY 2005 IPPS final rule, we became aware of the mismatch between the numerator and the denominator in the ``overhead rate'' calculation. Specifically, because the numerator in the ``overhead rate'' calculation does not include excluded area overhead hours, and the denominator in the ``overhead rate'' calculation does include the hours attributable to excluded areas, this results in an understatement of the amount of wage-related costs attributable to overhead that should be allocated to the excluded areas. That is, because we had not completely removed the amount of wage-related cost attributable to excluded areas from the denominator, the ``overhead rate'' was lower than it should be. A lower ``overhead rate'' has the unintended effect of artificially raising a hospital's average hourly wage because a lower amount of overhead attributable to excluded areas is removed from total allowable salaries. To the extent that a hospital has a higher ``excluded rate'' (that is, they provide a significant amount of services that are not covered under the IPPS, and therefore, have a high percentage of employee hours related to the excluded areas), this issue is more significant. For example, in the case of one hospital with an ``excluded rate'' of 96 percent, under the FY 2005 calculation, we identified (and removed) only 40 percent of the overhead wage- related costs as being attributable to excluded areas, whereas under the FY 2006 calculation, 93 percent of the hospital's overhead wage- related costs has been identified as being attributable to excluded areas, and therefore, are being removed for the FY 2006 wage index. Clearly, in the case of this hospital which predominantly provides services that are excluded from the IPPS, it is logical that the vast majority of its overhead costs are attributable to excluded areas of the hospital as well, and, therefore, these overhead costs should be removed from the hospital's average hourly wage used to determine the IPPS wage index.

Accordingly, in order to correct the discrepancy between the numerator and the denominator in the overhead rate calculation, and to correct the understatement of the excluded overhead wage-related costs, we believe that it is more appropriate to determine the amount of overhead wage-related costs associated with excluded areas that should be excluded from the wage index based on the ratio of allowable costs to allowable hours (that is, only hours related to IPPS areas of the hospital). Specifically, we are not including the hours associated with excluded areas in the denominator of the ``overhead rate'' calculation. While hospitals with small excluded areas relative to their IPPS areas should be minimally affected by the removal of the excluded area hours from the calculation, this change will serve to lower the average hourly wages of hospitals with relatively large excluded areas, more closely aligning them with costs allowed under the IPPS.

We believe that, despite the absence of a lengthy discussion of the policy in the proposed rule, the change in the overhead wage-related cost allocation noted in the FY 2006 IPPS proposed rule (70 FR 23373) provided hospitals with adequate notice of the change. Hospitals are sufficiently sophisticated to understand the implications of a proposal to exclude certain lines on the cost report from its calculations. In addition, the Average Hourly Wage Calculator, which included the revised overhead wage-related cost allocation, has been available on our Web site: http://www.cms.hhs.gov/providers/hipps/ippswage.asp since

shortly after the proposed rule went on public display on April 24, 2005. The tables included with the FY 2006 proposed rule also showed the average hourly wages and the wage indices resulting from the proposed modification. Finally, clearly the fact that a hospital association and other commenters provided comments on the proposal demonstrates that hospitals had actual notice of the change. Some commenters even computed the effect of the change on the calculation of their wage indices for FY 2006. In addition, even if some hospitals might object that they did not understand the change included in the FY 2006 proposed rule, we believe that the detailed steps used in calculating the wage index are interpretive rules that are not subject to the notice and comment rulemaking procedures of the Administrative Procedure Act. Clearly, we do not include each of the detailed steps and lines from the cost report in our regulations at Sec. 412.64(h), the section of the regulations requiring CMS to adjust the ``proportion of the Federal rate for inpatient operating costs that are attributable to wages and labor-related costs for area differences in hospital wage levels by a [wage index] factor.'' For these reasons, we believe that we have provided sufficient notice of the change in the ``overhead rate'' calculation.

Commenters are correct that some hospitals that wish to reclassify for FY 2007 could also be affected by decreased average hourly wages. However, we have analyzed our data, and have found that the impact of the change is limited. Specifically, approximately 42 hospitals in 11 labor market areas are receiving a decrease of 1.0 to 5.5 percent in their FY 2006 wage index as a result of this change in the calculation. These labor market areas are primarily in the New England and East North Central census regions. In addition, 10 rural hospitals and 18 urban hospitals are experiencing a decrease in their average hourly wages of between 10 percent and 45 percent. However, the ``excluded rates'' for these hospitals range between 74 percent up to and including 100 percent. While we note that CMS did provide a 30-day period after publication of the FY 2005 IPPS final rule (69 FR 49066) allowing hospitals to reconsider their geographic reclassification decisions, we provided this opportunity because of the number of changes between the proposed and final rules and the apparent confusion regarding application of the section 505 out-migration adjustment. We do not believe a similar extension is warranted in this case. Further, we do not agree that a 30-day window after publication of the final rule is necessary in order to

[[Page 47376]]

allow cancellations or reinstatements of reclassifications. The FY 2006 proposed rule change was clearly reflected in the wage index tables accompanying the proposed rule. Thus, hospitals were well aware of their proposed average hourly wages and proposed wage indices for FY 2006. Hospitals could review these wage tables, find the proposed average hourly wage and wage index listed for the hospital and wage area, and on the basis of such information, determine whether they wished to withdraw or retain a certain reclassification. Because of such notice, there is no need to provide a subsequent 30-day period for withdrawal or reinstatement. Further, we note that hospitals could use the Average Hourly Wage Calculator on the CMS Web site to determine exactly how the revised methodology affected the wage index. For the reasons stated above, we are finalizing our proposed decision to remove the excluded area hours on lines 8 and 8.01 from the overhead wage- related cost allocation.

G. Computation of the FY 2006 Blended Wage Index

For the final FY 2005 wage index, we used a blend of the occupational mix adjusted wage index and the unadjusted wage index. Specifically, we adjusted 10 percent of the FY 2005 wage index adjustment factor by a factor reflecting occupational mix. Given that 2003-2004 was the first time for the administration of the occupational mix survey, hospitals had a short timeframe for collecting their occupational mix survey data and documentation, the wage data were not in all cases from a 1-year period, and there was no baseline data for purposes of developing a desk review program, we found it prudent not to adjust the entire wage index factor by the occupational mix. However, we did find the data sufficiently reliable for applying an adjustment to 10 percent of the wage index. We found the data reliable because hospitals were given an opportunity to review their survey data and submit changes in the Spring of 2004, hospitals were already familiar with the BLS OES survey categories, hospitals were required to be able to provide documentation that could be used by fiscal intermediaries to verify survey data, and the results of our survey were consistent with the findings of the 2001 BLS OES survey, especially for nursing and physical therapy categories. In addition, we noted that we were moving cautiously with implementing the occupational mix adjustment in recognition of changing trends in hiring nurses, the largest group in the survey. We noted that some States had recently established floors on the minimum level of registered nurse staffing in hospitals in order to maintain licensure. In addition, in some rural areas, we believed that hospitals might be accounting for shortages of physicians by hiring more registered nurses. (A complete discussion of the FY 2005 wage index adjustment factor can be found in section III.G. of the FY 2005 IPPS final rule (69 FR 49052).)

In the FY 2005 final rule, we noted that while the statute required us to collect occupational mix data every 3 years, the statute does not specify how the occupational mix adjustment is to be constructed or applied. We are clarifying in this final rule that the October 1, 2004 deadline for implementing an occupational mix adjustment is not codified in section 1886(d)(3)(E) of the Act, which requires only a collection and measurement of occupational mix data, but rather stems from the effective date provisions in section 304(c) of the Medicare, Medicaid and SCHIP Benefits Improvement and Protection Act of 2000, Pub. L. 106-554 (BIPA). Although we believe that applying the occupational mix to 10 percent of the wage index factor fully implements the occupational mix adjustment, we also interpret BIPA as requiring only that we begin applying an adjustment by October 1, 2004. BIPA required the Secretary to complete, ``by not later than September 30, 2003, for application beginning October 1, 2004,'' both the collection of occupational mix data and the measurement of such data. (BIPA, section 304(c)(3).) Thus, even if adjusting 10 percent of the wage index for occupational mix were not (as we believe it to be) considered to be full implementation of the BIPA effective date, we certainly began our application of the adjustment as of October 1, 2004.

In addition, section 1886(d)(3)(E) of the Act provides broad authority for us to establish the factor we use to adjust hospital costs to take into account area differences in wage levels. The statute is clear that the wage index factor is to be ``established by the Secretary.'' The occupational mix is only one part of this wage index factor, which, for the most part, is calculated on the basis of average hourly wage data submitted by all hospitals in the United States. In exercising the Secretary's broad discretion to establish the factor that adjusts for geographic wage differences, in FY 2005 we adjusted 10 percent of such factor to account for occupational mix.

Indeed, we have often used percentage figures or blended amounts in exercising the Secretary's authority to establish the factor that adjusts for wage differences. For example, in the FY 2005 final rule, we implemented new mapping boundaries for assigning hospitals to the geographic labor market areas used for calculating the wage index. For hospitals that were harmed by the new geographic boundaries, we used a blended rate based on 50 percent of the wage index that would apply using the new geographic boundaries effective for FY 2005 and 50 percent of the wage index that would apply using the old geographic boundaries that were effective during FY 2004 (69 FR 49033). Similarly, beginning with FY 2000, we began phasing out costs related to GME and CRNAs from the wage index (64 FR 41505). Thus, for example, the FY 2001 wage index was based on a blend of 60 percent of an average hourly wage including these costs, and 40 percent of an average hourly wage excluding these costs (65 FR 47071).

As we proposed in the FY 2006 IPPS proposed rule, for FY 2006, we are again adjusting 10 percent of the wage index factor for the occupational mix. In computing the occupational mix adjustment for the final FY 2006 wage index, we used the occupational mix survey data that we collected for the FY 2005 wage index, replacing the survey data for 20 hospitals that submitted revised data, and excluding the survey data for hospitals with no corresponding Worksheet S-3 wage data for FY 2006 wage index. While we considered adjusting 100 percent of the wage index by the occupational mix, we did not believe it was appropriate to use first-year survey data to make such a large adjustment. As hospitals gain additional experience with the occupational mix survey, and as we develop more information upon which to audit the data we receive, we expect to increase the portion of the wage index that is adjusted.

As we did in the proposed rule, we also acknowledge the finding of the District Court opinion in Bellevue Hospital Center v. Leavitt, No. 04-8639 (S.D.N.Y, March 2005). Given that the Government has appealed the occupational mix portion of that decision, we believe it is appropriate to continue with our policy of adopting the policy we believe to be most prudent for occupational mix.

With 10 percent of the FY 2006 wage index adjusted for occupational mix, the national average hourly wage is $28.0037 and the Puerto Rico specific average hourly wage is $12.8055. The wage index values for 13 rural areas (27.7 percent) and 201 urban areas (52.1 percent) would decrease as a result of

[[Continued on page 47377]]

From the Federal Register Online via GPO Access [wais.access.gpo.gov] ]

[[pp. 47377-47426]] Medicare Program; Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2006 Rates

[[Continued from page 47376]]

[[Page 47377]]

the adjustment. These decreases would be minimal; the largest negative impact for a rural area would be 0.18 percent and for an urban area, 0.43 percent. Conversely, 31 rural areas (66.0 percent) and 176 urban areas (45.6 percent) would benefit from this adjustment, with 1 urban area increasing 2.2 percent and 1 rural area increasing 0.37 percent. As there are no significant differences between the FY 2005 and the FY 2006 occupational mix survey data and results, we believe it is appropriate to again apply the occupational mix to 10 percent of the final FY 2006 wage index. (See Appendix A to this final rule for further analysis of the impact of the occupational mix adjustment on the final FY 2006 wage index.)

Comment: Most commenters supported our proposal to adjust only 10 percent of the FY 2006 wage index for occupational mix. However, one commenter requested CMS to implement the occupational mix adjustment in a way that ensures that the adjustment does not negatively impact his hospital and other similar hospitals, providing no further elaboration for his suggestion, while two other commenters opposed applying any occupational mix adjustment at all until CMS performs a new survey. In contrast, a few commenters representing hospitals that would benefit from a 100 percent occupational mix adjustment to the wage index recommended the policy that would most behoove them (that is, a full implementation of the adjustment for the FY 2006 wage index). These commenters supported their proposal by noting that: (1) For FY 2006, hospitals were given an opportunity to revise or correct data originally submitted; (2) occupational mix data from FY 2005 were consistent with registered nurse and licensed practical nurse data from a AHA annual survey of hospitals; and (3) Congress intended for 100 percent of the wage index to be adjusted for occupational mix beginning October 1, 2004.

Response: We do not agree with the commenters recommending elimination of the occupational mix adjustment. As we stated in the proposed rule, given the FY 2005 and FY 2006 wage indices were based on the first year of survey data, as well as other stated considerations (see 70 FR 23375), we found survey results sufficiently robust to support an adjustment to 10 percent of the wage index, but did not believe it prudent to adjust the entire wage index by occupational mix. We refer readers to the proposed rule for a full discussion of our rationale. We continue to believe that the data are sufficient to support applying the occupational mix to 10 percent of the wage index. Moreover, we believe that by implementing the wage index in this manner, we are carrying out the Congressional requirement to begin applying an occupational mix to the wage index by October 1, 2004.

We do not agree with commenters that stated that the correction of data permitted for FY 2006 is sufficient to allow for a 100 percent adjustment in FY 2006. While hospitals were permitted to correct their data for FY 2006, only 20 out of the 3,541 hospitals did so. Further, the fact that hospitals were permitted to submit corrected data does not alleviate concerns that (a) the data continued to be derived from the first year of an occupational mix survey; or (b) that CMS had no historical baseline data for developing a robust audit program for such data. Given such concerns, we also believe it would be neither equitable nor appropriate to adjust 100 percent of the wage index when the occupational mix benefits hospitals, but 10 percent of the wage index when it does not. Instead, we continue to believe that the proposed, more moderate occupational mix adjustment is the most equitable and appropriate approach. As such, the FY 2006 wage index in this final rule is a blend of 10 percent of a wage index adjusted for occupational mix and 90 percent of an unadjusted wage index.

Comment: One commenter expressed concern regarding CMS' statement in the proposed rule that ``hospitals might be accounting for shortages of physicians by hiring more registered nurses'' (70 FR 23375). The commenter suggested that the statement is unsupported and implies a ``practice of downgrading care, especially since it uses 'registered nurses', not even nurse practitioners.'' The commenter requested that we delete the statement from the final rule.

Response: We did not intend to imply that hospitals that have increased their reliance on registered nurses provide downgraded care. Nursing schools and nursing associations acknowledge a significant increase in the number of registered nurses who are pursuing or have achieved advanced practice degrees as nurse practitioners, clinical nurse specialists, nurse midwives, and certified registered nurse anesthetists. Our statement merely acknowledged that hiring advanced practice registered nurses helps to mitigate problems with physician shortages by increasing the number of staff who are available to provide primary care, and that such hiring practices may have contributed to the higher than expected occupational mix reported by many rural hospitals.

The wage index values for FY 2006 (except those for hospitals receiving wage index adjustments under section 505 of Pub. L. 108-173) are shown in Tables 4A, 4B, 4C, and 4F in the Addendum to this final rule.

Tables 3A and 3B in the Addendum to this final rule list the 3-year average hourly wage for each labor market area before the redesignation of hospitals, based on FYs 2004, 2005, 2006 cost reporting periods. Table 3A lists these data for urban areas and Table 3B lists these data for rural areas. In addition, Table 2 in the Addendum to this final rule includes the adjusted average hourly wage for each hospital from the FY 2000 and FY 2001 cost reporting periods, as well as the FY 2002 period used to calculate the FY 2006 wage index. The 3-year averages are calculated by dividing the sum of the dollars (adjusted to a common reporting period using the method described previously) across all 3 years, by the sum of the hours. If a hospital is missing data for any of the previous years, its average hourly wage for the 3-year period is calculated based on the data available during that period.

The wage index values in Tables 4A, 4B, 4C, and 4F and the average hourly wages in Tables 2, 3A, and 3B in the Addendum to this final rule include the occupational mix adjustment.

Other Public Comments

Comment: One commenter stated that an ongoing concern is that the hospital wage index is applied to many provider types for which wage data are excluded from the wage index calculation. The commenter recommended that CMS separate wage indices for SNFs, IRFs, and IPFs by modifying the way the wage index data are reported on the Medicare cost report.

Response: We appreciate the comment, but note that the subject- matter of this final rule is the IPPS system and not the PPSs governing non-IPPS entities such as SNFs, IRFs, and IPFs. Therefore, we are not responding to this comment at this time. We suggest that the commenter raise his or her concerns as part of the rulemaking process for updating the respective facility's PPS.

H. Revisions to the Wage Index Based on Hospital Redesignation

1. General

Under section 1886(d)(10) of the Act, the Medicare Geographic Classification Review Board (MGCRB) considers applications by hospitals for geographic reclassification for purposes of payment under the IPPS. Hospitals must apply to

[[Page 47378]]

the MGCRB to reclassify by September 1 of the year preceding the year during which reclassification is sought. Generally, hospitals must be proximate to the labor market area to which they are seeking reclassification and must demonstrate characteristics similar to hospitals located in that area. The MGCRB issues its decisions by the end of February for reclassifications that become effective for the following fiscal year (beginning October 1). The regulations applicable to reclassifications by the MGCRB are located in Sec. Sec. 412.230 through 412.280.

Section 1886(d)(10)(D)(v) of the Act provides that, beginning with FY 2001, a MGCRB decision on a hospital reclassification for purposes of the wage index is effective for 3 fiscal years, unless the hospital elects to terminate the reclassification. Section 1886(d)(10)(D)(vi) of the Act provides that the MGCRB must use the 3 most recent years' average hourly wage data in evaluating a hospital's reclassification application for FY 2003 and any succeeding fiscal year.

Section 304(b) of Pub. L. 106-554 provides that the Secretary must establish a mechanism under which a statewide entity may apply to have all of the geographic areas in the State treated as a single geographic area for purposes of computing and applying a single wage index, for reclassifications beginning in FY 2003. The implementing regulations for this provision are located at Sec. 412.235.

Section 1886(d)(8)(B) of the Act requires the Secretary to treat a hospital located in a rural county adjacent to one or more urban areas as being located in the MSA to which the greatest number of workers in the county commute, if the rural county would otherwise be considered part of an urban area under the standards for designating MSAs and if the commuting rates used in determining outlying counties were determined on the basis of the aggregate number of resident workers who commute to (and, if applicable under the standards, from) the central county or counties of all contiguous MSAs. In light of the new CBSA definitions and the Census 2000 data that we implemented for FY 2005 (69 FR 49027), we undertook to identify those counties meeting these criteria. The eligible counties are identified under section III.H.5. of this preamble. 2. Effects of Reclassification

Section 1886(d)(8)(C) of the Act provides that the application of the wage index to redesignated hospitals is dependent on the hypothetical impact that the wage data from these hospitals would have on the wage index value for the area to which they have been redesignated. These requirements for determining the wage index values for redesignated hospitals is applicable both to the hospitals located in rural counties deemed urban under section 1886(d)(8)(B) of the Act and hospitals that were reclassified as a result of the MGCRB decisions under section 1886(d)(10) of the Act. Therefore, as provided in section 1886(d)(8)(C) of the Act,\9\ the wage index values were determined by considering the following:

\9\ Although section 1886(d)(8)(C)(iv)(I) of the Act also provides that the wage index for an urban area may not decrease as a result of redesignated hospitals if the urban area wage index is already below the wage index for rural areas in the State in which the urban area is located, the provision was effectively made moot by section 4410 of Pub. L. 105-33, which provides that the area wage index applicable to any hospital that is located in an urban area of a State may not be less than the area wage index applicable to hospitals located in rural areas in that State. For all-urban States, CMS established an imputed floor (69 FR 49109). Also, section 1886(d)(8)(C)(iv)(II) of the Act provides that an urban area's wage index may not decrease as a result of redesignated hospitals if the urban area is located in a State that is composed of a single urban area.

If including the wage data for the redesignated hospitals would reduce the wage index value for the area to which the hospitals are redesignated by 1 percentage point or less, the area wage index value determined exclusive of the wage data for the redesignated hospitals applies to the redesignated hospitals.

If including the wage data for the redesignated hospitals reduces the wage index value for the area to which the hospitals are redesignated by more than 1 percentage point, the area wage index determined inclusive of the wage data for the redesignated hospitals (the combined wage index value) applies to the redesignated hospitals.

If including the wage data for the redesignated hospitals increases the wage index value for the urban area to which the hospitals are redesignated, both the area and the redesignated hospitals receive the combined wage index value. Otherwise, the hospitals located in the urban area receive a wage index excluding the wage data of hospitals redesignated into the area.

The wage data for a reclassified urban hospital is included in both the wage index calculation of the area to which the hospital is reclassified (subject to the rules described above) and the wage index calculation of the urban area where the hospital is physically located.

Rural areas whose wage index values would be reduced by excluding the wage data for hospitals that have been redesignated to another area continue to have their wage index values calculated as if no redesignation had occurred (otherwise, redesignated rural hospitals are excluded from the calculation of the rural wage index).

The wage index value for a redesignated rural hospital cannot be reduced below the wage index value for the rural areas of the State in which the hospital is located. 3. Application of Hold Harmless Protection for Certain Urban Hospitals Redesignated as Rural

Section 401(a) of Pub. L. 106-113 (the Balanced Budget Refinement Act of 1999) amended section 1886(d)(8) of the Act by adding paragraph (E). Section 401(a) created a mechanism that permits an urban hospital to apply to the Secretary to be treated, for purposes of subsection (d), as being located in the rural area of the State in which the hospital is located. A hospital that is granted redesignation under section 1886(d)(8)(E) of the Act, as added by section 401 of Pub. L. 106-113, is therefore treated as a rural hospital for all purposes of payment under the Medicare IPPS, including the standardized amount, wage index, and disproportionate share calculations as of the effective date of the redesignation. Under current policy, as a result of an approved redesignation of an urban hospital as a rural hospital, the wage index data are excluded from the wage index calculation for the area where the urban hospital is geographically located and included in the rural hospital wage index calculation.

Last year, we became aware of an instance where the approved redesignation of an urban hospital as rural under section 1886(d)(8)(E) of the Act resulted in the hospital's data having an adverse impact on the rural wage index. We received a public comment noting that specific ``hold harmless'' provisions apply to reclassifications that occur under section 1886(d)(8)(B) and section 1886(d)(10) of the Act. That is, if a hospital is granted geographic reclassification under section 1886(d)(8)(B) or section 1886(d)(10) of the Act, there are certain rules that apply when the inclusion of the hospital's data results in a reduction of the reclassification area's wage index, and these rules are slightly different for urban areas versus rural areas. These rules are more fully described in the FY 2005 IPPS final rule (69 FR 49053). Generally stated, these rules prevent a rural area from being adversely affected as a result of reclassification. That is, if excluding the reclassifying hospitals'

[[Page 47379]]

wage data would decrease the wage index of the rural area, the reclassifying hospitals are included in the rural area's wage index. Otherwise, the reclassifying hospitals are excluded. For hospitals reclassifying out of urban areas, the rules provide that the wage data for the reclassified urban hospital are included in the wage index calculation of the urban area where the hospital is physically located.

The commenter recommended that we revise our regulations and apply similar hold harmless provisions and treat hospitals redesignated under section 1886(d)(8)(E) of the Act in the same manner as reclassifications under section 1886(d)(8)(B) and section 1886(d)(10) of the Act. In our continued effort to promote consistency, equity and to simplify our rules with respect to how we construct the wage indexes of rural and urban areas, we are persuaded that there is a need to modify our policy when hospital redesignations occur under section 1886(d)(8)(E) of the Act. Therefore, for the FY 2006 wage index, in the FY 2006 IPPS proposed rule, we proposed to apply the hold harmless rule that currently applies when rural hospitals are reclassifying out of the rural area (from rural to urban) to situations where hospitals are reclassifying into the rural area (from urban to rural under section 1886(d)(8)(E) of the Act). Thus, the rule would be that the wage data of the urban hospital reclassifying into the rural area are included in the rural area's wage index, if including the urban hospital's data increase the wage index of the rural area. Otherwise, the wage data are excluded. Similarly, we proposed to apply to these cases the rule that currently applies when urban hospitals reclassify under the MGCRB process. Thus, the wage data for an urban hospital reclassifying under section 1886(d)(8)((E) of the Act are always included in the wage index of the urban area where the hospital is located, and can also be included in the wage index of the rural area to which it is reclassifying (if doing so increases the rural area's wage index). In the FY 2006 IPPS proposed rule, we stated that we believe this proposal provides uniformity in the way geographic areas are treated under all types of reclassifications. In addition, we further stated that our proposal promotes predictability by alleviating fluctuations in the wage indexes due to a section 401 redesignation.

No commenters objected to extending hold harmless protection to urban hospitals that are redesignated as rural under section 401. Therefore, in this final rule, we are finalizing the policy to extend hold harmless protection to urban hospitals that are redesignated as rural under section 401.

We are including in the Addendum to this final rule Table 9C, which shows hospitals redesignated under section 1886(d)(8)(E) of the Act. 4. FY 2006 MGCRB Reclassifications

The MGCRB's review of FY 2006 reclassification requests resulted in 299 hospitals approved for wage index reclassifications for FY 2006. Because MGCRB wage index reclassifications are effective for 3 years, hospitals reclassified during FY 2004 or FY 2005 are eligible to continue to be reclassified based on prior reclassifications to current MSAs during FY 2006. There were 395 hospitals reclassified for wage index for FY 2005, and 94 hospitals reclassified for wage index in FY 2004. Some of the hospitals that reclassified in FY 2004 and FY 2005 have elected not to continue their reclassifications in FY 2006 because, under the new labor market area definitions, they are now physically located in the areas to which they previously reclassified. Of all of the hospitals approved for reclassification for FY 2004, FY 2005, and FY 2006, 631 hospitals are in a reclassification status for FY 2006.

Prior to FY 2004, hospitals had been able to apply to be reclassified for purposes of either the wage index or the standardized amount. Section 401 of Pub. L. 108-173 established that all hospitals will be paid on the basis of the large urban standardized amount, beginning with FY 2004. Consequently, all hospitals are paid on the basis of the same standardized amount, which made such reclassifications moot. Although there could still be some benefit in terms of payments for some hospitals under the DSH payment adjustment for operating IPPS, section 402 of Pub. L. 108-173 equalized DSH payment adjustments for rural and urban hospitals, with the exception that the rural DSH adjustment is capped at 12 percent (except that RRCs have no cap). (A detailed discussion of this application appears in section IV.I. of the preamble of the FY 2005 IPPS final rule (69 FR 49085).

Under Sec. 412.273, hospitals that have been reclassified by the MGCRB are permitted to withdraw their applications within 45 days of the publication of a proposed rule. The request for withdrawal of an application for reclassification or termination of an existing 3-year reclassification that would be effective in FY 2005 must be received by the MGCRB within 45 days of the publication of the proposed rule. If a hospital elects to withdraw its wage index application after the MGCRB has issued its decision, but prior to the above date, it may later cancel its withdrawal in a subsequent year and request the MGCRB to reinstate its wage index reclassification for the remaining fiscal year(s) of the 3-year period (Sec. 412.273(b)(2)(i)). The request to cancel a prior withdrawal must be in writing to the MGCRB no later than the deadline for submitting reclassification applications for the following fiscal year (Sec. 412.273(d)). For further information about withdrawing, terminating, or canceling a previous withdrawal or termination of a 3-year reclassification for wage index purposes, we refer the reader to Sec. 412.273, as well as the August 1, 2002, IPPS final rule (67 FR 50065) and the August 1, 2001 IPPS final rule (66 FR 39887).

Changes to the wage index that result from withdrawals of requests for reclassification, wage index corrections, appeals, and the Administrator's review process have been incorporated into the wage index values published in this final rule. These changes may affect not only the wage index value for specific geographic areas, but also the wage index value redesignated hospitals receive; that is, whether they receive the wage index that includes the data for both the hospitals already in the area and the redesignated hospitals. Further, the wage index value for the area from which the hospitals are redesignated may be affected.

Applications for FY 2007 reclassifications are due to the MGCRB by September 1, 2005. We note that this is also the deadline for canceling a previous wage index reclassification withdrawal or termination under Sec. 412.273(d). Applications and other information about MGCRB reclassifications may be obtained, beginning in Mid-July 2005, via the CMS Internet Web site at: http://cms.hhs.gov/providers/prrb/mgcinfo.asp , or by calling the MGCRB at (410) 786-1174. The mailing

address of the MGCRB is: 2520 Lord Baltimore Drive, Suite L, Baltimore, MD 21244-2670. 5. FY 2006 Redesignations Under Section 1886(d)(8)(B) of the Act

Beginning October 1, 1988, section 1886(d)(8)(B) of the Act required us to treat a hospital located in a rural county adjacent to one or more urban areas as being located in the MSA if certain criteria were met. Prior to FY 2005, the rule was that a rural county adjacent to one or more urban areas would be treated as being located in the MSA to which the greatest number of workers in the county commute, if the rural county

[[Page 47380]]

would otherwise be considered part of an urban area under the standards published in the Federal Register on January 3, 1980 (45 FR 956) for designating MSAs (and NECMAs), and if the commuting rates used in determining outlying counties (or, for New England, similar recognized areas) were determined on the basis of the aggregate number of resident workers who commute to (and, if applicable under the standards, from) the central county or counties of all contiguous MSAs (or NECMAs). Hospitals that met the criteria using the January 3, 1980 version of these OMB standards were deemed urban for purposes of the standardized amounts and for purposes of assigning the wage data index.

On June 6, 2003, OMB announced the new CBSAs based on Census 2000 data. For FY 2005, we used OMB's 2000 CBSA standards and the Census 2000 data to identify counties qualifying for redesignation under section 1886(d)(8)(B) for the purpose of assigning the wage index to the urban area. We presented this listing, effective for discharges occurring on or after October 1, 2004 (FY 2005), in Chart 6 of the FY 2005 final rule (69 FR 49057). However, Chart 6 in the FY 2005 final rule contained a printing error in which we misidentified the redesignation areas for two counties that qualified for redesignation under section 1886(d)(8)(B) of the Act. The list of rural counties qualifying to be urban in that Chart 6 incorrectly listed the redesignation CBSAs for Monroe, PA and Walworth, WI. This error was made only in the chart and not in the application of the rules; that is, we correctly applied the rules to the correct rural counties qualifying to be urban for FY 2005.

In addition, we discovered that, in the FY 2005 IPPS final rule, we had erroneously printed the names of the entire Metropolitan Statistical Areas rather than the Metropolitan Division names. Because we recognized Metropolitan Divisions as MSAs in the FY 2005 IPPS final rule (69 FR 49029), we should have printed the division names for the following counties: Henry, FL; Starke, IN; Henderson, TX; Fannin, TX; and Island, WA.

The chart below contains the corrected listing of the rural counties designated as urban under section 1886(d)(8)(B) of the Act that we are using for FY 2006. For discharges occurring on or after October 1, 2005, hospitals located in the first column of this chart will be redesignated for purposes of using the wage index of the urban area listed in the second column.

Comment: Several commenters urged CMS to permit hospitals located in counties redesignated under section 1886(d)(8)(B) of the Act to waive or reject the redesignation if the redesignation proves to be detrimental or otherwise undesirable to the qualifying hospital. They cited examples in which hospitals with special designations, such as rural referral centers, SCHs, MDHs, and CAHs, where their status is dependent on being located in a rural area, lost their special designation when they were reclassified to an urban area under section 1886(d)(8)(B) of the Act.

Response: We considered this comment and are responding to it only insofar as it relates to section 1886(d) hospitals, such as rural referral centers, SCHs, and MDHs, located in Lugar counties. We refer readers to the section on CAHs in this final rule for information on how CMS treats CAHs in Lugar counties. The statute specifically states that ``(f)or purposes of this subsection, the Secretary shall treat a hospital located in a rural county adjacent to one or more urban areas as being located in (a) urban metropolitan statistical area * * *.'' Therefore, all section 1886(d) hospitals located in Lugar counties are deemed urban and such classification cannot be waived, except if a hospital is eligible for an out-migration adjustment. In order for a section 1886(d) hospital to retain its special designation when the area in which it is located is redesignated from rural to urban, a hospital must apply for reclassification under Sec. 412.103(a). We encourage a hospital seeking reclassification under this section to submit a complete application in writing to its CMS Regional Office.

Rural Counties Redesignated as Urban Under Section 1886(d)(8)(B) of the Act [Based on CBSAs and Census 2000 Data]

Rural county

CBSA

Cherokee, AL........................ Rome, GA. Macon, AL........................... Auburn-Opelika, AL. Talladega, AL....................... Anniston-Oxford, AL. Hot Springs, AR..................... Hot Springs, AR. Windham, CT......................... Hartford-West Hartford-East Hartford, CT. Bradford, FL........................ Gainesville, FL. Flagler, FL......................... Deltona-Daytona Beach-Ormond Beach, FL. Hendry, FL.......................... West Palm Beach-Boca Raton-Boynton, FL. Levy, FL............................ Gainesville, FL. Walton, FL.......................... Fort Walton Beach-Crestview-Destin, FL. Banks, GA........................... Gainesville, GA. Chattooga, GA....................... Chattanooga, TN-GA. Jackson, GA......................... Atlanta-Sandy Springs-Marietta, GA. Lumpkin, GA......................... Atlanta-Sandy Springs-Marietta, GA. Morgan, GA.......................... Atlanta-Sandy Springs-Marietta, GA. Peach, GA........................... Macon, GA. Polk, GA............................ Atlanta-Sandy Springs-Marietta, GA. Talbot, GA.......................... Columbus, GA-AL. Bingham, ID......................... Idaho Falls, ID. Christian, IL....................... Springfield, IL. DeWitt, IL.......................... Bloomington-Normal, IL. Iroquois, IL........................ Kankakee-Bradley, IL. Logan, IL........................... Springfield, IL. Mason, IL........................... Peoria, IL. Ogle, IL............................ Rockford, IL. Clinton, IN......................... Lafayette, IN. Henry, IN........................... Indianapolis, IN. Spencer, IN......................... Evansville, IN-KY.

[[Page 47381]]

Starke, IN.......................... Gary, IN. Warren, IN.......................... Lafayette, IN. Boone, IA........................... Ames, IA. Buchanan, IA........................ Waterloo-Cedar Falls, IA. Cedar, IA........................... Iowa City, IA. Allen, KY........................... Bowling Green, KY. Assumption Parish, LA............... Baton Rouge, LA. St. James Parish, LA................ Baton Rouge, LA. Allegan, MI......................... Holland-Grand Haven, MI. Montcalm, MI........................ Grand Rapids-Wyoming, MI. Oceana, MI.......................... Muskegon-Norton Shores, MI. Shiawassee, MI...................... Lansing-East Lansing, MI. Tuscola, MI......................... Saginaw-Saginaw Township North, MI. Fillmore, MN........................ Rochester, MN. Dade, MO............................ Springfield, MO. Pearl River, MS..................... Gulfport-Biloxi, MS. Caswell, NC......................... Burlington, NC. Granville, NC....................... Durham, NC. Harnett, NC......................... Raleigh-Cary, NC. Lincoln, NC......................... Charlotte-Gastonia-Concord, NC-SC. Polk, NC............................ Spartanburg, NC. Los Alamos, NM...................... Santa Fe, NM. Lyon, NV............................ Carson City, NV. Cayuga, NY.......................... Syracuse, NY. Columbia, NY........................ Albany-Schenectady-Troy, NY. Genesee, NY......................... Rochester, NY. Greene, NY.......................... Albany-Schenectady-Troy, NY. Schuyler, NY........................ Ithaca, NY. Sullivan, NY........................ Poughkeepsie-Newburgh-Middletown, NY. Wyoming, NY......................... Buffalo-Niagara Falls, NY. Ashtabula, OH....................... Cleveland-Elyria-Mentor, OH. Champaign, OH....................... Springfield, OH. Columbiana, OH...................... Youngstown-Warren-Boardman, OH-PA. Cotton, OK.......................... Lawton, OK. Linn, OR............................ Corvallis, OR. Adams, PA........................... York-Hanover, PA. Clinton, PA......................... Williamsport, PA. Greene, PA.......................... Pittsburgh, PA. Monroe, PA.......................... Allentown-Bethlehem-Easton, PA-NJ. Schuylkill, PA...................... Reading, PA. Susquehanna, PA..................... Binghamton, NY. Clarendon, SC....................... Sumter, SC. Lee, SC............................. Sumter, SC. Oconee, SC.......................... Greenville, SC. Union, SC........................... Spartanburg, SC. Meigs, TN........................... Cleveland, TN. Bosque, TX.......................... Waco, TX. Falls, TX........................... Waco, TX. Fannin, TX.......................... Dallas-Plano-Irving, TX. Grimes, TX.......................... College Station-Bryan, TX. Harrison, TX........................ Longview, TX. Henderson, TX....................... Dallas-Plano-Irving, TX. Milam, TX........................... Austin-Round Rock, TX. Van Zandt, TX....................... Dallas-Plano-Irving, TX. Willacy, TX......................... Brownsville-Harlingen, TX. Buckingham, VA...................... Charlottesville, VA. Floyd, VA........................... Blacksburg-Christiansburg-Radford, VA. Middlesex, VA....................... Virginia Beach-Norfolk-Newport News, VA. Page, VA............................ Harrisonburg, VA. Shenandoah, VA...................... Winchester, VA-WV. Island, WA.......................... Seattle-Bellevue-Everett, WA. Mason, WA........................... Olympia, WA. Wahkiakum, WA....................... Longview, WA. Jackson, WV......................... Charleston, WV. Roane, WV........................... Charleston, WV. Green, WI........................... Madison, WI. Green Lake, WI...................... Fond du Lac, WI. Jefferson, WI....................... Milwaukee-Waukesha-West Allis, WI. Walworth, WI........................ Milwaukee-Waukesha-West Allis, WI.

[[Page 47382]]

As in the past, hospitals redesignated under section 1886(d)(8)(B) of the Act are also eligible to be reclassified to a different area by the MGCRB. Affected hospitals were permitted to compare the reclassified wage index for the labor market area in Table 4C in the Addendum of the May 4, 2005 proposed rule into which they have been reclassified by the MGCRB to the wage index for the area to which they are redesignated under section 1886(d)(8)(B) of the Act. Hospitals were provided the opportunity to withdraw from an MGCRB reclassification within 45 days of the publication of the FY 2006 IPPS proposed rule (May 4, 2005). 6. Reclassifications Under Section 508 of Pub. L. 108-173

Under section 508 of Pub. L. 108-173, a qualifying hospital could appeal the wage index classification otherwise applicable to the hospital and apply for reclassification to another area of the State in which the hospital is located (or, at the discretion of the Secretary, to an area within a contiguous State). We implemented this process through notices published in the Federal Register on January 6, 2004 (69 FR 661) and February 13, 2004 (69 FR 7340). Such reclassifications are applicable to discharges occurring during the 3-year period beginning April 1, 2004 and ending March 31, 2007. Under section 508(b), reclassifications under this process do not affect the wage index computation for any area or for any other hospital and cannot be effected in a budget neutral manner.

Comment: Some commenters indicated that hospitals currently receiving a section 508 reclassification are eligible to reclassify to that same area under the standard reclassification process as a result of the new labor market definitions that we adopted for FY 2005. The commenters pointed out that the governing regulations indicate that ``if a hospital is already reclassified to a given geographic area for wage index purposes for a 3-year period, and submits an application to the same area for either the second or third year of the 3-year period, that application will not be approved.'' These commenters expressed concern that the MGCRB will deny these hospitals reclassification for FY 2007 if there is no change in the regulations to address this issue.

Response: We appreciate the commenters' interest in this matter. Hospitals that indicate in their MGCRB applications that they agree to waive their section 508 reclassification for the first 6 months of FY 2007 if they are granted a 3-year reclassification under the traditional MGCRB process will not be subject to the regulation cited above. Thus, in applying for a 3-year MGCRB reclassification beginning in FY 2007, hospitals that are already reclassified to the same area under section 508 should indicate in their MGCRB reclassification requests that if they receive the MGCRB reclassification, they will forfeit the section 508 reclassification for the first 6 months of FY 2007.

Comment: Many commenters expressed concern regarding the timing overlaps between section 508 of Pub. L. 108-173 and the FY 2007 reclassifications. The commenters pointed out that section 508 of Pub. L. 108-173 required the Secretary to develop a one-time special reclassification procedure that allowed hospitals meeting specified criteria to be reclassified from April 1, 2004, through March 31, 2007. They further stated that some hospitals that qualified for reclassification under section 508 may qualify for geographic reclassification under one of the opportunities available under the regulations in 42 CFR part 412, subpart L. Because pending reclassifications will expire in the middle of a Federal fiscal year, the commenters requested that CMS clarify when the hospitals should apply for reclassification under an opportunity under subpart L. Commenters stated that, unless CMS establishes an accommodation for section 508 hospitals, hospitals will be confronted with a difficult dilemma: Forfeiting 6 months of section 508 reclassification to be able to reclassify for FY 2007; or postponing reclassification until FY 2008 and being without reclassification for the 6 months between April 1 and September 30, 2007. The commenters believed that both of these options would carry significant financial consequences for hospitals. The commenters urged CMS to implement a solution that does not require hospitals to make such a difficult choice, and would provide them with the full benefits of the section 508 reclassification.

Response: We appreciate the commenters' suggestions and their interest in this matter. Under 1886(d)(10)(D)(v) of the Act, CMS has the authority to ``establish procedures'' under which a hospital may elect to terminate a reclassification before the end of a 3-year period. Based on comments and on a careful review of the statute, we have decided to exercise this authority to establish a procedural rule for section 508 hospitals to retain their section 508 reclassification through its expiration on March 31, 2007 and reclassify under a subpart L opportunity for the second half of FY 2007. The following procedural rules will apply for section 508 hospitals that wish to reclassify for the second half of FY 2007:

For section 508 hospitals applying for individual reclassification under 42 CFR 412.230--

(1) Hospitals must apply for reclassification through the MGCRB by the September 1, 2005 deadline.

(2) Section 508 hospitals that are approved by the MGCRB for reclassification will have 45 days from the date the FY 2007 IPPS proposed rule is published to cancel their section 1886(d)(10) reclassifications for either the first 6 months of FY 2007 or for the entire fiscal year. Hospitals should note that if they fail to cancel their section 1886(d)(10) reclassification by the deadline, they will not receive their section 508 wage adjustment in FY 2007. To further clarify--

Hospitals that cancel their section 1886(d)(10) reclassification for the first 6 months receive their section 508 reclassifications for October 2006 through March 2007 and their section 1886(d)(10) reclassifications for April through September 2007.

Hospitals that cancel their section 1886(d)(10) reclassification for the entire year will receive their section 508 reclassification for October 2006 through March 2007 and their home area wage index for April through September 2007.

Hospitals that do not cancel their section 1886(d)(10) reclassifications will receive their section 1886(d)(10) reclassification, not their section 508 reclassification, for the entire fiscal year.

Hospital groups that include a section 508 hospital would also be permitted to submit section 1886(d)(10) reclassification applications by the September 1, 2005 deadline. However, in order for a group reclassification to be approved, either of the following conditions would need to be met:

(1) The section 508 hospital that is part of the group must waive its section 508 reclassification for the first half of FY 2007. This is necessary because the regulations at Sec. Sec. 412.232 and 412.234 state that all hospitals in a county must apply for reclassification as a group. The hospitals either agree to receive the same reclassification or they fail to qualify as a group. The Administrator upheld this policy in an MGCRB appeal for FY 2006.

(2) Each member of the group agrees in writing, at the time the application is submitted September 1, 2005, that they cancel the group reclassification if granted for the first 6 months of FY 2007. The section 1886(d)(10) reclassification will be effective only

[[Page 47383]]

April through September 2007. Under this scenario, the section 508 hospital receives its section 508 reclassification from October 2006 through March 2007 and the remainder of the group receives the home wage index for that time period. For April through September 2007, the section 508 hospital and the remainder of the group receive the group reclassification. The group will have the opportunity to cancel the April through September 2007 group reclassification within 45 days of publication of the proposed rule.

We would apply a similar rule for purposes of the out-migration adjustment. The statute states that a hospital cannot receive an out- migration adjustment if it is simultaneously reclassified under section 1886(d)(10) of the Act. Therefore, hospitals that are not reclassified during any part of FY 2007 will, by default, receive an out-migration adjustment during that time period.

We show the reclassifications effective under the one-time appeal process in Table 9B in the Addendum to this final rule.

I. FY 2006 Wage Index Adjustment Based on Commuting Patterns of Hospital Employees

In accordance with the broad discretion under section 1886(d)(13) of the Act, as added by section 505 of Pub. L. 108-173, beginning with FY 2005, we established a process to make adjustments to the hospital wage index based on commuting patterns of hospital employees. The process, outlined in the FY 2005 IPPS final rule (69 FR 49061), provides for an increase in the wage index for hospitals located in certain counties that have a relatively high percentage of hospital employees who reside in the county but work in a different county (or counties) with a higher wage index. Such adjustments to the wage index are effective for 3 years, unless a hospital requests to waive the application of the adjustment. A county will not lose its status as a qualifying county due to wage index changes during the 3-year period, and counties will receive the same wage index increase for those 3 years. However, a county that qualifies in any given year may no longer qualify after the 3-year period, or it may qualify but receive a different adjustment to the wage index level. Hospitals that receive this adjustment to their wage index are not eligible for reclassification under section 1886(d)(8) or section 1886(d)(10) of the Act. Adjustments under this provision are not subject to the IPPS budget neutrality requirements under section 1886(d)(3)(E) or section 1886(d)(8)(D) of the Act.

Comment: One commenter proposed that CMS allow hospitals that reclassify and receive a diluted wage index to receive the out- migration adjustment provided it does not exceed the actual wage index for the area to which they are reclassified.

Response: The statute specifically states that hospitals that receive an out-migration adjustment are ineligible for reclassification under section 1886(d)(8) or section 1886(d)(10) of the Act.

Hospitals located in counties that qualify for the wage index adjustment will receive an increase in the wage index that is equal to the average of the differences between the wage indices of the labor market area(s) with higher wage indices and the wage index of the resident county, weighted by the overall percentage of hospital workers residing in the qualifying county who are employed in any labor market area with a higher wage index. We have employed the prereclassified wage indices in making these calculations.

Hospitals located in the qualifying counties identified in Table 4J in the Addendum to this final rule that have not already reclassified through section 1886(d)(10) of the Act, redesignated through section 1886(d)(8) of the Act, received a section 508 reclassification, or requested to waive the application of the out-migration adjustment will receive the wage index adjustment listed in the table for FY 2006. We used the same formula described in the FY 2005 final rule (69 FR 49064) to calculate the out-migration adjustment. This adjustment was calculated as follows:

Step 1. Subtract the wage index for the qualifying county from the wage index for the higher wage area(s).

Step 2. Divide the number of hospital employees residing in the qualifying county who are employed in such higher wage index area by the total number of hospital employees residing in the qualifying county who are employed in any higher wage index area. Multiply this result by the result obtaining in Step 1.

Step 3. Sum the products resulting from Step 2 (if the qualifying county has workers commuting to more than one higher wage area).

Step 4. Multiply the result from Step 3 by the percentage of hospital employees who are residing in the qualifying county and who are employed in any higher wage index area.

The adjustments calculated for qualifying hospitals are listed in Table 4J in the Addendum to this final rule. These adjustments are effective for each county for a period of 3 fiscal years. Hospitals that received the adjustment in FY 2005 will be eligible to retain that same adjustment for FY 2006 and FY 2007. For hospitals in newly qualified counties, adjustments to the wage index are effective for 3 years, beginning with discharges occurring on or after October 1, 2005.

As previously noted, hospitals receiving the wage index adjustment under section 1886(d)(13)(F) of the Act are not eligible for reclassification under sections 1886(d)(8) or (d)(10) of the Act, or under section 508 of Pub. L. 108-173, unless they waive such out- migration adjustment. As announced in the FY 2005 final rule as well as the proposed rule for FY 2006, hospitals redesignated under section 1886(d)(8) of the Act or reclassified under section 1886(d)(10) of the Act or under section 508 of Pub. L. 108-173 were deemed to have chosen to retain their redesignation or reclassification, unless they explicitly notified CMS that they elected to receive the out-migration adjustment instead within 45 days from the publication of the FY 2006 IPPS proposed rule (May 4, 2005). Under Sec. 412.273, hospitals that have been reclassified by the MGCRB were permitted to terminate existing 3-year reclassifications within 45 days of the May 4, 2005 proposed rule. Hospitals that are eligible to receive the out-migration wage index adjustment and that withdraw their application for reclassification automatically receive the wage index adjustment listed in Table 4J in the Addendum to this final rule. Requests for withdrawal of an application for reclassification or termination of an existing 3- year reclassification will be effective in FY 2006 and had to have been received by the MGCRB within 45 days of the publication of the FY 2006 IPPS proposed rule. Requests to waive section 1886(d)(8) redesignations for FY 2006 had to have been received by CMS within 45 days of the publication of the FY 2006 IPPS proposed rule. In addition, hospitals that wished to retain their redesignation/reclassification under section 1886(d)(8), section 1886(d)(10), or section 508 (instead of receiving the out-migration adjustment) for FY 2006 did not need to submit a formal request to CMS; they automatically retain their redesignation/reclassification status for FY 2006.

Comment: Commenters expressed opposition to and support of CMS' interpretation of the law that hospitals will receive the same out- migration adjustment in each of the 3 years of eligibility for the adjustment. One

[[Page 47384]]

commenter recommended that CMS maintain its policy to keep the out- migration adjustment unchanged to minimize uncertainties and instability in Medicare reimbursement to hospitals. Other commenters recommended that CMS revise its policy so that the out-migration adjustment will be recalculated each year based on updated wage data and the new wage indices.

Response: We appreciate the comments we received regarding this issue. The governing statute specifically states that the wage index increase ``shall be effective for a period of 3 fiscal years.'' We have interpreted this to mean that the adjustment shall be identical for 3 years. If we were to recalculate the out-migration adjustment each year based on updated wage data as suggested, counties could potentially be deemed ineligible for the wage index adjustment if the average hourly wage for all hospitals in the labor market area exceeded the average hourly wages for all hospitals in the county. Therefore, we have elected to maintain our policy to keep the out-migration adjustment associated with a particular county unchanged.

Comment: One commenter requested that we clarify the removal of several providers from Table 4J between the May 4, 2005 Federal Register publication and the revised table posted on the CMS Web site on June 1, 2005.

Response: There were some errors for CBSAs and imputed rural floors and these errors had an effect on the out-migration calculations shown in Table 4J of the proposed rule. We posted the corrected adjustments on the CMS Web site on June 1, 2005. Hospitals were also notified of the corrected out-migration adjustments via the Listserv and a Hospital Open Door Forum on June 2, 2005.

Comment: Commenters requested that CMS make available the hospital commuting data used to compute the out-migration adjustment.

Response: We plan to make the data used for determining the qualifying counties and the out-migration adjustment available after the publication of this final rule on the CMS Web site at: http://www.cms.gov .

Comment: Commenters requested that CMS implement a policy similar to the policy established for FY 2005 that allows hospitals to withdraw or reinstate their geographic applications within 30 days of the date that the final rule is published. Several commenters believed there is still a likelihood that revisions made between the proposed and final rules may affect a hospital's choice of whether to accept the out- migration or a reclassification.

Response: First, we note that cancellation and reinstatement rules for geographic reclassifications are procedural rules that are not subject to notice and comment rulemaking. Second, we note that it has been our longstanding policy that our procedural rules on withdrawals or terminations of reclassifications require such terminations and withdrawals be made within 45 days of the proposed rule (Sec. 412.273). However, FY 2005 was an exceptional circumstance due to the extensive changes to the wage index as a result of our adoption of the new labor market areas. We noted that this was a limited circumstance, and we did not expect to extend the withdrawal date beyond 45 days after the proposed rule in future years. We do not believe the exceptional circumstance that existed for FY 2005 exists for FY 2006, given the changes to the labor market areas have been adopted. Therefore, we are continuing with our longstanding policy that terminations of reclassifications are required to be made within 45 days of the proposed rule. As we have explained in previous preamble discussions (see, for example, 56 FR 43241, August 30, 1991), the 45- day deadline provides a reasonable time to take withdrawals or terminations into account in developing the final wage index and prospective payment rates.

J. Requests for Wage Index Data Corrections

In the FY 2005 IPPS final rule (68 FR 27194), we revised the process and timetable for application for development of the wage index, beginning with the FY 2005 wage index. The preliminary and unaudited Worksheet S-3 wage data and occupational mix survey files were made available on October 8, 2004 through the Internet on the CMS Web site at: http://cms.hhs.gov/providers/hipps/ippswage.asp. In a

memorandum dated October 6, 2004, we instructed all Medicare fiscal intermediaries to inform the IPPS hospitals they service of the availability of the wage index data files and the process and timeframe for requesting revisions (including the specific deadlines listed below). We also instructed the fiscal intermediaries to advise hospitals that these data are also made available directly through their representative hospital organizations.

If a hospital wished to request a change to its data as shown in the October 8, 2004 wage and occupational mix data files, the hospital was to submit corrections along with complete, detailed supporting documentation to its fiscal intermediary by November 29, 2004. Hospitals were notified of this deadline and of all other possible deadlines and requirements, including the requirement to review and verify their data as posted on the preliminary wage index data file on the Internet, through the October 6, 2004 memorandum referenced above.

In the October 6, 2004 memorandum, we also specified that a hospital could only request revisions to the occupational mix data for the reporting period that the hospital used in its original FY 2005 wage index occupational mix survey. That is, a hospital that submitted occupational mix data for the 12-month reporting period could not switch to submitting data for the 4-week reporting period and vice versa. Further, a hospital could not submit an occupational mix survey for the periods beginning before January 1, 2003, or after January 11, 2004. In addition, a hospital that did not submit an occupational mix survey for the FY 2005 wage index was not permitted to submit a survey for the FY 2006 wage index.

The fiscal intermediaries notified the hospitals by mid-February 2005 of any changes to the wage index data as a result of the desk reviews and the resolution of the hospitals' late November 2004 change requests. The fiscal intermediaries also submitted the revised data to CMS by mid-February 2005. CMS published the proposed wage index public use files that included hospitals' revised wage data on February 25, 2005. In a memorandum also dated February 25, 2005, we instructed fiscal intermediaries to notify all hospitals regarding the availability of the proposed wage index public use files and the criteria and process for requesting corrections and revisions to the wage index data. Hospitals had until March 14, 2005 to submit requests to the fiscal intermediaries for reconsideration of adjustments made by the fiscal intermediaries as a result of the desk review, and to correct errors due to CMS's or the fiscal intermediary's mishandling of the wage index data. Hospitals were also required to submit sufficient documentation to support their requests.

After reviewing requested changes submitted by hospitals, fiscal intermediaries transmitted any additional revisions resulting from the hospitals' reconsideration requests by April 15, 2005. The deadline for a hospital to request CMS intervention in cases where the hospital disagreed with the fiscal intermediary's policy interpretations was April 22, 2005.

Hospitals were also instructed to examine Table 2 in the Addendum to

[[Page 47385]]

the proposed rule. Table 2 of the proposed rule contained each hospital's adjusted average hourly wage used to construct the wage index values for the past 3 years, including the FY 2002 data used to construct the FY 2006 wage index. We noted that the hospital average hourly wages shown in Table 2 only reflected changes made to a hospital's data and transmitted to CMS by February 23, 2005.

The final wage data public use file was released in early May 2005 to hospital associations and the public on the Internet at http:/www.cms.hhs.gov/providers/hipps/ippswage.asp. The May 2005 public use

file was made available solely for the limited purpose of identifying any potential errors made by CMS or the fiscal intermediary in the entry of the final wage data that result from the correction process described above (revisions submitted to CMS by the fiscal intermediaries by April 15, 2005). If, after reviewing the May 2005 final file, a hospital believed that its wage data were incorrect due to a fiscal intermediary or CMS error in the entry or tabulation of the final wage data, it was provided the opportunity to send a letter to both its fiscal intermediary and CMS that outlined why the hospital believed an error exists and to provide all supporting information, including relevant dates (for example, when it first became aware of the error). These requests had to be received by CMS and the fiscal intermediaries by no later than June 10, 2005. The fiscal intermediary reviewed requests upon receipt and contacted CMS immediately to discuss its findings.

After the release of the May 2005 wage index data file, changes to the hospital wage data were only made in those very limited situations involving an error by the fiscal intermediary or CMS that the hospital could not have known about before its review of the final wage index data file. Specifically, neither the intermediary nor CMS accepted the following types of requests:

Requests for wage data corrections that were submitted too late to be included in the data transmitted to CMS by fiscal intermediaries on or before April 15, 2005.

Requests for correction of errors that were not, but could have been, identified during the hospital's review of the February 25, 2005 wage index data file.

Requests to revisit factual determinations or policy interpretations made by the fiscal intermediary or CMS during the wage index data correction process.

Verified corrections to the wage index received timely by CMS and the fiscal intermediaries (that is, by June 10, 2005) have been incorporated into the final wage index of this final rule and are effective October 1, 2005.

We created the processes described above to resolve all substantive wage index data correction disputes before we finalize the wage and occupational mix data for the FY 2006 payment rates. Accordingly, hospitals that did not meet the procedural deadlines set forth above will not be afforded a later opportunity to submit wage index data corrections or to dispute the fiscal intermediary's decision with respect to requested changes. Specifically, our policy is that hospitals that do not meet the procedural deadlines set forth above will not be permitted to challenge later, before the Provider Reimbursement Review Board, the failure of CMS to make a requested data revision. (See W.A. Foote Memorial Hospital v. Shalala, No. 99-CV- 75202-DT (E.D. Mich. 2001) and Palisades General Hospital v. Thompson, No. 99-1230 (D.D.C. 2003.) We refer the reader also to the FY 2000 final rule (64 FR 41513) for a discussion of the parameters for appealing to the PRRB for wage index data corrections.

Again, we believe the wage index data correction process described above provides hospitals with sufficient opportunity to bring errors in their wage index data to the fiscal intermediaries' attention. Moreover, because hospitals had access to the final wage index data by early May 2005, they had the opportunity to detect any data entry or tabulation errors made by the fiscal intermediary or CMS before the development and publication of the final FY 2006 wage index in this final rule, and the implementation of the FY 2006 wage index on October 1, 2005. If hospitals availed themselves of the opportunities afforded to provide and make corrections to the wage data, the wage index implemented on October 1 should be accurate. Nevertheless, in the event that errors are identified by hospitals and brought to our attention after June 10, 2005, we retain the right to make midyear changes to the wage index under very limited circumstances.

Specifically, in accordance with Sec. 412.64(k)(1) of our existing regulations, we make midyear corrections to the wage index for an area only if a hospital can show that: (1) the fiscal intermediary or CMS made an error in tabulating its data; and (2) the requesting hospital could not have known about the error or did not have an opportunity to correct the error, before the beginning of the fiscal year. For purposes of this provision, ``before the beginning of the fiscal year'' means by the June deadline for making corrections to the wage data for the following fiscal year's wage index. This provision is not available to a hospital seeking to revise another hospital's data that may be affecting the requesting hospital's wage index for the labor market area. As indicated earlier, since CMS makes the wage data available to a hospital on the CMS Web site prior to publishing both the proposed and final IPPS rules, and the fiscal intermediaries notify hospitals directly of any wage data changes after completing their desk reviews, we do not expect that midyear corrections would be necessary. However, under our current policy, if the correction of a data error changes the wage index value for an area, the revised wage index value will be effective prospectively from the date the correction is made.

In the FY 2006 IPPS proposed rule, we proposed to revise Sec. 412.64(k)(2) to specify that a change to the wage index can be made retroactive to the beginning of the Federal fiscal year only when: (1) The fiscal intermediary or CMS made an error in tabulating data used for the wage index calculation; (2) the hospital knew about the error and requested that the fiscal intermediary and CMS correct the error using the established process and within the established schedule for requesting corrections to the wage data, before the beginning of the fiscal year for the applicable IPPS update (that is, by the June 10, 2005 deadline for the FY 2006 wage index); and (3) CMS agreed that the fiscal intermediary or CMS made an error in tabulating the hospital's wage data and the wage index should be corrected. We proposed this change because there may be instances in which a hospital identifies an error in its wage data and submits a correction request using all appropriate procedures and by the June deadline, CMS agrees that the fiscal intermediary or CMS caused the error in the hospital's wage data and that the wage index must be corrected, but CMS fails to publish or implement the corrected wage index value by the beginning of the Federal fiscal year. We made this proposed revision to Sec. 412.64(k)(2) because we believe that it is appropriate and fair. We also believe that, unlike a generalized retroactive policy, the situations where this will occur will be minimal, thus minimizing the administrative burden associated with such retroactive corrections. In those circumstances where a hospital requests a correction to its wage data before CMS calculates the final wage index (that is, by the June deadline), and CMS acknowledges that the error in the

[[Page 47386]]

hospital's wage data caused by CMS's or the fiscal intermediary's mishandling of the data, we believe that the hospital should not be penalized by our delay in publishing or implementing the correction. As with our current policy, we indicated that the proposed provision would not be available to a hospital seeking to revise another hospital's data. In addition, the provision could not be used to correct prior years' wage data; it could only be used for the current Federal fiscal year. In other situations, we continue to believe that it is appropriate to make prospective corrections to the wage index in those circumstances where a hospital could not have known about or did not have the opportunity to correct the fiscal intermediary's or CMS's error before the beginning of the fiscal year (that is, by the June deadline).

We are making this change to Sec. 412.64(k)(2) effective on October 1, 2005, that is, beginning with the FY 2006 wage index. We note that, as with prospective changes to the wage index, the final retroactive correction will be made irrespective of whether the change increases or decreases a hospital's payment rate. In addition, we note that the policy of retroactive adjustment will still apply in those instances where a judicial decision reverses a CMS denial of a hospital's wage data revision request.

In addition, in the FY 2006 IPPS proposed rule, we proposed to correct the FY 2005 wage index retroactively (that is, from October 1, 2004) on a one-time only basis for a limited circumstance using the authority provided under section 903(a)(1) of Pub. L. 108-173. This provision authorizes the Secretary to make retroactive changes to items and services if failure to apply such changes would be contrary to the public interest. However, as indicated, our current regulations at Sec. 412.64(k)(1) allow only for a prospective correction to the hospitals' area wage index values. We proposed to correct the FY 2005 wage index retroactively in the limited circumstance where a hospital meets all of the following criteria: (1) The fiscal intermediary or CMS made an error in tabulating a hospital's FY 2005 wage index data; (2) the hospital informed the fiscal intermediary or CMS, or both, about the error, following the established schedule and process for requesting corrections to its FY 2005 wage index data; and (3) CMS agreed before October 1 that the fiscal intermediary or CMS made an error in tabulating the hospital's wage data and the wage index should be corrected by the beginning of the Federal fiscal year (that is, by October 1, 2004), but CMS was unable to publish the correction by the beginning of the fiscal year.

On December 30, 2004, we published in the Federal Register a correction notice to the FY 2005 IPPS final rule that included the corrected wage data for four hospitals that meet all of the three above stated criteria (69 FR 78526). These corrections were effective January 1, 2005. As noted, our current regulations allow only for a prospective correction to the hospitals' area wage index values. However, we believe that, in the limited circumstance mentioned above, a retroactive correction to the FY 2005 wage index is appropriate and meets the condition of section 903(a)(1) of Pub. L. 108-173 that ``failure to apply the change retroactively would be contrary to the public interest.''

Comment: Several commenters supported CMS' proposal to correct the FY 2005 wage index retroactive to October 1, 2004, using the authority provided under section 903(a)(1) of Pub. L. 108-173 on a one-time only basis for the limited circumstance where a hospital meets the first two criteria specified in the proposal. However, the commenters requested that CMS amend the proposed policy to delete the third criterion that CMS must have agreed before October 1 that the fiscal intermediary or CMS made an error in tabulating the hospital's wage data. The commenters were concerned that if CMS could not notify hospitals before October 1 that the wage data would be corrected, the hospital would not be eligible for the retroactive correction to the FY 2005 wage index.

Response: We believe it is important to retain the requirement that CMS must have notified the hospital before October 1 that an error was made in calculating the wage index for an area for the correction to be made retroactively to October 1. The October 1 date is relevant because it is the first day of the new fiscal year. Once the fiscal year begins, we believe it is important to only make changes to the wage index prospectively, as has been CMS' longstanding policy as stated in the FY 1984 IPPS final rule (49 FR 258, January 3, 1984), unless it is clear that CMS determined that either it or the fiscal intermediary made an error prior to the beginning of the fiscal year and intended to pay hospitals using a different wage index. With respect to the specific requirements for making FY 2005 wage index corrections retroactive to October 1, 2004, we will accept letters, e-mails, and other written evidence from hospitals demonstrating that, prior to October 1, 2004, CMS agreed that an error was made to the wage index and intended to pay the hospital at the corrected wage index effective October 1, 2004.

Comment: Two commenters urged CMS to retroactively apply the policy that we are finalizing in this final rule to extend hold harmless protections to urban hospitals that are redesignated as rural under section 401 to the FY 2005 IPPS wage index.

Response: Retroactive wage corrections are intended to correct errors made in a previous year. In this case, we made a change to the regulations prospectively. Because the regulation change is unrelated to errors that were not corrected, we do not believe a retroactive wage index correction is warranted.

Comment: One commenter, a group of hospitals within a single CBSA, believed that the proposed retroactive wage index corrections should be expanded to include geographic classification errors. The commenter indicated that CMS made an error in tabulating the FY 2005 wage index data for the CBSA when it incorrectly categorized one provider as belonging to another CBSA. The commenter added that the geographic classification error had the effect of lowering the wage index of the CBSA and inflating the wage index for the other CBSA. The commenter indicated that CMS was given notice of the error prior to October 1, 2004, but the correction was changed prospectively effective January 1, 2005, rather than retrospectively.

Response: We agree that both geographic classification and reclassification technical errors should be corrected retroactive to the beginning of the fiscal year and that the special rule for FY 2005 should apply if the circumstances are the same as those that we are applying to the wage index. This would apply in cases where the wage index of an area has been miscalculated because of the improper assignment of a particular hospital to a labor market area.

Beginning with FY 2006, a hospital could receive a retroactive adjustment to its wage index for a geographic classification or reclassification error if the circumstances included in Sec. 412.64(k)(2) exist. Generally stated, the following circumstances must be present.

For classification/reclassification errors made during the proposed rule:

(1) CMS made a technical error in assigning the hospital to a geographic labor market area. (The error made must be truly technical in nature and could not include any disputes about policy or cases where a hospital disagrees with

[[Page 47387]]

the MGCRB or CMS' reclassification decisions.)

(2) The hospital notifies CMS of the technical error using the formal comment process and during the comment period on the proposed rule. (This period is different from the period for requesting wage index corrections, as wage index data are posted on the CMS Web site and must follow a certain schedule set by CMS--for example, for FY 2006, tabulation errors were required to have been identified by June 10, 2005.)

(3) The error was not corrected in the final rule.

(4) The hospital again notifies CMS of the geographic assignment error, via written correspondence or e-mail following the publication of the final rule, and CMS agrees prior to October 1 that an error was made.

For classification/reclassification errors made for the first time during the final rule:

(1) CMS made a technical error in the final rule in assigning the hospital to a geographic labor market area; and

(2) The hospital notifies CMS of the error via written correspondence or e-mail, following the publication of the final rule, and CMS agrees prior to October 1 that an error was made.

In addition, we also agree that geographic classification or reclassification errors that resulted in an incorrect wage index for FY 2005 should also be corrected retroactively (that is, from October 1, 2004) on a one-time only basis for a limited circumstance using the authority provided under section 903(a)(1) of Pub. L. 108-173. This provision authorizes the Secretary to make retroactive changes to items and services if failure to apply such changes would be contrary to the public interest. Again, we believe it would not be in the public interest for us to pay hospitals using an incorrect wage index when the geographic classification/reclassification error was brought to our attention and we agreed prior to the beginning of FY 2005 that the error should be corrected. For FY 2005, we will make corrections to the wage index for geographic classification errors retroactive to October 1, 2004 in the following circumstances:

For classification/reclassification errors made during the FY 2005 IPPS proposed rule:

(1) CMS made a technical error in the tables of the FY 2005 proposed rule (69 FR 28752, May 18, 2004) in assigning a hospital to a geographic labor market area;

(2) The hospital notified CMS of the error, via written correspondence or e-mail during the comment period on the proposed rule and using the procedures for submitting formal comments;

(3) The error was not corrected in the tables accompanying the FY 2005 final rule (69 FR 49690); and

(4) The hospital notified CMS of the error via written correspondence or e-mail following the publication of the final rule, CMS agreed prior to October 1, 2004, that an error was made, CMS agreed that the error should be corrected by the beginning of the Federal fiscal year (that is, by October 1, 2004), but CMS was unable to publish the correction by the beginning of such fiscal year.

For geographic assignment errors made for the first time during the FY 2005 final rule:

(1) CMS made a technical error in the tables of the FY 2005 final rule (69 FR 49690) in assigning a hospital to a geographic labor market area; and

(2) The hospital notified CMS of the error via written correspondence or e-mail following the publication of the final rule, CMS agreed prior to October 1, 2004, that an error was made, CMS agreed that the error should be corrected by the beginning of the Federal fiscal year (that is, by October 1, 2004), but CMS was unable to publish the correction by the beginning of such fiscal year.

IV. Rebasing and Revision of the Hospital Market Baskets

A. Background

Effective for cost reporting periods beginning on or after July 1, 1979, we developed and adopted a hospital input price index (that is, the hospital market basket for operating costs). Although ``market basket'' technically describes the mix of goods and services used to produce hospital care, this term is also commonly used to denote the input price index (that is, cost category weights and price proxies combined) derived from that market basket. Accordingly, the term ``market basket'' as used in this document refers to the hospital input price index.

The terms ``rebasing'' and ``revising,'' while often used interchangeably, actually denote different activities. ``Rebasing'' means moving the base year for the structure of costs of an input price index (for example, in this final rule, we are shifting the base year cost structure for the IPPS hospital index from FY 1997 to FY 2002). ``Revising'' means changing data sources, or price proxies, used in the input price index.

The percentage change in the market basket reflects the average change in the price of goods and services hospitals purchase in order to furnish inpatient care. We first used the market basket to adjust hospital cost limits by an amount that reflected the average increase in the prices of the goods and services used to provide hospital inpatient care. This approach linked the increase in the cost limits to the efficient utilization of resources.

Since the inception of the IPPS, the projected change in the hospital market basket has been the integral component of the update factor by which the prospective payment rates are updated every year. An explanation of the hospital market basket used to develop the prospective payment rates was published in the Federal Register on September 1, 1983 (48 FR 39764). We also refer the reader to the August 1, 2002 Federal Register (67 FR 50032) in which we discussed the previous rebasing of the hospital input price index.

The hospital market basket is a fixed weight, Laspeyres-type price index that is constructed in three steps. First, a base period is selected (in this final rule, FY 2002) and total base period expenditures are estimated for a set of mutually exclusive and exhaustive spending categories based upon type of expenditure. Then the proportion of total operating costs that each category represents is determined. These proportions are called cost or expenditure weights. Second, each expenditure category is matched to an appropriate price or wage variable, referred to as a price proxy. In nearly every instance, these price proxies are price levels derived from publicly available statistical series that are published on a consistent schedule, preferably at least on a quarterly basis.

Finally, the expenditure weight for each cost category is multiplied by the level of its respective price proxy. The sum of these products (that is, the expenditure weights multiplied by their price levels) for all cost categories yields the composite index level of the market basket in a given period. Repeating this step for other periods produces a series of market basket levels over time. Dividing an index level for a given period by an index level for an earlier period produces a rate of growth in the input price index over that time period.

The market basket is described as a fixed-weight index because it describes the change in price over time of the same mix of goods and services purchased to provide hospital services in a base period. The effects on total expenditures resulting from changes in the quantity or mix of goods and services (intensity) purchased subsequent to the base period are not

[[Page 47388]]

measured. For example, shifting a traditionally inpatient type of care to an outpatient setting might affect the volume of inpatient goods and services purchased by the hospital, but would not be factored into the price change measured by a fixed weight hospital market basket. In this manner, the market basket measures only the pure price change. Only when the index is rebased using a more recent base period would the quantity and intensity effects be captured in the cost weights. Therefore, we rebase the market basket periodically so the cost weights reflect changes in the mix of goods and services that hospitals purchase (hospital inputs) to furnish inpatient care between base periods. We last rebased the hospital market basket cost weights effective for FY 2003 (67 FR 50032, August 1, 2002), with FY 1997 data used as the base period for the construction of the market basket cost weights.

B. Rebasing and Revising the Hospital Market Basket

1. Development of Cost Categories and Weights a. Medicare Cost Reports

The major source of expenditure data for developing the rebased and revised hospital market basket cost weights is the FY 2002 Medicare cost reports. These cost reports are from IPPS hospitals only. They do not reflect data from hospitals excluded from the IPPS or CAHs. The IPPS cost reports yield seven major expenditure or cost categories: wages and salaries, employee benefits, contract labor, pharmaceuticals, professional liability insurance (malpractice), blood and blood products, and a residual ``all other.''

Chart 1.--Major Cost Categories Found in Medicare Cost Reports

FY 1997-based FY 2002-based Major cost categories

market basket market basket

Wages and salaries......................

48.965

45.590 Employee benefits.......................

10.597

11.189 Contract labor..........................

2.094

3.214 Professional Liability Insurance

0.840

1.589 (Malpractice).......................... Pharmaceuticals.........................

5.416

5.855 Blood and blood products................

0.875

1.082 All other...............................

31.213

31.481

b. Other Data Sources

In addition to the Medicare cost reports, other sources of data used in developing the market basket weights are the Benchmark Input- Output Tables (I-Os) created by the Bureau of Economic Analysis, U.S. Department of Commerce, and the Business Expenses Survey developed by the Bureau of the Census, U.S. Department of Commerce, from its Economic Census.

New data for these sources are scheduled for publication every 5 years, but may take up to 7 years after the reference year. Only an Annual I-O is produced each year, but the Annual I-O contains less industry detail than does the Benchmark I-O. When we rebased the market basket using FY 1997 data in the FY 2003 IPPS final rule, the 1997 Benchmark I-O was not yet available. Therefore, we did not incorporate data from that source into the FY 1997-based market basket (67 FR 50033). However, we did use a secondary source, the 1997 Annual Input- Output tables. The third source of data, the 1997 Business Expenditure Survey (now known as the Business Expenses Survey) was used to develop weights for the utilities and telephone services categories.

The 1997 Benchmark I-O data are a much more comprehensive and complete set of data than the 1997 Annual I-O estimates. The 1997 Annual I-O is an update of the 1992 I-O tables, while the 1997 Benchmark I-O is an entirely new set of numbers derived from the 1997 Economic Census. The 2002 Benchmark Input-Output tables are not yet available. Therefore, as we proposed in the FY 2006 IPPS proposed rule, we use the 1997 Benchmark I-O data in the FY 2002-based market basket, to be effective for FY 2006. Instead of using the less detailed, less accurate Annual I-O data, we aged the 1997 Benchmark I-O data forward to FY 2002. The methodology we used to age the data involves applying the annual price changes from the price proxies to the appropriate cost categories. We repeat this practice for each year.

The ``all other'' cost category is further divided into other hospital expenditure category shares using the 1997 Benchmark Input- Output tables. Therefore, the ``all other'' cost category expenditure shares are proportional to their relationship to ``all other'' totals in the I-O tables. For instance, if the cost for telephone services were to represent 10 percent of the sum of the ``all other'' I-O (see below) hospital expenditures, then telephone services would represent 10 percent of the market basket's ``all other'' cost category. 2. PPS--Selection of Price Proxies

After computing the FY 2002 cost weights for the rebased hospital market basket, it was necessary to select appropriate wage and price proxies to reflect the rate-of-price change for each expenditure category. With the exception of the Professional Liability proxy, all the indicators are based on Bureau of Labor Statistics (BLS) data and are grouped into one of the following BLS categories:

Producer Price Indexes--Producer Price Indexes (PPIs) measure price changes for goods sold in other than retail markets. PPIs are preferable price proxies for goods that hospitals purchase as inputs in producing their outputs because the PPIs would better reflect the prices faced by hospitals. For example, we use a special PPI for prescription drugs, rather than the Consumer Price Index (CPI) for prescription drugs because hospitals generally purchase drugs directly from the wholesaler. The PPIs that we use measure price change at the final stage of production.

Consumer Price Indexes--Consumer Price Indexes (CPIs) measure change in the prices of final goods and services bought by the typical consumer. Because they may not represent the price faced by a producer, we used CPIs only if an appropriate PPI was not available, or if the expenditures were more similar to those of retail consumers in general rather than purchases at the wholesale level. For example, the CPI for food purchased

[[Page 47389]]

away from home is used as a proxy for contracted food services.

Employment Cost Indexes--Employment Cost Indexes (ECIs) measure the rate of change in employee wage rates and employer costs for employee benefits per hour worked. These indexes are fixed-weight indexes and strictly measure the change in wage rates and employee benefits per hour. Appropriately, they are not affected by shifts in employment mix.

We evaluated the price proxies using the criteria of reliability, timeliness, availability, and relevance. Reliability indicates that the index is based on valid statistical methods and has low sampling variability. Timeliness implies that the proxy is published regularly, at least once a quarter. Availability means that the proxy is publicly available. Finally, relevance means that the proxy is applicable and representative of the cost category weight to which it is applied. The CPIs, PPIs, and ECIs selected meet these criteria.

Chart 2 sets forth the complete market basket including cost categories, weights, and price proxies. For comparison purposes, the corresponding FY 1997-based market basket is listed as well. A summary outlining the choice of the various proxies follows the chart.

Chart 2.--FY 2002-Based PPS Hospital Market Basket Cost Categories, Weights, and Proxies With FY 1997-Based Market Basket Used for Comparison

Rebased FY FY 1997-Based 2002-based Expense categories

hospital

hospital Rebased FY 2002-based hospital market basket market basket market basket price proxies weights

weights

1. Compensation...............................

61.656

59.993 ................................

A. Wages and Salaries*....................

50.686

48.171 ECI--Wages and Salaries, Civilian Hospital Workers.

B. Employee Benefits*.....................

10.970

11.822 ECI--Benefits, Civilian Hospital Workers. 2. Professional Fees*.........................

5.401

5.510 ECI--Compensation for Professional, Specialty & Technical Workers. 3. Utilities..................................

1.353

1.251 ................................

A. Fuel, Oil, and Gasoline................

0.284

0.206 PPI Refined Petroleum Products.

B. Electricity............................

0.833

0.669 PPI Commercial Electric Power.

C. Water and Sewerage.....................

0.236

0.376 CPI-U Water & Sewerage Maintenance. 4. Professional Liability Insurance...........

0.840

1.589 CMS Professional Liability Insurance Premium Index. 5. All Other..................................

30.749

31.657 ................................

A. All Other Products.....................

19.537

20.336 ................................ (1) Pharmaceuticals...................

5.416

5.855 PPI Prescription Drugs. (2) Direct Purchase Food..............

1.370

1.664 PPI Processed Foods & Feeds. (3) Contract Service Food.............

1.274

1.180 CPI-U Food Away From Home. (4) Chemicals.........................

2.604

2.096 PPI Industrial Chemicals. (5) Blood and Blood Products**........

0.875 .............. ................................ (6) Medical Instruments...............

2.192

1.932 PPI Medical Instruments & Equipment. (7) Photographic Supplies.............

0.204

0.183 PPI Photographic Supplies. (8) Rubber and Plastics...............

1.668

2.004 PPI Rubber & Plastic Products. (9) Paper Products....................

1.355

1.905 PPI Converted Paper & Paperboard Products. (10) Apparel..........................

0.583

0.394 PPI Apparel. (11) Machinery and Equipment..........

1.040

0.565 PPI Machinery & Equipment. (12) Miscellaneous Products**.........

0.956

2.558 PPI Finished Goods less Food and Energy.

B. All Other Services.....................

11.212

11.321 ................................ (1) Telephone Services................

0.398

0.458 CPI-U Telephone Services. (2) Postage...........................

0.857

1.300 CPI-U Postage. (3) All Other: Labor Intensive*.......

5.438

4.228 ECI--Compensation for Private Service Occupations. (4) All Other: Non-Labor Intensive....

4.519

5.335 CPI-U All Items.

Total.....................................

100.000

100.000 ................................

* Labor-Related. ** Blood and blood products, previously a separate cost category, is now contained within Miscellaneous Products in the FY 2002-based market basket. See discussion in section IV.B.2.r., miscellaneous products, as well as comment and response on blood and blood products that follow this section.

a. Wages and Salaries

For measuring the price growth of wages in the FY 2002-based market basket, as we proposed, we used the ECI for wages and salaries for civilian hospital workers as the proxy for wages in the hospital market basket. This same proxy was used for the FY 1997-based market basket. b. Employee Benefits

The FY 2002-based hospital market basket uses the ECI for employee benefits for civilian hospital workers. This is the same proxy that was used in the FY 1997-based market basket. c. Nonmedical Professional Fees

The ECI for compensation for professional and technical workers in private industry is applied to this category because it includes occupations such as management and consulting, legal, accounting and engineering services. The same proxy was used in the FY 1997-based market basket. d. Fuel, Oil, and Gasoline

The percentage change in the price of gas fuels as measured by the PPI (Commodity Code 0552) is applied to this component. The same proxy was used in the FY 1997-based market basket. e. Electricity

The percentage change in the price of commercial electric power as measured by the PPI (Commodity Code 0542) is applied to this component. The same

[[Page 47390]]

proxy was used in the FY 1997-based market basket. f. Water and Sewerage

The percentage change in the price of water and sewerage maintenance as measured by the CPI for all urban consumers (CPI Code CUUR0000SEHG01) is applied to this component. The same proxy was used in the FY 1997-based market basket. g. Professional Liability Insurance

The FY 2002-based index uses the percentage change in the hospital professional liability insurance (PLI) premiums as estimated by the CMS Hospital Professional Liability Index, which we use as a proxy in the Medicare Economic Index (68 FR 63244), for the proxy of this category. Similar to the Physicians Professional Liability Index, we attempt to collect commercial insurance premiums for a fixed level of coverage, holding nonprice factors constant (such as a change in the level of coverage). In the FY 1997-based market basket, the same price proxy was used.

We continue to research options for improving our proxy for professional liability insurance. This research includes exploring various options for expanding our current survey, including the identification of another entity that would be willing to work with us to collect more complete and comprehensive data. We are also exploring other options such as third party or industry data that might assist us in creating a more precise measure of PLI premiums. We have not yet identified a preferred option. Therefore, we did not make any changes to the proxy in this rule. h. Pharmaceuticals

The percentage change in the price of prescription drugs as measured by the PPI (PPI Code PPI32541DRX) is used as a proxy for this category. This is a special index produced by BLS and is the same proxy used in the FY 1997-based market basket. i. Food: Direct Purchases

The percentage change in the price of processed foods and feeds as measured by the PPI (Commodity Code 02) is applied to this component. The same proxy was used in the FY 1997-based market basket. j. Food: Contract Services

The percentage change in the price of food purchased away from home as measured by the CPI for all urban consumers (CPI Code CUUR0000SEFV) is applied to this component. The same proxy was used in the FY 1997-based market basket. k. Chemicals

The percentage change in the price of industrial chemical products as measured by the PPI (Commodity Code 061) is applied to this component. While the chemicals hospitals purchase include industrial as well as other types of chemicals, the industrial chemicals component constitutes the largest proportion by far. Thus, we believe that Commodity Code 061 is the appropriate proxy. The same proxy was used in the FY 1997-based market basket. l. Medical Instruments

The percentage change in the price of medical and surgical instruments as measured by the PPI (Commodity Code 1562) is applied to this component. The same proxy was used in the FY 1997-based market basket. m. Photographic Supplies

The percentage change in the price of photographic supplies as measured by the PPI (Commodity Code 1542) is applied to this component. The same proxy was used in the FY 1997-based market basket. n. Rubber and Plastics

The percentage change in the price of rubber and plastic products as measured by the PPI (Commodity Code 07) is applied to this component. The same proxy was used in the FY 1997-based market basket. o. Paper Products

The percentage change in the price of converted paper and paperboard products as measured by the PPI (Commodity Code 0915) is used. The same proxy was used in the FY 1997-based market basket. p. Apparel

The percentage change in the price of apparel as measured by the PPI (Commodity Code 381) is applied to this component. The same proxy was used in the FY 1997-based market basket. q. Machinery and Equipment

The percentage change in the price of machinery and equipment as measured by the PPI (Commodity Code 11) is applied to this component. The same proxy was used in the FY 1997-based market basket. r. Miscellaneous Products

The percentage change in the price of all finished goods less food and energy as measured by the PPI (Commodity Code SOP3500) is applied to this component. Using this index removes the double-counting of food and energy prices, which are already captured elsewhere in the market basket. The same proxy was used in the FY 1997-based market basket. The weight for this cost category is higher than in the FY 1997-based market basket because the weight for blood and blood products (1.082) is added to it. In the FY 1997-based market basket, we included a separate cost category for blood and blood products, using the BLS PPI (Commodity Code 063711) for blood and derivatives as a price proxy. A review of recent trends in the PPI for blood and derivatives suggests that its movements may not be consistent with the trends in blood costs faced by hospitals. While this proxy did not match exactly with the product hospitals are buying, its trend over time appears to be reflective of the historical price changes of blood purchased by hospitals. However, an apparent divergence over recent periods led us to reevaluate whether the PPI for blood and derivatives was an appropriate measure of the changing price of blood. We ran test market baskets classifying blood in three separate cost categories: blood and blood products, contained within chemicals as was done for the FY 1992-based market basket, and within miscellaneous products. These categories use as proxies the following PPIs: the PPI for blood and blood derivatives, the PPI for chemicals, and the PPI for finished goods less food and energy, respectively. Of these three market baskets, the market basket with blood in miscellaneous products and its associated proxy, the PPI for finished goods less food and energy, moved very similar to the market basket with blood as a separate category. In addition, the impact on the overall market basket by using different proxies for blood was negligible, mostly due to the relatively small weight for blood in the market basket. Therefore, we chose the PPI for finished goods less food and energy for the blood proxy because we believe it will best be able to proxy price changes (not quantities or required tests) associated with blood purchased by hospitals. We will continue to evaluate this proxy for its appropriateness and will explore the development of alternative price indexes to proxy the price changes associated with this cost.

Comment: Some commenters questioned the CMS proposal to remove blood and blood products as a separate cost category and add its weight to the miscellaneous products cost category of

[[Page 47391]]

the hospital market basket. A few commenters supported this move only as a temporary measure until a more appropriate blood and blood products PPI can be developed by the BLS.

Response: We studied different cost categories that might be used until we have had the opportunity to evaluate whether the BLS' PPI for Blood and Organ Banks (NAICS 621991), which is still in development, may be an appropriate price proxy that could be proposed for blood and blood products. The alternative cost categories we considered were Blood and Blood Products, Chemicals, and Miscellaneous Products. We considered placing blood and blood products in the ``other products'' subcategory because blood is a product purchased by hospitals. From 2001 to 2003 the percent changes in the price proxies for these respective cost categories were:

Chart 3.--Annual Growth Rates for Three Possible Price Proxies

Cost category

Proxy

2001-2002 2002-2003

Chemicals..................................... Industrial Chemicals............

-0.9

11.3 Blood......................................... Blood and Derivatives...........

-7.2

-11.4 Miscellaneous Products........................ Finished Goods less Food and

0.1

0.2 Energy.

In discussions with the blood banking industry we were presented data that the cost of blood had been increasing over the 2001-2003 period. In addition, an analysis of Medicare Cost Report data indicated that the cost weight for blood was increasing had increased from 1.023 in 2001 to 1.082 in 2002. Neither of these data sources supported the trends in the PPI for blood and derivatives over this period. In addition, we had previously determined that the PPI for Industrial Chemicals was not an appropriate price proxies for the change in blood prices (67 FR 50035). We believed the PPI for finished goods less food and energy was an appropriate proxy because it has a more stable measure than the others considered, and had not exhibited negative price movements in recent periods and currently serves as a proxy for all product costs that are small or without a specific price proxy.

We ran test market baskets using the most recent forecast (2005q2, with history through 2005q1). The three market baskets were identical, except that the blood weight was in its own cost category, in chemicals, or in miscellaneous products, respectively. As shown in Chart 4, the annual increases in the market baskets were similar, regardless of which cost category contained the market basket weight for blood and blood products. Therefore, even if blood and blood products were its own cost category, it would have little effect on the market basket update factor.

Chart 4.--Market Basket Increase With Blood and Blood Products located in:

Blood and

Miscellaneous blood products Chemicals products

2000............................................................

3.2

3.3

3.2 2001............................................................

4.2

4.2

4.1 2002............................................................

3.7

3.6

3.7 2003............................................................

3.9

4.2

4.0 2004............................................................

3.9

4.0

3.9 Average: 2000-2004..............................................

3.8

4.0

3.8

We are adopting the PPI for finished goods less food and energy as the price proxy for blood and blood products because our analysis shows that this price proxy most accurately reflects changes in costs of blood products. We note that the BLS is developing a Producer Price Index for Blood and Organ Banks. We look forward to evaluating this index when it is ready for use. s. Telephone

The percentage change in the price of telephone services as measured by the CPI for all urban consumers (CPI Code CUUR0000SEED) is applied to this component. The same proxy was used in the FY 1997-based market basket. t. Postage

The percentage change in the price of postage as measured by the CPI for all urban consumers (CPI Code CUUR0000SEEC01) is applied to this component. The same proxy was used in the FY 1997-based market basket. u. All Other Services: Labor Intensive

The percentage change in the ECI for compensation paid to service workers employed in private industry is applied to this component. The same proxy was used in the FY 1997-based market basket. v. All Other Services: Nonlabor Intensive

The percentage change in the all-items component of the CPI for all urban consumers (CPI Code CUUR0000SA0) is applied to this component. The same proxy was used in the FY 1997-based market basket.

For further discussion of the rationales for choosing many of the specific price proxies, we refer the reader to the August 1, 2002 final rule (67 FR 50037).

[[Page 47392]]

Chart 5.--FY 1997-Based and FY 2002-Based Prospective Payment Hospital Operating Index Percent Change, FY 2000 Through FY 2008

Rebased FY 2002-based FY 1997-based Fiscal year (FY)

hospital market basket market basket

Historical data:

FY 2000.............................

3.2

3.3

FY 2001.............................

4.1

4.3

FY 2002.............................

3.7

3.8

FY 2003.............................

4.0

3.9

FY 2004.............................

3.9

3.9

Average FYs 2000-2004...............

3.8

3.8 Forecast:

FY 2005.............................

4.2

4.2

FY 2006.............................

3.7

3.7

FY 2007.............................

3.1

3.2

FY 2008.............................

2.9

3.0

Average FYs 2005-2008...............

3.5

3.5

TL0505.SIM

Prior to the publication of the FY 2006 IPPS proposed rule, we had been actively working with our forecasting firm, Global Insight, Inc. (GII), to improve the forecasting accuracy of the market baskets. GII is a nationally recognized economic and financial forecasting firm that contracts with CMS to forecast the components of the market baskets. Among other services GII provides to CMS, GII calculates projected inflation factors for price proxies using models that take into account national and global economic trends.

Over the last several years, dramatic fluctuations in the price of certain costs have made it difficult to forecast price proxy inflation. This uncertainty has resulted in market basket forecast error greater than 0.25 percentage points in FY 2001, FY 2003, and FY 2004. The driving force behind much of this uncertainty has been the instability of energy costs, which, in a global economy, have an indirect effect on wages and other costs as well as a direct effect on utility prices. With our input and consultation, GII recently evaluated and modified their forecasting models to help improve their accuracy. Using these improved forecasting models, GII calculated updated inflation factors for the major cost categories in Chart 6.

Chart 6.--Comparison of the 4 Quarter Moving Average Percent Changes for Several Cost Category Weights Between the FY 2006 IPPS Proposed and Final Rules

GII 2004q4 GII 2005q2 FY 2002-based forecast of FY forecast of FY Expense category

cost weights 2006 (Proposed 2006 (Final Rule)

Rule)

Total--PPS02....................................................

100.000

3.2

3.7

Compensation................................................

59.993

3.5

3.9

Utilities...................................................

1.251

0.8

3.6

Professional Fees...........................................

5.510

3.6

4.3

Professional liability insurance............................

1.589

8.4

7.8

All Other...................................................

31.657

2.4

3.0

All Other Products..........................................

20.336

2.3

3.2

All Other Services..........................................

11.321

2.4

2.6

In the FY 2006 IPPS proposed rule, we forecasted a market basket update of 3.2 percent. Based on our updated forecasting model, we are forecasting a market basket update of 3.7 percent for FY 2006.

Comment: Several commenters requested that CMS review and revise the methodology used to determine the projected FY 2006 market basket. They are concerned that the previously proposed FY 2006 update of 3.2 percent is a dramatic underestimation. They emphasized the importance of a reliable projection methodology in order to ensure equitable payments.

Response: We recognize the importance of a reliable forecasting methodology. As discussed above, we have worked with our forecasting firm, GII, to modify and improve GII's forecasting models to help improve their accuracy. The final FY 2006 update of 3.7 percent reflects these modifications.

Comment: Several commenters requested that CMS make the calculation of the projected FY 2006 available to the public.

Response: We have summarized our calculation of the market basket update in Chart 6 above. 3. Labor-Related Share

Under section 1886(d)(3)(E) of the Act, the Secretary estimates from time to time the proportion of payments that are labor-related. ``The Secretary shall adjust the proportion (as estimated by the Secretary from time to time) of hospitals' costs which are attributable to wages and wage-related costs of the DRG prospective payment rates * * *.'' We refer to the proportion of hospitals' costs that are attributable to wages and

[[Page 47393]]

wage-related costs as the ``labor-related share.''

The labor-related share is used to determine the proportion of the national PPS base payment rate to which the area wage index is applied. As we proposed in the FY 2006 IPPS proposed rule, we are continuing to use our current methodology of defining the labor-related share as the national average proportion of operating costs that are attributable to wages and salaries, fringe benefits, professional fees, contract labor, and labor intensive services. Therefore, we calculate the labor-related share by adding the relative weights for these operating cost categories. We continue to believe, as we have stated in the past, that these operating cost categories likely are related to, are influenced by, or vary with the local markets. Our definition of the labor-related share therefore continues to be consistent with section 1886(d)(3) of the Act. As we proposed, we are removing postage costs from the FY 2002-based labor-related share.

Using the cost category weights that we determined in section IV.B. of this preamble, we calculated a labor-related share of 69.731 percent, using the FY 2002-based PPS market basket. Accordingly, in this final rule, we are implementing a labor-related share of 69.7 percent for discharges occurring on or after October 1, 2005. We note that section 403 of Pub. L. 108-173 amended sections 1886(d)(3)(E) and 1886(d)(9)(C)(iv) of the Act to provide that the Secretary must employ 62 percent as the labor-related share unless this employment ``would result in lower payments than would otherwise be made.''

Comment: One commenter suggested that we decrease the labor-related share from 62 percent to 50 percent for those hospitals with wage indices under 1.0.

Response: As stated above, the 62 percent labor-related share provision was established by section 403 of Pub. L. 108-173. This provision was mandated by Congress and, therefore, CMS has no authority to modify it.

As we proposed, we also are updating the labor-related share for Puerto Rico. Consistent with our methodology for determining the national labor-related share, we add the Puerto Rico-specific relative weights for wages and salaries, fringe benefits, and contract labor. Because there are no Puerto Rico-specific relative weights for professional fees and labor intensive services, we use the national weights. In the proposed rule, we observed that, rather than using a Puerto Rico-specific labor-related share, another option would be to apply the national labor-related share to the Puerto Rico-specific rate. In the proposed rule, we also noted that we were still reviewing our data and had not yet calculated the updated Puerto Rico-specific labor-related share percentage. Therefore, in the proposed rule, the labor-related and nonlabor-related portions of the Puerto Rico-specific standardized amount listed in Table 1C of the Addendum to the proposed rule reflected the current FY 2005 labor-related share for Puerto Rico of 71.3 percent. We solicited comments on our proposal to update the labor-related share for Puerto Rico.

After publication of the proposed rule, we calculated an updated labor-related share of 58.7 percent for Puerto Rico and posted it on the CMS Web site at http://www.cms.hhs.gov/providers/hipps. We did not

receive any public comments on the proposed updated labor-share for Puerto Rico. Accordingly, we are adopting an updated Puerto Rico labor- related share of 58.7 percent, which is reflected in the Table 1C of the Addendum of this final rule.

Unlike the 1997 Annual I-O which was based on Standard Industrial Codes (SIC), the 1997 Benchmark I-O is categorized using the North American Industrial Classification System (NAICS). This change required us to classify all cost categories under NAICS, including a reevaluation of labor-related costs on the NAICS definitions. Chart 7 compares the FY 1992-based labor-related share, the current measure, with the FY 2002-based labor-related share. When we rebased the market basket to reflect FY 1997 data, we did not change the labor-related share (67 FR 50041). Therefore, the FY 1992-based labor-related share is the current measure.

Chart 7.--Labor-Related Share: FY 1992-Based and FY 2002-Based

FY 1992-based FY 2002-based Cost category

weight

weight

Difference

Wages and salaries..............................................

50.244

48.171

-2.073 Fringe benefits.................................................

11.146

11.822

0.676 Nonmedical professional fees....................................

2.127

5.510

3.383 Postal services*................................................

0.272 ..............

-0.272 Other labor-intensive services**................................

7.277

4.228

-3.049

Total labor-related.........................................

71.066

69.731

-1.335

Total nonlabor-related..................................

28.934

30.269

1.335

*No longer considered to be labor-related. **Other labor-intensive services includes landscaping services, services to buildings, detective and protective services, repair services, laundry services, advertising, auto parking and repairs, physical fitness facilities, and other government enterprises.

Although we are continuing to calculate the labor-related share by adding the relative weights of the labor-related operating cost categories, we continue to evaluate alternative methodologies. In the May 9, 2002 Federal Register (67 FR 31447), we discussed our research on the methodology for the labor-related share. This research involved analyzing the compensation share (the sum of wages and salaries and benefits) separately for urban and rural hospitals, using regression analysis to determine the proportion of costs influenced by the area wage index, and exploring alternative methodologies to determine whether all or only a portion of professional fees and nonlabor intensive services should be considered labor-related.

Our original analysis, which appeared in the May 9, 2002 Federal Register (67 FR 31447) and which focused mainly on edited FY 1997 hospital data, found that the compensation share of costs for hospitals in rural areas was higher on average than the compensation share for hospitals in urban areas. We also researched whether only a proportion of the costs in professional fees and labor-intensive services should be considered labor-related, not the entire cost categories. However, there was not sufficient information available to make this determination.

[[Page 47394]]

Our finding that the average compensation share of costs for rural hospitals was higher than the average compensation for urban hospitals was validated consistently through our regression analysis. Regression analysis is a statistical technique that determines the relationship between a dependent variable and one or more independent variables. We tried several regression specifications in an effort to determine the proportion of costs that are influenced by the area wage index. Furthermore, MedPAC raised the possibility that regression may be an alternative to the current market basket methodology. In our initial regression specification (in log form), Medicare operating cost per Medicare discharge was the dependent variable and the independent variables were the area wage index, the case-mix index, the ratio of residents per bed (as proxy for IME status), and a dummy variable that equaled one if the hospital was located in a metropolitan area with a population of 1 million or more. (A dummy variable represents the presence or absence of a particular characteristic.) This regression produced a coefficient for all hospitals for the area wage index of 0.638 (which is equivalent to the labor share and can be interpreted as an elasticity because of the log specification) with an adjusted R- squared of 64.3. (Adjusted R-squared is a measure of how well the regression model fits the data.) While, on the surface, this appeared to be a reasonable result, this same specification for urban hospitals had a coefficient of 0.532 (adjusted R-squared = 53.2) and a coefficient of 0.709 (adjusted R-squared = 36.4) for rural hospitals. This highlighted some apparent problems with the specification because the overall regression results appeared to be masking underlying problems. It did not seem reasonable that urban hospitals would have a labor share below their actual compensation share or that the discrepancy between urban and rural hospitals would be this large. When we standardized the Medicare operating cost per Medicare discharge for case-mix, the fit, as measured by adjusted R-squared, fell dramatically and the urban/rural discrepancy became even larger.

Based on this initial result, we tried two modifications to the FY 1997 regressions to correct for the underlying problems. First, we edited the data differently to determine whether a few reports were causing the inconsistent results. We found when we tightened the edits, the wage index coefficient was lower and the fit was worse. When we loosened the edits, we found higher wage index coefficients and still a worse fit. Second, we added additional variables to the regression equation to attempt to explain some of the variation that was not being captured. We found the best fit occurred when the following variables were added: the occupancy rate, the number of hospital beds, a dummy variable that equals one if the hospital is privately owned and zero otherwise, a dummy variable that equals one if the hospital is government-controlled and zero otherwise, the Medicare length of stay, the number of FTEs per bed, and the age of fixed assets. The result of this specification was a wage index coefficient of 0.620 (adjusted R- squared = 68.7), with the regression on rural hospitals having a coefficient of 0.772 (adjusted R-squared = 45.0) and the regression on urban hospitals having a coefficient of 0.474 (adjusted R-squared = 60.9). Neither of these alternatives seemed to help the underlying difficulties with the regression analysis.

Subsequent to the work described above, we have undertaken the research necessary to reevaluate the current assumptions used in determining the labor-related share. We ran regressions applying the previous specifications to more recent data (FY 2001 and FY 2002), and, as described below, we ran regressions using alternative specifications. In the FY 2006 IPPS proposed rule, we solicited comments on this research and any information that is available to help determine the most appropriate measure.

The first step in our regression analysis to determine the proportion of hospitals' costs that varied with labor-related costs was to edit the data, which had significant outliers in some of the variables we used in the regressions. We originally began with an edit that excluded the top and bottom 5 percent of reports based on average Medicare cost per discharge and number of discharges. We also used edits to exclude reports that did not meet basic criteria for use, such as having costs greater than zero for total, operating, and capital for the overall facility and just the Medicare proportion. We also used an edit that required that the hospital occupancy rate, length of stay, number of beds, FTEs, and overall and Medicare discharges be greater than zero. Finally, we excluded reports with occupancy rates greater than one.

Our regression specification (in log form) was Medicare operating cost per Medicare discharge as the dependent variable (the same dependent variable we used in the regression analysis described in the May 9, 2002 Federal Register) with the independent variables being the compensation per FTE, the ratio of interns and residents per bed (as proxy for IME status), the occupancy rate, the number of hospital beds, a dummy variable that equals one if the hospital is privately owned and is zero otherwise, a dummy variable that equals one if the hospital is government-controlled and is zero otherwise, the Medicare length of stay, the number of FTEs per bed, the age of fixed assets, and a dummy variable that equals one if the hospital is located in a metropolitan area with a population of 1 million or more. This is a similar model to the one described in the May 9, 2002 Federal Register (67 FR 31447) as having the best fit, with two notable exceptions. First, the area wage index is replaced by compensation per FTE, where compensation is the sum of hospital wages and salaries, contract labor costs, and benefits. The area wage index is a payment variable computed by averaging wages across all hospitals within each MSA, whereas compensation per FTE differs from one hospital to the next. Second, the case-mix index is no longer included as a regressor because it is correlated with other independent variables in the regression. In other words, the other independent variables are capturing part of the effect of the case-mix index. We made these two specification changes in an attempt to only use cost variables to explain the variation in Medicare operating costs per discharge. We believe this is appropriate in order to compare to the results we are getting from the market basket methodology, which is based solely on cost data. As we will show below, the use of payment variables on the right-hand side of the equation appears to be producing less reasonable results when cost data are used.

The revised specification for FY 2002 produced a coefficient for all hospitals for compensation per FTE of 0.673 (which is roughly equivalent to the labor share and can be interpreted as an elasticity because of the log specification) with an adjusted R-squared of 63.7. The coefficient result for FY 2001 is 64.5, with an adjusted R-squared of 65.2. (For comparison, a separate regression for FY 2002 with the log area wage index and log case-mix index included in the set of regressors displays a log area wage index coefficient of 75.6 (adjusted R-squared = 67.7).) For FY 2001, the coefficient for the log area wage index is 72.3 (adjusted R-squared = 67.9). On the surface, these seem to be reasonable results. However, a closer look reveals some problems. In

[[Page 47395]]

FY 2001, the coefficient for urban hospitals was 59.6 (adjusted R- squared = 57.3), and the coefficient for rural hospitals was 61.3 (adjusted R-squared = 50.6). On the other hand, in FY 2002, the coefficient for urban hospitals increased to 69.2 (adjusted R-squared = 55.9), and the coefficient for rural hospitals decreased to 58.2 (adjusted R-squared = 46.0). The results for FY 2001 seem reasonable, but not when compared with the results for FY 2002. Furthermore, for FY 2002 the compensation share of costs for hospitals in rural areas was higher on average than the compensation share for hospitals in urban areas. Rural areas had an average compensation share of 63.3 percent, while urban areas had a share of 60.5 percent. This compares to a share of 61.2 percent for all hospitals.

Due to these problems, we do not believe the regression analysis is producing sound enough evidence at this point for us to make the decision to change from the current method for calculating the labor- related share. We continue to analyze these data and work on alternative specifications, including working with MedPAC, who in the past have done similar analysis in their studies of payment adequacy. In the FY 2006 IPPS proposed rule, we solicited comments on this approach, given the difficulties we have encountered.

We also continue to look into ways to refine our market basket approach to more accurately account for the proportion of costs influenced by the local labor market. Specifically, we are looking at the professional fees and labor-intensive cost categories to determine if only a proportion of the costs in these categories should be considered labor-related, not the entire cost category. Professional fees include management and consulting fees, legal services, accounting services, and engineering services. Labor-intensive services are mostly building services, but also include other maintenance and repair services.

We conducted preliminary research into whether the various types of professional fees are more or less likely to be purchased locally. Through contact with a handful of hospitals in only two States, we asked for the percentages of their advertising, legal, and management and consulting services that they purchased locally, regionally, or nationally. The results were quite consistent across all of the hospitals, indicating most advertising and legal services are purchased locally or regionally and nearly all management and consulting services are purchased nationally. Although the results of our research are instructive, as we have stated in the past, we believe that items should not be excluded from the labor-related share merely because they could be purchased nationally (68 FR 45467). We do plan to expand our efforts in this area to determine whether it would be appropriate in the future to modify our methodology for calculating the labor-related share. In the FY 2006 IPPS proposed rule, we solicited data or studies that would be helpful in this analysis. However, we indicated that we were unsure if we would be able to finish this analysis in time for inclusion in this FY 2006 IPPS final rule.

Comment: Several commenters objected to our proposal to change the labor-related share to 69.7 percent and requested that CMS maintain a labor-related share of 71.1 percent. The commenters provided similar reasons for rejecting this provision of the proposed rule. Generally, the commenters were concerned that the new lower labor share would negatively impact urban hospitals and several commenters stated that CMS should postpone changing the labor share until the agency has finished researching they are finished researching different labor- related share methodologies. In addition, commenters noted that the budget neutral manner in which CMS proposed to implement this labor share change would increase the standardized amount for all hospitals. They believed this is unfair as the increased amount would provide an additional benefit to rural hospitals that are already advantaged by many provisions of Pub. L. 108-173, including section 403 which sets the labor share at 62 percent for hospitals with a wage index less than or equal to 1.0.

Response: Section 404 of Pub. L. 108-173 requires the Secretary to update the weights used in the IPPS operating and capital market baskets, including the labor-related share, to reflect the most current available data. Therefore, we are directed by statute to update the labor share and cannot maintain the labor share at the outdated percentage of 71.1. Since the FY 2003 IPPS final rule was issued, CMS has continued to evaluate alternative labor-related share methodologies. Given this research, we believe our existing methodology of calculating the labor-related share is the most appropriate methodology at this time. Our alternative methodologies did not produce the sound evidence needed to justify changing our existing methodology. Specifically, our regression results were inconsistent and highlighted underlying data problems that were not evident in our market basket labor-related share methodology. We are confident that our current model is the best method presently available to appropriately capture the changing cost structures hospitals have faced over the last ten year period (1992 to 2002). Therefore, we are establishing the labor share at 69.7 percent.

In addition, we are implementing this revised and rebased labor share in a budget neutral manner, but consistent with section 1886(d)(3)(E) of the Act, we are not taking into account the additional payments that will be made as a result of hospitals with a wage index less than or equal to 1.0 being paid using a labor-related share lower than the labor-related share of hospitals with a wage index greater than 1.0. Section 1886(d)(3)(E) of the Act directs us to determine a labor related share that reflects the ``proportion * * * of hospitals'' costs which are attributable to wages and wage-related costs.'' In addition, section 1886(d)(3)(E) of the Act requires that we implement the wage index adjustment in a budget neutral manner. However, section 403 of Pub. L. 108-173, which sets the labor-related share at 62 percent for hospitals with a wage index less than or equal to 1.0, also provides that the Secretary shall calculate the budget neutrality adjustment for the wage index as if the Pub. L. 108-173 had not been enacted. Therefore, for purposes of the budget neutrality adjustment, section 403 of Pub. L. 108-173 prohibits us from taking into account the additional payments that will be made as a result of hospitals with a wage index less than or equal to 1.0 being paid using a labor-related share of 62 percent. While we recognize that this does have the effect of increasing the standardized amount applicable to all hospitals, the statute requires this implementation methodology.

As mentioned previously in the proposed rule, we proposed to continue to calculate the labor-related share by adding the relative weights of the operating cost categories that are related to, influenced by, or vary with the local labor markets. These categories include wages and salaries, fringe benefits, professional fees, contract labor and labor-intensive services. Using this methodology, we calculated a labor-related share of 69.731, which we are using for FY 2006.

Comment: One commenter requested that CMS continue to include postage in the labor-related share.

Response: We do not believe that we should continue to include postage costs in the labor-related share as postage fees are set at nationally uniform rates and are not affected by local purchasing power of hospitals. The cost of postage is primarily

[[Page 47396]]

influenced by weight of the package and the distance the package is traveling (National Zone Chart Program Technical Guide 2003-2004, http://www.ribbs.usps.gov/files/Zone_Charts/ZCTECHNICAL_GUIDE.PDF,

page 2). For example, the cost of mailing a package from Boston, MA to Baltimore, MD (approximately 450 miles) is the same price as mailing a package from Long Beach, NC to Baltimore, MD (approximately 450 miles) (http://postcalc.usps.gov/).

Comment: One commenter argued that geographical differences in costs of goods and services such as food, energy, telephone services, pharmaceuticals, and supplies are attributable to local differences in wages and hence should be included in the labor-related share.

Response: We believe that the commenter may have misunderstood a statement in the notice of proposed rulemaking. Previously, we stated that our current methodology is to define the labor-related share as the national average proportion of operating costs that are related to, influenced by, or vary with local labor markets. As we have stated in previous rules and clarified in this final rule, it is more accurate to say that we define the labor-related share as the national average proportion of operating costs that are attributable to wages and salaries, fringe benefits, professional fees, contract labor, and labor intensive services. These costs are included in the labor-related share because they are labor intensive, and therefore, are ``hospitals'' costs that are attributable to wages and wage-related costs.'' As was stated previously, we believe that, with the exclusion of postage, the costs included in the labor-related share are, in fact, related to, influenced by, or vary with local labor markets. However, hospital costs are not necessarily ``attributable to wages and wage-related costs ``merely because they may be related to, may be influenced by, or may vary with local labor markets. Therefore, it would be incorrect to say that all costs that are related to, influenced by, or vary with the local labor market must be included in the labor-related share merely because they are related to, influenced by, or vary with the local labor market.

We include only labor-intensive inputs in the labor-related share (55 FR 36046). Although the costs of goods and services such as food, pharmaceuticals, energy, telephone services, and supplies may vary by geographic area, these items are not labor-intensive inputs. Thus, we disagree with the commenter's argument that these items should be included in the labor-related share.

Comment: Several commenters suggested that we include professional liability insurance (PLI) in the labor-related share since these costs are included in the wage index. The commenters also claimed that professional liability insurance costs are wage related.

Response: The wage index includes, as a fringe benefit cost, PLI for those policies that list actual names or specific titles of covered employees (59 FR 45358). The benefit cost weight in the market basket, included in the labor-related share, is also based on the same wage index benefit data. Therefore, the labor-related share includes these PLI costs. General PLI coverage maintained by hospitals is not recognized as a wage-related cost for purposes of the wage index or labor-related share.

Although general PLI costs do vary by geographic region, they are not labor-intensive inputs. The variance in general PLI costs is primarily influenced by state legislation and risk level, not by local wage rates. In fact, areas with high wage indices may have low relative PLI costs. For example, the malpractice geographic price indices, used in the Medicare physician payment system, for San Francisco, Los Angeles, and Boston regions are below 1, while their hospital wage indices for comparable areas are much greater than 1.

Comment: Several commenters requested that CMS explain why the labor-related share is fluctuating between FYs 1992, 1997, and 2002- based market baskets. They stated these changes raise questions about the (1) veracity of the data, (2) the change in base cost data, (3) effect of proxy changes on the trending, (4) consistency of CMS' methodology, and (5) other factors. They specifically requested that CMS explain in more detail the change in the other labor-intensive services cost weight.

Response: In addition to the official market basket weights published in the Federal Register, CMS also analyzed the weights based on different trimming methodologies and on a matched sample of hospitals over time. These weights exhibited the same trends as our published weights. Specifically, the compensation cost weight, the largest component in the labor-related share, from 1997 to 2002 steadily declined in all instances.

The decline in the nonmedical professional fees from 1992 to 1997 reflects hospital purchasing patterns' and a change in the data source used to derive this weight. The FY 1992-based market basket used the American Hospital Association Survey data while the FY 1997-based market basket used the 1997 Bureau of Economic Analysis' Annual I-O Tables. As stated in the FY 2003 IPPS final rule (67 FR 50034), if CMS had used the Annual I-O Tables to calculate the FY 1992 nonmedical professional fees component, the proportion would have been similar to the FY 1997 share. The FY 2002 nonmedical professional fees cost category is based on 1997-Benchmark I-O data trended forward using the ECI for Compensation for Private Service Occupations.

The decline in the other labor intensive cost category from 1997 to 2002 is a result of hospitals purchasing patterns and substituting the 1997 Benchmark I-O data for the 1997 Annual I-O data. The 1997 Benchmark I-O data are a much more comprehensive and complete set of data than the 1997 Annual I-O estimates. The 1997 Annual I-O is an update of the 1992 I-O tables, while the 1997 Benchmark I-O is an entirely new set of numbers derived from the 1997 Economic Census. The 1997-Benchmark I-O is also based on the 1997 North American Industrial Classification System while the 1997 Annual I-O is based on the 1987 Standard Industrial Classification System.

CMS has maintained a relatively consistent methodology for calculating the hospital market basket cost weights. However, the methodology is periodically modified to include more comprehensive data sources and/or price proxies. These methodological changes, as well as their impacts, are published in the Federal Register. In most instances, the modifications have a small effect on the total market basket update.

Finally, approximately 85 percent of the labor-related shares (FY 1992, FY 1997, and FY 2002) are based on Medicare Cost Report data submitted by hospitals.

C. Separate Market Basket for Hospitals and Hospital Units Excluded From the IPPS

1. Hospitals Paid Based on Their Reasonable Costs

On August 7, 2001, we published a final rule in the Federal Register (66 FR 41316) establishing the PPS for IRFs, effective for cost reporting periods beginning on or after January 1, 2002. On August 30, 2002, we published a final rule in the Federal Register (67 FR 55954) establishing the PPS for LTCHs, effective for cost reporting periods beginning on or after October 1, 2002. On November 15, 2004, we published a final rule in the Federal Register (69 FR 66922) establishing the PPS for the IPFs,

[[Page 47397]]

effective for cost reporting periods beginning on or after January 1, 2005.

Prior to being paid under a PPS, IRFs, LTCHs, and IPFs were reimbursed solely under the reasonable cost-based system under Sec. 413.40 of the regulations, which impose rate-of-increase limits. Children's and cancer hospitals and religious nonmedical health care institutions (RNHCIs) are still reimbursed solely under the reasonable cost-based system, subject to the rate-of-increase limits. Under these limits, an annual target amount (expressed in terms of the inpatient operating cost per discharge) is set for each hospital based on the hospital's own historical cost experience trended forward by the applicable rate-of-increase percentages. To the extent an LTCH or IPF receives a blend of reasonable cost-based payment and the Federal prospective payment rate amount, the reasonable cost portion of the payment is also subject to the applicable rate-of-increase percentage. Section 1886(b)(3)(B) (ii) of the Act sets the percentage increase of the limits, which in certain years was based upon the market basket percentage increase. Beginning in FY 2003 and subsequent years, the applicable rate-of-increase is the market basket increase. The market basket currently (and historically) used is the excluded hospital operating market basket, representing the cost structure of rehabilitation, long-term care, psychiatric, children's, and cancer hospitals (FY 2003 final rule, 67 FR 50042).

In the FY 2006 IPPS proposed rule, we indicated that because IRFs, LTCHs, and some IPFs are now paid under a PPS, we were considering developing a separate market basket for these hospitals that contains both operating and capital costs. (The IPF PPS was implemented recently for cost reporting periods beginning on or after January 1, 2005; therefore, all IPFs will soon be paid under the IPF PPS.) We indicated that we would publish any proposal to use a revised separate market basket for each of these types of hospitals when we propose the next update of their respective PPS rates. Children's and cancer hospitals are two of the remaining three types of hospitals excluded from the IPPS that are still being paid based solely on their reasonable costs, subject to target amounts. (RNHCIs, the third type of IPPS-excluded entity still subject to target amounts, are reimbursed under Sec. 403.752(a) of the regulations.) Because there are a small number of children's and cancer hospitals and RNHCIs, which receive in total less than 1 percent of all Medicare payments to hospitals and because these hospitals provide limited Medicare cost report data, in the FY 2006 IPPS proposed rule, we did not propose to create a separate market basket specifically for these hospitals. Under the broad authority in sections 1886(b)(3)(A) and (B), 1886(b)(3)(E), and 1871 of the Act, we proposed to use the FY 2002 IPPS operating market basket percentage increase to update the target amounts for children's and cancer hospitals and the market basket for RNHCIs under Sec. 403.752(a) of the regulations. This proposal reflected our belief that it is best to use an index that most closely represents the cost structure of children's and cancer hospitals and RNHCIs. The FY 2002 cost weights for wages and salaries, professional liability, and ``all other'' for children's and cancer hospitals are noticeably closer to those in the IPPS operating market basket than those in the excluded hospital market basket, which is based on the cost structure of IRFs, LTCHs, IPFs, and children's and cancer hospitals and RNHCIs. Therefore, as proposed, for this final rule we are using the IPPS operating market basket to update the target amounts for children's and cancer hospitals and the market basket for RNHCIs under Sec. 403.752(a) of the regulations. However, when we compare the weights for LTCHs and IPFs to the weights for IPPS hospitals, we did not find them comparable. Therefore, we did not believe it was appropriate to use the IPPS market basket for LTCHs and IPFs to update the portion of their payment that is based on reasonable cost.

For similar reasons, we indicated in the proposed rule that we are considering at some other date proposing a separate market basket to update the adjusted Federal payment amount for IRFs, LTCHs, and IPFs. We expect that these changes would be proposed in separate proposed rules for each of these three hospital types. We envision that these changes should apply to the adjusted Federal payment rate, and not the portion of the payment that is based on a facility-specific (or reasonable cost) payment to the extent such a hospital or unit is paid under a blend methodology. In other words, to the extent any of these hospitals are paid under a blend methodology whereby a percentage of the payment is based on reasonable cost principles, we would not propose to make changes to the existing methodology for developing the market basket for the reasonable cost portion of the payment because this portion of the payment is being phased out, if it is not already a nonexistent feature of the PPSs for IRFs, LTCHs, and IPFs. As indicated in the proposed rule, we do not believe that it makes sense to propose to create an entirely new methodology for creating the market basket index which updates the ``reasonable cost'' portion of a blend methodology since the ``reasonable cost portion'' will last at most for 1 or 3 additional years (1 year for LTCHs paid under a blend methodology since some LTCHs only have 1 year remaining in their transition, and 3 years for IPFs since existing IPFs paid under a blend methodology only have 3 years remaining under a blend methodology). However, the same cannot be said for the adjusted Federal payment amount. In the case of the IRF PPS, all IRFs are paid at 100 percent of the adjusted Federal payment amount and will continue to be paid based on 100 percent of this amount under current law. In the LTCH PPS, most LTCHs (98 percent) are already paid at 100 percent of the adjusted Federal payment amount. In the case of the few LTCHs that are paid under a blend methodology for cost reporting periods beginning on or after October 1, 2006, payment will be based entirely on the adjusted Federal prospective payment rate. In the case of IPFs, new IPFs (as defined in Sec. 412.426(c)) will be paid at 100 percent of the adjusted Federal prospective payment rate (the Federal per diem payment amount), while all others will continue to transition to 100 percent of the Federal per diem payment amount. In any event, even those transitioning will be at 100 percent of the adjusted Federal prospective payment rate in 3 years.

Comment: One commenter supported CMS evaluation of a potential new market basket for LTCHs and other post-acute care providers. However, they cautioned CMS to look at the distinct attributes and price inputs of various providers, claiming the price inputs of LTCHs are linked more closely to those of acute care hospitals than other types of providers. They also recommended that CMS use FY 2002 hospital data to calculate the excluded hospital with capital market basket in the 2007 LTCH rate year payment update.

Response: In the RY 2007 LTCH proposed rule, we plan to propose a new market basket for updating the LTCH prospective payments which may be based on 2002 data. The proposed methodology used to create this market basket will be described in detail and is likely going to be similar to the market basket described in the IRF FY 2006 proposed rule. We will also present any additional analysis we have conducted on the differing cost structures of LTCHs

[[Page 47398]]

and other types of providers. This proposed rule will be subject to comments.

Comment: Several commenters disagreed with CMS proposal to use the FY 2002 IPPS operating market basket to update the target amounts for children's and cancer hospitals. One commenter recommended CMS implement a separate market basket for cancer hospitals that would recognize the actual cost increases experienced by these institutions. The commenters contended that the existing excluded market basket falls short of reflecting the annual cost increases actually experienced by cancer hospitals. They have determined this shortfall to be specific cost weights and relative price proxies of pharmaceuticals and compensation. Another commenter recommended using the excluded hospital market basket until new market baskets are implemented for IRFs, IPFs, and LTCHs.

Response: Due to the small number of children's and cancer hospitals and RNCHIs (less than 80 in 2002) and limited reporting, we believe we are unable to create a representative market basket for those hospitals still being paid based solely on their reasonable costs, subject to target amounts. Therefore, we proposed to use the FY 2002 IPPS operating market basket percentage increase to update the target amounts for children's and cancer hospitals and the market basket for RNHCIs under Sec. 403.752(a) of the regulations because this market basket most closely represented the cost structure of children's and cancer hospitals and RNHCIs.

Chart 8 compares the limited data available on median salary, median pharmaceutical, and median professional liability insurance (PLI) cost weights (as a percent of operating costs) for cancer and children's hospitals and RNCHTs; IPPS hospitals; and IRFs, LTCHs, and IPFs. As indicated, the cost structure for cancer and children's hospitals and RNCHIs is more like the cost structure for IPPS hospitals than that for IRFs, LTCHs, and IPFs. Because both the excluded and IPPS market baskets use the same price proxies, a difference in update would be due to the base cost structure. Therefore, by choosing a market basket that most closely represents the cost structures of cancer and children's hospitals and RNCHIs, we are reflecting the annual cost increases experienced by these hospitals.

Chart 8.--Comparison of 2002 Median Cost Weights From the Medicare Cost Reports

Cancer and children's IPPS

IRFs, hospitals hospitals LTCHs, and and RNCHIs

IPFs

Salary Cost Weight.................. 49.486 46.278 55.263 (Number of providers)...............

(68) (3889) (591) Pharmaceutical Cost Weight.......... 6.053 5.453 4.992 (Number of providers)...............

(56) (3891) (585) PLI Cost Weight..................... 1.050 1.099 0.922 (Number of Providers)...............

(75) (2341) (279)

\1\ Costs were included if they were greater than zero and less than operating costs. \2\ Salary cost weights exclude contract labor costs. \3\ The cost weights presented here are medians, which is different than the market basket cost weights which are means (they are calculated by dividing total expenditures for all hospitals by total operating costs for all hospitals).

We will continue to monitor the cost structures of children's and cancer hospitals and RNHCIs to ensure the IPPS hospital market basket adequately reflects these hospitals purchasing patterns. We do not believe it is necessary to postpone the implementation of the IPPS market basket to update the target limits for children's and cancer hospitals and RNCHIs until a new market basket has been implemented to update IRFs, LTCHs, and IPFs payments. The latter group of hospitals are, or soon will be, reimbursed under a PPS that will not affect the reimbursement of children's and cancer hospitals and RNCHIs.

Chart 9 compares the updates for the FY 2002-based IPPS operating market basket, the index we proposed to use to update the target amounts for children's and cancer hospitals, and RNHCIs, with a FY 2002-based excluded hospital market basket that is based on the current methodology (that is, based on the cost structure of IRFs, LTCHs, IPFs, and children's and cancer hospitals). Although the percent change in the IPPS operating market basket is typically lower than the percent change in the FY 2002-based excluded hospital market basket (see charts), we believe it is important to use the market basket that most closely reflects the cost structure of children's and cancer hospitals and RNCHIs. In the FY 2006 IPPS proposed rule, we invited comments on our proposal to use the proposed FY 2002 IPPS operating market basket to update the target amounts for children's and cancer hospitals reimbursed under sections 1886(b)(3)(A) and (b)(3)(E) of the Act and the market basket for RNHCIs under Sec. 403.752(a) of the regulations. The forecasts are based on the GII 2nd quarter, 2005 forecast with historical data through the 1st quarter of 2005, incorporating two more quarters of historical data than published in the FY 2006 IPPS proposed rule. (As we indicated earlier, GII is a nationally recognized economic and financial forecasting firm that contracts with CMS to forecast the components of the market baskets.)

[[Page 47399]]

Chart 9.--FY 2002-Based IPPS and FY 2002-Based Excluded Hospital Operating Index Percent Change, FYs 2000 Through 2007

Rebased FY FY 2002-based 2002-based excluded Fiscal year

IPPS operating hospital market basket market basket

Historical Data:

FY 2000.............................

3.2

3.3

FY 2001.............................

4.1

4.3

FY 2002.............................

3.7

4.2

FY 2003.............................

4.0

4.1

FY 2004.............................

3.9

4.0 ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤ Forecast:

FY 2005.............................

4.2

4.2

FY 2006.............................

3.7

3.8

FY 2007.............................

3.1

3.4 ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤

2. Excluded Hospitals Paid Under a Blend Methodology

As we discuss in greater detail in Appendix B to this final rule, in the past, hospitals and hospital units excluded from the IPPS have been paid based on their reasonable costs, subject to TEFRA limits. However, some of these categories of excluded hospitals and hospital units are now paid under their own PPSs. Specifically, existing LTCHs and existing IPFs are or will be transitioning from reasonable cost- based payments (subject to the TEFRA limits) to prospective payments under their respective PPSs. Under the respective transition period methodologies for the LTCH PPS and the IPF PPS, which are described below, payment is based, in part, on a decreasing percentage of the reasonable cost-based payment amount, which is subject to the TEFRA limits and an increasing percentage of the Federal prospective payment rate. In general, LTCHs and IPFs whose PPS payment is comprised in part of a reasonable cost-based payment will have those reasonable cost- based payment amounts limited by the hospital's TEFRA ceiling.

Effective for cost reporting periods beginning on or after October 1, 2002, LTCHs are paid under the LTCH PPS, which was implemented with a 5-year transition period, transitioning existing LTCHs to a payment based on the fully Federal prospective payment rate (August 30, 2002; 67 FR 55954). However, an existing LTCH may elect to be paid at 100 percent of the Federal prospective rate at the start of any of its cost reporting periods during the 5-year transition period. A ``new'' LTCH is paid based on 100 percent of the standard Federal rate. Effective for cost reporting periods beginning on or after January 1, 2005, IPFs, as defined in Sec. 412.426(c), are paid under the IPF PPS under which they receive payment based on a prospectively determined Federal per diem rate that is based on the sum of the average routine operating, ancillary, and capital costs for each patient day of psychiatric care in an IPF, adjusted for budget neutrality. During a 3-year transition period, existing IPFs are paid based on a blend of the reasonable cost- based payments and the Federal prospective per diem base rate. For cost reporting periods beginning on or after January 1, 2008, existing IPFs are to be paid based on 100 percent of the Federal per diem rate. A ``new'' IPF, as defined in Sec. 412.426(c), is paid based on 100 percent of the Federal per diem payment amount. Any LTCHs or IPFs that receive a PPS payment that includes a reasonable cost-based payment during its respective transition period will have that portion of its payment subject to the TEFRA limits.

Under the broad authority of sections 1886(b)(3)(A) and (b)(3)(B) of the Act, as was proposed, for LTCHs and IPFs that are transitioning to the fully Federal prospective payment rate, we are using the rebased FY 2002-based excluded hospital market basket to update the reasonable cost-based portion of their payments. The market basket update is described in detail below. We do not believe the IPPS operating market basket should be used for the update to the reasonable cost-based portion of the payments to LTCHs or IPFs because this market basket does not reflect the cost structure of LTCHs and IPFs. Chart 8 compares the median salary, median pharmaceutical, and median professional liability insurance cost weights for IPPS hospitals and IRFs, LTCHs, and IPFs.

Comment: One commenter endorsed the CMS proposal to rebase the excluded hospital market basket, stating that rebasing the excluded hospital market basket improves accuracy and predictability of the LTCH PPS. The commenter also hoped that the forecast for the final rule for FY 2006 will be higher than the proposed rule's forecast of 3.2 percent.

Response: We agree that the market baskets should be periodically rebased to ensure they adequately reflect the purchasing patterns of hospitals and the price increases associated with providing hospital services. The 2002-based excluded hospital's FY 2006 forecast was run on the GII second quarter forecast for 2005, with historical data through the first quarter of 2005, incorporating two more quarters of historical data than published in the FY 2006 IPPS proposed rule. The forecast for FY 2006 for the FY 2002-based excluded hospital market basket is 3.8 percent. 3. Development of Cost Categories and Weights for the FY 2002-Based Excluded Hospital Market Basket a. Medicare Cost Reports

In this final rule, as was proposed, the major source of expenditure data for developing the rebased and revised excluded hospital market basket cost weights is the FY 2002 Medicare cost reports. We chose FY 2002 as the base year because we believe this is the most recent, relatively complete year (with a 90-percent reporting rate) of Medicare

[[Page 47400]]

cost report data. These cost reports are from rehabilitation, psychiatric, long-term care, children's, cancer, and RNHCIs. They do not reflect data from IPPS hospitals or CAHs. These are the same hospitals included in the FY 1997-based excluded hospital market basket, except for RNHCIs. Due to insufficient Medicare cost report data for these excluded hospitals, their cost reports yield only four major expenditure or cost categories: Wages and salaries, pharmaceuticals, professional liability insurance (malpractice), and a residual ``all other.''

Since the cost weights for the FY 2002-based excluded hospital market basket are based on facility costs, as we proposed, in this final rule, we are using those cost reports for IRFs, LTCHs, and children's, cancer, and RNHCIs whose Medicare average length of stay is within 15 percent (that is, 15 percent higher or lower) of the total facility average length of stay for the hospital. We use a less stringent edit for Medicare length of stay for IPFs, requiring the average length of stay to be within 30 or 50 percent (depending on the total facility average length of stay) of the total facility length of stay. This allows us to increase our sample size by over 150 reports and produce a cost weight more consistent with the overall facility. The edit we applied to IPFs when developing the FY 1997-based excluded hospital market basket was based on the best available data at the time.

We believe that limiting our sample to hospitals with a Medicare average length of stay within a comparable range of the total facility average length of stay provides a more accurate reflection of the structure of costs for Medicare treatments. Our method results in including in our data set hospitals with a share of Medicare patient days relative to total patient days that was approximately three times greater than for those hospitals excluded from our sample. Our goal is to measure cost shares that are reflective of case-mix and practice patterns associated with providing services to Medicare beneficiaries.

As was proposed, cost weights for benefits, contract labor, and blood and blood products were derived using the FY 2002-based IPPS market basket. This is necessary because these data are poorly reported in the cost reports for non-IPPS hospitals. For example, the ratio of the benefit cost weight to the wages and salaries cost weight was applied to the excluded hospital wages and salaries cost weight to derive a benefit cost weight for the excluded hospital market basket.

Chart 10.--Major Cost Categories Found in Excluded Hospital Medicare Cost Reports

FY 1997-based FY 2002-based excluded

excluded Major cost categories

hospital

hospital market basket market basket

Wages and salaries......................

51.998

57.037 Professional Liability Insurance

0.805

1.504 (Malpractice).......................... Pharmaceuticals.........................

6.940

5.940 All other...............................

40.257

35.519

b. Other Data Sources

In addition to the Medicare cost reports, the other source of data used in developing the excluded hospital market basket weights is the Benchmark Input-Output Tables (I-Os) created by the Bureau of Economic Analysis, U.S. Department of Commerce.

New data for this source are scheduled for publication every 5 years, but may take up to 7 years after the reference year. Only an Annual I-O is produced each year, but the Annual I-O contains less industry detail than does the Benchmark I-O. When we rebased the excluded hospital market basket using FY 1997 data in the FY 2003 IPPS final rule, the 1997 Benchmark I-O was not yet available. Therefore, as was proposed, for this final rule, we did not incorporate data from that source into the FY 1997-based excluded hospital market basket (67 FR 50033). However, we did use a secondary source, the 1997 Annual Input-Output tables. The third source of data, the 1997 Business Expenditure Survey (now known as the Business Expenses Survey), was used to develop weights for the utilities and telephone services categories.

The 1997 Benchmark I-O data are a much more comprehensive and complete set of data than the 1997 Annual I-O estimates. The 1997 Annual I-O is an update of the 1992 I-O tables, while the 1997 Benchmark I-O is an entirely new set of numbers derived from the 1997 Economic Census. The 2002 Benchmark Input-Output tables are not yet available. Therefore, we used the 1997 Benchmark I-O data in the FY 2002-based excluded hospital market basket, to be effective for FY 2006. Instead of using the less detailed, less accurate Annual I-O data, we aged the 1997 Benchmark I-O data forward to FY 2002. As was proposed, the methodology we used to age the data for this final rule involves applying the annual price changes from the price proxies to the appropriate cost categories. We repeat this practice for each year.

The ``all other'' cost category is further divided into other hospital expenditure category shares using the 1997 Benchmark Input- Output tables. Therefore, the ``all other'' cost category expenditure shares are proportional to their relationship to ``all other'' totals in the I-O tables. For instance, if the cost for telephone services were to represent 10 percent of the sum of the ``all other'' I-O (see below) hospital expenditures, then telephone services would represent 10 percent of the market basket's ``all other'' cost category. The remaining detailed cost categories under the residual ``all other'' cost category were derived using the 1997 Benchmark Input-Output Tables aged to FY 2002 using relative price changes. 4. FY 2002-Based Excluded Hospital Market Basket--Selection of Price Proxies

After computing the FY 2002 cost weights for the rebased excluded hospital market basket, it is necessary to select appropriate wage and price proxies to reflect the rate-of-price change for each expenditure category. With the exception of the Professional Liability proxy, as was proposed, all the indicators are based on Bureau of Labor Statistics (BLS) data and are grouped into one of the following BLS categories:

Producer Price Indexes--Producer Price Indexes (PPIs) measure price changes for goods sold in other than retail markets. PPIs are preferable price proxies for goods that hospitals purchase as inputs in producing their outputs because the PPIs would better reflect the prices faced by hospitals. For example, we use a special PPI for

[[Page 47401]]

prescription drugs, rather than the Consumer Price Index (CPI) for prescription drugs because hospitals generally purchase drugs directly from the wholesaler. The PPIs that we use measure price change at the final stage of production.

Consumer Price Indexes--Consumer Price Indexes (CPIs) measure change in the prices of final goods and services bought by the typical consumer. Because they may not represent the price faced by a producer, we used CPIs only if an appropriate PPI was not available, or if the expenditures were more similar to those of retail consumers in general rather than purchases at the wholesale level. For example, the CPI for food purchased away from home is used as a proxy for contracted food services.

Employment Cost Indexes--Employment Cost Indexes (ECIs) measure the rate of change in employee wage rates and employer costs for employee benefits per hour worked. These indexes are fixed-weight indexes and strictly measure the change in wage rates and employee benefits per hour. Appropriately, they are not affected by shifts in employment mix. We made no changes to the proposed price proxies in this final rule. We evaluated the price proxies using the criteria of reliability, timeliness, availability, and relevance. Reliability indicates that the index is based on valid statistical methods and has low sampling variability. Timeliness implies that the proxy is published regularly, at least once a quarter. Availability means that the proxy is publicly available. Finally, relevance means that the proxy is applicable and representative of the cost category weight to which it is applied. The CPIs, PPIs, and ECIs selected meet these criteria and, therefore, we believe they continue to be the best measure of price changes for the cost categories to which they are applied.

Chart 11 sets forth the complete FY 2002-based excluded hospital market basket including cost categories, weights, and price proxies. For comparison purposes, the corresponding FY 1997-based excluded hospital market basket is listed as well. A summary outlining the choice of the various proxies follows the charts.

Chart 11.--FY 2002-Based Excluded Hospital Market Basket Cost Categories, Weights, and Proxies With FY 1997- Based Excluded Hospital Market Basket Used for Comparison

FY 1997-based FY 2002-based excluded

excluded Expense categories

hospital

hospital FY 2002-based excluded hospital market basket market basket market basket price proxies weights

weights

1. Compensation...............................

63.251

71.035

C. Wages and Salaries*....................

51.998

57.037 ECI--Wages and Salaries, Civilian Hospital Workers.

D. Employee Benefits*.....................

11.253

13.998 ECI--Benefits, Civilian Hospital Workers. 2. Professional Fees*.........................

4.859

3.543 ECI--Compensation for Professional, Specialty & Technical Workers. 3. Utilities..................................

1.296

0.804

A. Fuel, Oil, and Gasoline................

0.272

0.132 PPI Refined Petroleum Products.

B. Electricity............................

0.798

0.430 PPI Commercial Electric Power.

C. Water and Sewerage.....................

0.226

0.242 CPI-U Water & Sewerage Maintenance. 4. Professional Liability Insurance...........

0.805

1.504 CMS Professional Liability Insurance Premium Index. 5. All Other..................................

29.790

23.114

B. All Other Products.....................

19.680

15.836 (1) Pharmaceuticals...........................

6.940

5.940 PPI Prescription Drugs. (2) Direct Purchase Food......................

1.233

1.070 PPI Processed Foods & Feeds. (3) Contract Service Food.....................

1.146

0.759 CPI-U Food Away From Home. (4) Chemicals.................................

2.343

1.347 PPI Industrial Chemicals. (5) Blood and Blood Products**................

0.821 (6) Medical Instruments.......................

1.972

1.242 PPI Medical Instruments & Equipment. (7) Photographic Supplies.....................

0.184

0.118 PPI Photographic Supplies. (8) Rubber and Plastics.......................

1.501

1.289 PPI Rubber & Plastic Products. (9) Paper Products............................

1.219

1.225 PPI Converted Paper & Paperboard Products. (10) Apparel..................................

0.525

0.253 PPI Apparel. (11) Machinery and Equipment..................

0.936

0.364 PPI Machinery & Equipment. (12) Miscellaneous Products**.................

0.860

2.230 PPI Finished Goods less Food and Energy.

B. All Other Services.....................

10.110

7.279 (1) Telephone Services........................

0.382

0.295 CPI-U Telephone Services. (2) Postage...................................

0.771

0.836 CPI-U Postage. (3) All Other: Labor Intensive*...............

4.892

2.718 ECI-Compensation for Private Service Occupations. (4) All Other: Non-Labor Intensive............

4.065

3.430 CPI-U All Items.

Total.....................................

100.000

100.000

* Labor-Related ** Blood and blood products, previously a separate cost category, is now contained within Miscellaneous Products in the FY 2002-based excluded hospital market basket.

a. Wages and Salaries

For measuring the price growth of wages in the FY 2002-based excluded hospital market basket, we used the ECI for wages and salaries for civilian hospital workers as the proxy for wages. This same proxy was used for the FY 1997-based excluded hospital market basket. b. Employee Benefits

The FY 2002-based excluded hospital market basket uses the ECI for employee benefits for civilian hospital workers. This is the same proxy that was used in

[[Page 47402]]

the FY 1997-based excluded hospital market basket. c. Nonmedical Professional Fees

The ECI for compensation for professional and technical workers in private industry is applied to this category because it includes occupations such as management and consulting, legal, accounting and engineering services. The same proxy was used in the FY 1997-based excluded hospital market basket. d. Fuel, Oil, and Gasoline

The percentage change in the price of gas fuels as measured by the PPI (Commodity Code 0552) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. e. Electricity

The percentage change in the price of commercial electric power as measured by the PPI (Commodity Code 0542) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. f. Water and Sewerage

The percentage change in the price of water and sewerage maintenance as measured by the CPI for all urban consumers (CPI Code CUUR0000SEHG01) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. g. Professional Liability Insurance

The FY 2002-based excluded hospital market basket uses the percentage change in the hospital professional liability insurance (PLI) premiums as estimated by the CMS Hospital Professional Liability Index for the proxy of this category. Similar to the Physicians Professional Liability Index, we attempt to collect commercial insurance premiums for a fixed level of coverage, holding nonprice factors constant (such as a change in the level of coverage). In the FY 1997-based excluded hospital market basket, the same price proxy was used.

We continue to research options for improving our proxy for professional liability insurance. This research includes exploring various options for expanding our current survey, including the identification of another entity that would be willing to work with us to collect more complete and comprehensive data. We are also exploring other options such as third party or industry data that might assist us in creating a more precise measure of PLI premiums. At this time, we have not yet identified a preferred option. Therefore, we are not making any changes to the proxy in this final rule. h. Pharmaceuticals

The percentage change in the price of prescription drugs as measured by the PPI (PPI Code PPI32541DRX) is used as a proxy for this category. This is a special index produced by BLS and is the same proxy used in the FY 1997-based excluded hospital market basket. i. Food: Direct Purchases

The percentage change in the price of processed foods and feeds as measured by the PPI (Commodity Code 02) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. j. Food: Contract Services

The percentage change in the price of food purchased away from home as measured by the CPI for all urban consumers (CPI Code CUUR0000SEFV) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. k. Chemicals

The percentage change in the price of industrial chemical products as measured by the PPI (Commodity Code 061) is applied to this component. While the chemicals hospitals purchase include industrial as well as other types of chemicals, the industrial chemicals component constitutes the largest proportion by far. Thus, we believe that Commodity Code 061 is the appropriate proxy. The same proxy was used in the FY 1997-based excluded hospital market basket. l. Medical Instruments

The percentage change in the price of medical and surgical instruments as measured by the PPI (Commodity Code 1562) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. m. Photographic Supplies

The percentage change in the price of photographic supplies as measured by the PPI (Commodity Code 1542) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. n. Rubber and Plastics

The percentage change in the price of rubber and plastic products as measured by the PPI (Commodity Code 07) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. o. Paper Products

The percentage change in the price of converted paper and paperboard products as measured by the PPI (Commodity Code 0915) is used. The same proxy was used in the FY 1997-based excluded hospital market basket. p. Apparel

The percentage change in the price of apparel as measured by the PPI (Commodity Code 381) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. q. Machinery and Equipment

The percentage change in the price of machinery and equipment as measured by the PPI (Commodity Code 11) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. r. Miscellaneous Products

The percentage change in the price of all finished goods less food and energy as measured by the PPI (Commodity Code SOP3500) is applied to this component. Using this index removes the double-counting of food and energy prices, which are already captured elsewhere in the market basket. The same proxy was used in the FY 1997-based excluded hospital market basket. The weight for this cost category is higher than in the FY 1997-based index because it also includes blood and blood products. In the FY 1997-based excluded hospital market basket, we included a separate cost category for blood and blood products, using the BLS PPI (Commodity Code 063711) for blood and derivatives as a price proxy. A review of recent trends in the PPI for blood and derivatives suggests that its movements may not be consistent with the trends in blood costs faced by hospitals. While this proxy did not match exactly with the product hospitals are buying, its trend over time appears to be reflective of the historical price changes of blood purchased by hospitals. However, an apparent divergence over recent periods led us to reevaluate whether the PPI for blood and derivatives was an appropriate measure of the changing price of blood. We ran test market baskets classifying blood in three separate cost categories: blood and blood products, contained within chemicals as was done for the FY 1992-based index, and within miscellaneous products. These categories use as proxies the following PPIs: the PPI for blood and blood products, the PPI for chemicals, and the PPI for finished goods less food and energy, respectively. These three market

[[Page 47403]]

baskets moved similarly. The impact on the overall market basket by using different proxies for blood was negligible, mostly due to the relatively small weight for blood in the market basket. Therefore, we chose the PPI for finished goods less food and energy for the blood proxy because we believe it will best be able to proxy price changes (not quantities or required tests) associated with blood purchased by hospitals. We will continue to evaluate this proxy for its appropriateness and will explore the development of alternative price indexes to proxy the price changes associated with this cost.

We received several comments on including blood and blood products costs in miscellaneous products cost weight. These comments were addressed in section IV.B.1.b.2 of this final rule and are applicable to the FY 2002-based excluded hospital market basket as well because our rationale for how we treat blood and blood products in the IPPS market basket is the same as in the FY 2002-based excluded hospital market basket. s. Telephone

The percentage change in the price of telephone services as measured by the CPI for all urban consumers (CPI Code CUUR0000SEED) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. t. Postage

The percentage change in the price of postage as measured by the CPI for all urban consumers (CPI Code CUUR0000SEEC01) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. u. All Other Services: Labor Intensive

The percentage change in the ECI for compensation paid to service workers employed in private industry is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket. v. All Other Services: Nonlabor Intensive

The percentage change in the all-items component of the CPI for all urban consumers (CPI Code CUUR0000SA0) is applied to this component. The same proxy was used in the FY 1997-based excluded hospital market basket.

For further discussion of the rationale for choosing many of the specific price proxies, we refer the reader to the August 1, 2002 final rule (67 FR 50037).

Chart 12 compares the updates for the FY 2002-based excluded hospital market basket (based on the cost structures of IRFs, LTCHs, IPFs, children's and cancer hospitals, and RNCHIs), the index we proposed to use to update the reasonable cost-based portion of IPF and LTCH payments and which we are adopting in this final rule, with a FY 1997-based excluded hospital market basket (based on the cost structure of IRFs, LTCHs, IPFs, and children's and cancer hospitals).

Chart 12.--FY 1997-Based and FY 2002-Based Excluded Hospital Operating Index Percent Change, FY 2000 Through FY 2008

FY 2002-based FY 1997-based excluded

excluded Fiscal Year (FY)

hospital

hospital market basket market basket

Historical data:

FY 2000.............................

3.3

3.3

FY 2001.............................

4.3

4.3

FY 2002.............................

4.2

3.9

FY 2003.............................

4.1

4.0

FY 2004.............................

4.0

3.9 ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤ Forecast:

FY 2005.............................

4.2

4.2

FY 2006.............................

3.8

3.8

FY 2007.............................

3.4

3.2

FY 2008.............................

3.2

3.0 ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤

TL0505.SIM

D. Frequency of Updates of Weights in IPPS Hospital Market Basket

Section 404 of Pub. L. 108-173 (MMA) requires CMS to report in this final rule the research that has been done to determine a new frequency for rebasing the hospital market basket. Specifically, section 404 states:

``(a) More frequent updates in weights. After revising the weights used in the hospital market basket under section 1886(b)(3)(B)(iii) of the Social Security Act (42 U.S.C. 1395ww(b)(3)(B)(iii)) to reflect the most current data available, the Secretary shall establish a frequency for revising such weights, including the labor share, in such market basket to reflect the most current data available more frequently than once every 5 years; and

``(b) Incorporation of explanation in rulemaking. The Secretary shall include in the publication of the final rule for payment for inpatient hospitals services under section 1886(d) of the Social Security Act (42 U.S.C. 1395ww(d)) for fiscal year 2006, an explanation of the reasons for, and options considered, in determining the frequency established under subsection (a).''

This section of the final rule discusses the research we have done to fulfill this requirement, and sets forth a rebasing frequency that makes optimal use of available data.

Our past practice has been to monitor the appropriateness of the market basket on a consistent basis in order to rebase and revise the index when necessary. The decision to rebase and revise the index has been driven in large part by the availability of the data necessary to produce a complete index. In the past, we have supplemented the Medicare cost report data that are available on an annual basis with Bureau of the Census hospital expense data that are typically available only every 5 years (usually in

[[Page 47404]]

years ending in 2 and 7). Because of this, we have generally rebased the index every 5 years. However, prior to the requirement associated with section 404 of Pub. L. 108-173, there was no legislative requirement regarding the timing of rebasing the hospital market basket nor was there a hard rule that we used in determining this frequency. ProPAC, one of MedPAC's predecessor organizations, submitted a report to the Secretary on April 1, 1985, that supported periodic rebasing at least every 5 years.

The most recent rebasing of the hospital market basket was just 3 years ago, for the FY 2003 update. Since its inception with the hospital PPS in FY 1984, the hospital market basket has been rebased several times (FY 1987 update, FY 1991 update, FY 1997 update, FY 1998 update, and FY 2003 update). One of the reasons we believe it appropriate to rebase the index on a periodic basis is that rebasing (as opposed to revising, as explained in section IV.A. of this preamble) tends to have only a minor impact on the actual percentage increase applied to the PPS update. There are two major reasons for this: (1) The cost category weights tend to be relatively stable over shorter term periods (3 to 5 years); and (2) the update is based on a forecast, which means the individual price series tend not to grow as differently as they have in some historical periods.

We focused our research in two major areas. First, we reviewed the frequency and availability of the data needed to produce the market basket. Second, we analyzed the impact on the market basket of determining the market basket weights under various frequencies. We did this by developing market baskets that had base years for every year between 1997 and 2002, and then analyzed how different the market basket percent changes were over various periods. We used the results from these areas of research to assist in our determination of a new rebasing frequency. Based on this analysis, as we proposed in the FY 2006 IPPS proposed rule, we would rebase the hospital market basket every 4 years. This would mean the next rebasing would occur for the FY 2010 update.

As we have described in numerous Federal Register documents over the past few decades, the hospital market basket weights are the compilation of data from more than one data source. When we are discussing rebasing the weights in the hospital market basket, there are two major data sources: (1) the Medicare cost reports; and (2) expense surveys from the Bureau of the Census (the Economic Census is used to develop data for the Bureau of Economic Analysis' input-output series).

Each Medicare-participating hospital submits a Medicare cost report to CMS on an annual basis. It takes roughly 2 years before ``nearly complete'' Medicare cost report data are available. For example, approximately 90 percent of FY 2002 Medicare cost report data were available in October 2004 (only 50 percent of FY 2003 data was available), although only 20 percent of these reports were settled. We choose FY 2002 as the base year because we believe this is the most recent, relatively complete year (with a 90 percent reporting rate) of Medicare cost report data. In developing the hospital market basket weights, we have used the Medicare cost reports to determine the weights for six major cost categories (wages, benefits, contract labor, pharmaceuticals, professional liability, and blood and blood products). In FY 2002, these six categories accounted for 68.5 percent of the hospital market basket. Therefore, it is possible to develop a new set of market basket weights for these categories on an annual basis, but with a substantial lag (for the FY 2006 update, we consider the latest year of historical data to be FY 2002).

The second source of data is the U.S. Department of Commerce, Bureau of Economic Analysis' Benchmark Input-Output (I-O) table. These data are published every 5 years with a more significant lag than the Medicare cost reports. For example, the 1997 Benchmark I-O tables were not published until the beginning of 2003. We have sometimes used data from a third data source, the Bureau of the Census' Business Expenses Survey (BES), which is also published every 5 years. The BES data are used as an input into the I-O data, and thus are published a few months prior to the release of the I-O. However, the BES contains only a fraction of the detail contained in the I-O.

Chart 13 below takes into consideration the expected availability of these major data sources and summarizes how they could be incorporated into the development of future market basket weights.

Chart 13.--Expected Future Data Availability for Major Data Sources Used in the Hospital Market Basket

PPS FY Update................................. FY 2006 FY 2007 FY 2008 FY 2009 FY 2010 FY 2011 Market Basket Base Year....................... FY 2002 FY 2003 FY 2004 FY 2005 FY 2006 FY 2007 Medicare Cost Report Data Available........... FY 2002 FY 2003 FY 2004 FY 2005 FY 2006 FY 2007 I-O Data Available............................ 1997 1997 1997 1997 1997 2002 BES Data Available............................ 1997 1997 1997 1997 1997 2002 Number of Years Data Must Be Aged.............

5

6

7

8

9

5

FPS FY Update............................................ FY 2012 FY 2013 FY 2014 FY 2015 FY 2016 Market Basket Base Year.................................. FY 2008 FY 2009 FY 2010 FY 2011 FY 2012 Medicare Cost Report Data Available...................... FY 2008 FY 2009 FY 2010 FY 2011 FY 2012 I-O Data Available....................................... 2002 2002 2002 2002 2007 BES Data Available....................................... 2002 2002 2002 2002 2007 Number of Years Data Must Be Aged........................

6

7

8

9

5

It would be necessary to age the I-O or BES data to the year for which cost report data are available using the price changes between those periods. While not a preferred method in developing the market basket weights, we have done this in the past when rebasing the index. For instance, we have aged the 1997 Benchmark I-O data for this final rule.

As the table clearly indicates, the most optimal rebasing frequency from a data availability standpoint is every 5 years. That is, if we were to next rebase for the FY 2011 update, we could use the 2002 Benchmark I-O data that would recently be available. In order to match the Medicare cost report data that would be available at that time (FY 2007 data), we would have to age the I-O data to FY 2007. However, this would be aging the data only 5 years, whereas if the rebasing frequency was determined to be every 4 years, we would have to

[[Page 47405]]

age 1997 I-O data to FY 2006. While aging data over 5 years is problematic (there can be significant utilization and intensity changes over that length period, as opposed to only one or two years), it would be significantly worse to age data over an 8-year or 9-year period. If we were on a 5-year rebasing frequency, for the FY 2016 update, we would use cost report data for FY 2012 and the newly available 2007 I-O data. Again, the I-O data would have to be aged only 5 years to match the cost report data.

We systematically examined at the implications of determining a rebasing frequency of every 3 or 4 years. Considering a frequency of 3 years first, we would next rebase for the FY 2009 update using FY 2005 Medicare cost report data and 1997 I-O data (the same data currently being used in the FY 2002-based market basket). This is problematic because the 1997 I-O data would need to be aged 8 years to match the cost report data. The next two rebasings would be for the FY 2012 update (using FY 2008 cost report data and 2002 I-O data) and FY 2015 (using FY 2011 cost report data and 2002 I-O data). This means that while we are making optimal use of the Medicare cost report data, we would be forced to use the same I-O data in consecutive rebasings and would have to age that data as much as 9 years to use the same year as the cost report data.

For a rebasing frequency of every 4 years, our next rebasing would be for the FY 2010 update using FY 2006 Medicare cost report data and 1997 I-O data. This is also problematic because the 1997 I-O data would need to be aged 9 years to match the cost report data. The next two rebasings would be for the FY 2014 update (using FY 2010 cost report data and 2002 I-O data) and FY 2018 (using FY 2014 cost report data and 2007 I-O data). Again, this frequency would make optimal use of the Medicare cost report data but would require aging of the I-O data between 7 and 9 years in order to match the cost report data.

It is clear from this analysis that neither the 3-year nor 4-year rebasing frequencies optimize the timeliness of the data relative to rebasing every 5 years. In addition, when comparing the 3-year and 4- year rebasing frequencies, no one method stands out as being significantly improved over another. Thus, this analysis does not lead us to draw any definitive conclusions as to a rebasing frequency more appropriate than every 5 years.

Our second area of research in determining a new rebasing frequency was to analyze the impact on the market basket of determining the market basket weights under various frequencies. We did this by using the current historical data that are available (both Medicare cost report and I-O) to develop market baskets with base year weights for each year between FY 1997 and FY 2002. We then analyzed how differently the market baskets moved over various historical periods.

Approaching the analysis this way allowed us to develop six hypothetical market baskets with different base years (FY 1997, FY 1998, FY 1999, FY 2000, FY 2001, and FY 2002). As we have done when developing the official market baskets, we used Medicare cost report data where available. Thus, cost report data were used to determine the weights for wages and salaries, benefits, contract labor, pharmaceuticals, blood and blood products, and all other costs. We used the 1997 Benchmark I-O data to fill out the remainder of the market basket weights (note that this produces a different index for FY 1997 than the official FY 1997-based hospital market basket that used the Annual 1997 I-O data), aging the data to the appropriate year to match the cost report data. This means the FY 2002-based index used in this analysis matches the FY 2002-based market basket we are using in this final rule. Chart 14 shows the weights from these hypothetical market baskets:

BILLING CODE 4120-01-P

[[Page 47406]]

[GRAPHIC] [TIFF OMITTED] TR12AU05.002

BILLING CODE 4120-01-C

Note that the weights remain relatively stable between periods. It is for this reason that we believe defining the market basket as a Laspeyres-type, fixed-weight index is appropriate. Because the weights in the market basket are generally for aggregated costs (for example, wages and salaries for all employees), there is not much volatility in the weights between periods, especially over shorter time spans. As the results of this analysis will show, rebasing the market basket more frequently than every 5 years is expected to have little impact on the overall percent change in the hospital market basket.

Using these hypothetical market baskets, we can produce market basket percent changes over historical periods to determine what is the impact of using various base periods. In our analysis, we consider the hypothetical FY 1997-based index to be the benchmark measure and the other indexes to indicate the impact of rebasing over various frequencies. The hypothetical FY 2000-based index would reflect the impact of rebasing every 3 years, the hypothetical FY 2001-based index would reflect the impact of rebasing every 4 years, and the hypothetical FY 2002-based index would reflect the

[[Page 47407]]

impact of rebasing every 5 years. Chart 15 shows the results of these comparisons.

Chart 15.--Comparison of Hypothetical Market Baskets, FY 1997 Through FY 2002 Base Years, Percent Changes, FY 1998 Through FY 2004

Percent Change in Hypothetical Market Baskets Federal Fiscal Year

----------------------------------------------------------------------------------------------- FY 1997-based FY 1998-based FY 1999-based FY 2000-based FY 2001-based FY 2002-based

1998....................................................

2.7

2.6

2.6

2.6

2.6

2.6 1999....................................................

2.7

2.5

2.5

2.5

2.5

2.5 2000....................................................

3.2

3.3

3.3

3.3

3.3

3.2 2001....................................................

4.2

4.2

4.2

4.2

4.2

4.1 2002....................................................

3.8

3.8

3.8

3.7

3.7

3.7 2003....................................................

3.9

4.0

4.0

4.0

4.0

4.0 2004....................................................

3.9

3.8

3.9

3.8

3.9

3.9 ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤

Source: Global Insight, Inc, 2nd Qtr. 2005;@USMACRO/CNTL0605 @CISSIM/TL0505.SIM.

It is clear from this comparison that there is little difference between the indexes, and, for some FYs, there would be no difference in the market basket update factor if we had rebased the market basket more frequently. In particular, there is no difference in the hypothetical indexes based between FY 2000 and FY 2002. This suggests that setting the rebasing frequency to 3, 4, or 5 years will have little or no impact on the resulting market basket. As we found when analyzing data availability, this portion of our research does not suggest that rebasing the market basket more frequently than every 5 years results in an improved market basket or that there is any noticeable difference between rebasing every 3 or 4 years.

Market basket rebasing is a 1-year to 2-year long process that includes data processing, analytical work, methodology reevaluation, and regulatory process. After developing a rebased and revised market basket, there are extensive internal review processes that a rule must undergo, both in proposed and final form. Once the proposed rule has been published, there is a 60-day comment period set aside for the public to respond to the proposed rule. After comments are received, we then require adequate time to research and reply to all comments submitted. The last part of the regulatory process is the 60-day requirement that is, the final rule must be published 60 days before the provisions of the rule can become effective.

We would like to rebase all of our indexes (PPS operating, PPS capital, excluded hospital with capital, SNFs, HHAs, and Medicare Economic Index) on a regular schedule. Therefore, if we were to choose a 3-year rebasing schedule, we would have to rebase more than one index at a time. This may potentially limit the amount of time and resources we could devote to the market basket rebasing process. In addition, we recognize that, in the future, we may be required to develop additional market baskets that would require frequent rebasing.

Given the number of market baskets we are responsible for rebasing and revising, the regulatory process for each, and the availability of source data, we believe that while it is not necessary, rebasing and revising the hospital market baskets every 4 years is the most appropriate frequency to meet the legislative requirement.

Comment: A few commenters stated there is no compelling reason to rebase the market basket for the FY 2006 update. They requested that CMS begin its 4-year rebasing schedule, beginning with the FY 2007 update (4 years after the last rebasing of the hospital market for the FY 2003 update).

Response: Section 404(a) of Pub. L. 108-173 directs the Secretary to establish a frequency for rebasing the market basket after updating the weights used in the IPPS operating and capital market baskets to reflect the most current available data. Section 404(b) of the Pub. L. 108-173 provides that the Secretary shall include his explanation of the reasons for the frequency of market basket updates in the FY 2006 IPPS final rule. We believe that section 404 of Pub. L. 108-173 requires that we rebase the market basket in the FY 2006 IPPS final rule because we are required to establish a schedule for rebasing the market basket in the FY 2006 IPPS final rule, but may not establish the schedule until after we have rebased the market basket to reflect the most current data available.

Comment: MedPAC urged the Secretary to propose legislation to repeal section 404 of Pub. L. 108-173 requiring the more frequent updating of the market basket. CMS' analysis shows that updating the weights more frequently then every 5 years would make only small differences in its market basket forecasts. In addition, some of the data used in developing the market basket is only available every 5 years, thus a 4-year rebasing schedule could make the market basket weights even more out of date due to the timing of these data sources. Therefore, MedPAC concluded that updating the weights more often than once every 5 years is unnecessary and potentially counterproductive. Other commenters also requested that CMS continue with a 5-year rebasing schedule.

Response: As described in this rule, we agree with the commenters that rebasing the hospital market basket more frequently than every 5 years is unnecessary. However, section 404 of Pub. L. 108-173 requires a shorter frequency, which CMS has set at every 4 years.

E. Capital Input Price Index Section

The Capital Input Price Index (CIPI) was originally described in the September 1, 1992 Federal Register (57 FR 40016). There have been subsequent discussions of the CIPI presented in the May 26, 1993 (58 FR 30448), September 1, 1993 (58 FR 46490), May 27, 1994 (59 FR 27876), September 1, 1994 (59 FR 45517), June 2, 1995 (60 FR 29229), September 1, 1995 (60 FR 45815), May 31, 1996 (61 FR 27466), and August 30, 1996 (61 FR 46196) issues of the Federal Register. The August 1, 2002 (67 FR 50032) rule discussed the most recent revision and rebasing of the CIPI to a FY 1997 base year, which reflects the capital cost structure facing hospitals in that year.

[[Page 47408]]

In this final rule, we are revising and rebasing the CIPI to a FY 2002 base year to reflect the more recent structure of capital costs in hospitals. Unlike the PPS operating market basket, we do not have FY 2002 Medicare cost report data available for the development of the capital cost weights, due to a change in the FY 2002 cost reporting requirements. Rather, we used hospital capital expenditure data for the capital cost categories of depreciation, interest, and other capital expenses for FY 2001 and aged these data to a FY 2002 base year using the relevant vintage-weighted price proxies. As with the FY 1997-based index, we have developed two sets of weights in order to calculate the FY 2002-based CIPI. The first set of weights identifies the proportion of hospital capital expenditures attributable to each expenditure category, while the second set of weights is a set of relative vintage weights for depreciation and interest. The set of vintage weights is used to identify the proportion of capital expenditures within a cost category that is attributable to each year over the useful life of the capital assets in that category. A more thorough discussion of vintage weights is provided later in this section.

Both sets of weights are developed using the best data sources available. In reviewing source data, we determined that the Medicare cost reports provided accurate data for all capital expenditure cost categories. We used the FY 2001 Medicare cost reports for PPS hospitals, aged to FY 2002, excluding expenses from hospital-based subproviders, to determine weights for all three cost categories: depreciation, interest, and other capital expenses. We compared the weights determined from the Medicare cost reports to the 2002 Bureau of the Census' Business Expenses Survey and found the weights to be similar to those developed from the Medicare cost reports.

Lease expenses are not broken out as a separate cost category in the CIPI, but are distributed among the cost categories of depreciation, interest, and other, reflecting the assumption that the underlying cost structure of leases is similar to capital costs in general. As was done in previous rebasings of the CIPI, we assumed 10 percent of lease expenses are overhead and assigned them to the other capital expenses cost category as overhead. The remaining lease expenses were distributed to the three cost categories based on the proportion of depreciation, interest, and other capital expenses to total capital costs, excluding lease expenses.

Depreciation contains two subcategories: building and fixed equipment and movable equipment. The split between building and fixed equipment and movable equipment was determined using the Medicare cost reports. This methodology was also used to compute the FY 1997-based index.

Total interest expense cost category is split between government/ nonprofit and profit interest. The FY 1997-based CIPI allocated 85 percent of the total interest cost weight to government/nonprofit interest, proxied by average yield on domestic municipal bonds, and 15 percent to for-profit interest, proxied by average yield on Moody's Aaa bonds (67 FR 50044). The methodology used to derive this split is explained in the June 2, 1995 issue of the Federal Register (60 FR 29233).

We derived the split using the relative FY 2001 Medicare cost report data on interest expenses for government/nonprofit and profit hospitals. Based on these data, we applied a 75/25 split between government/nonprofit and profit interest. We believe it is important that this split reflects the latest relative cost structure of interest expenses. The split of 75/25 had little (less than 0.1 percent in any given year) or no effect on the annual capital market basket percent change in both the historical and forecasted periods.

Chart 16 presents a comparison of the FY 2002-based CIPI capital cost weights and the FY 1997-based CIPI capital cost weights.

Chart 16.--Comparison of FY 1997-Based and FY 2002-Based CIPI Cost Category Weights

FY 2002

FY 1997 Expense categories

weights

weights

Price proxy

Total......................................

100.00

100.00 Total depreciation.........................

74.58

71.35 Building and fixed equipment depreciation..

36.23

34.22 Boeckh Institutional Construction Index--vintage weighted (23 years). Movable equipment depreciation.............

38.35

37.13 PPI for machinery and equipment-- vintage weighted (11 years). Total interest.............................

19.86

23.46 Government/nonprofit interest..............

14.90

19.94 Average yield on domestic municipal bonds (Bond Buyer 20 bonds)-- vintage weighted (23 years). For-profit interest........................

4.97

3.52 Average yield on Moody's Aaa bonds-- vintage weighted (23 years). Other......................................

5.55

5.19 CPI-U--Residential Rent.

Because capital is acquired and paid for over time, capital expenses in any given year are determined by both past and present purchases of physical and financial capital. The vintage-weighted CIPI is intended to capture the long-term consumption of capital, using vintage weights for depreciation (physical capital) and interest (financial capital). These vintage weights reflect the proportion of capital purchases attributable to each year of the expected life of building and fixed equipment, movable equipment, and interest. We used the vintage weights to compute vintage-weighted price changes associated with depreciation and interest expense.

Vintage weights are an integral part of the CIPI. Capital costs are inherently complicated and are determined by complex capital purchasing decisions, over time, based on such factors as interest rates and debt financing. In addition, capital is depreciated over time instead of being consumed in the same period it is purchased. The CIPI accurately reflects the annual price changes associated with capital costs, and is a useful simplification of the actual capital investment process. By accounting for the vintage nature of capital, we are able to provide an accurate, stable annual measure of price changes. Annual nonvintage price changes for capital are unstable due to the volatility of interest rate changes and, therefore, do not reflect the actual annual price changes for Medicare capital-related costs. CMS' CIPI reflects the underlying stability of the capital acquisition process and provides hospitals with the ability to plan for changes in capital payments.

[[Page 47409]]

To calculate the vintage weights for depreciation and interest expenses, we needed a time series of capital purchases for building and fixed equipment and movable equipment. We found no single source that provides the best time series of capital purchases by hospitals for all of the above components of capital purchases. The early Medicare cost reports did not have sufficient capital data to meet this need. While the AHA Panel Survey provided a consistent database back to 1963, it did not provide annual capital purchases. The AHA Panel Survey provided a time series of depreciation expenses through 1997 which could be used to infer capital purchases over time. From 1998 to 2001, hospital depreciation expenses were calculated by multiplying the AHA Annual Survey total hospital expenses by the ratio of depreciation to total hospital expenses from the Medicare cost reports. Beginning in 2001, the AHA Annual survey began collecting depreciation expenses. We expect to be able to use these data in future rebasings.

In order to estimate capital purchases from AHA data on depreciation expenses, the expected life for each cost category (building and fixed equipment, movable equipment, and interest) is needed to calculate vintage weights. We used FY 2001 Medicare cost reports to determine the expected life of building and fixed equipment and movable equipment. The expected life of any piece of equipment can be determined by dividing the value of the asset (excluding fully depreciated assets) by its current year depreciation amount. This calculation yields the estimated useful life of an asset if depreciation were to continue at current year levels, assuming straight-line depreciation. From the FY 2001 cost reports, the expected life of building and fixed equipment was determined to be 23 years, and the expected life of movable equipment was determined to be 11 years. The FY 1997-based CIPI showed the same expected life for the two categories of depreciation.

Between the publication of the FY 2006 IPPS proposed rule and this final rule, we conducted a further review of the methodology used to derive the useful life of an asset. Based on this brief analysis into the capital cost structures of hospitals, we are not changing the expected life of fixed and moveable assets for the final rule.

As proposed, we used the building and fixed equipment and movable equipment weights derived from FY 2001 Medicare cost reports to separate the depreciation expenses into annual amounts of building and fixed equipment depreciation and movable equipment depreciation. Year- end asset costs for building and fixed equipment and movable equipment were determined by multiplying the annual depreciation amounts by the expected life calculations from the FY 2001 Medicare cost reports. We then calculated a time series back to 1963 of annual capital purchases by subtracting the previous year asset costs from the current year asset costs. From this capital purchase time series, we were able to calculate the vintage weights for building and fixed equipment and movable equipment. Each of these sets of vintage weights is explained in detail below.

For building and fixed equipment vintage weights, the real annual capital purchase amounts for building and fixed equipment derived from the AHA Panel Survey were used. The real annual purchase amount was used to capture the actual amount of the physical acquisition, net of the effect of price inflation. This real annual purchase amount for building and fixed equipment was produced by deflating the nominal annual purchase amount by the building and fixed equipment price proxy, the Boeckh Institutional Construction Index. Because building and fixed equipment have an expected life of 23 years, the vintage weights for building and fixed equipment are deemed to represent the average purchase pattern of building and fixed equipment over 23-year periods. With real building and fixed equipment purchase estimates available back to 1963, we averaged sixteen 23-year periods to determine the average vintage weights for building and fixed equipment that are representative of average building and fixed equipment purchase patterns over time. Vintage weights for each 23-year period are calculated by dividing the real building and fixed capital purchase amount in any given year by the total amount of purchases in the 23- year period. This calculation is done for each year in the 23-year period, and for each of the sixteen 23-year periods. We used the average of each year across the sixteen 23-year periods to determine the 2002 average building and fixed equipment vintage weights for the FY 2002-based CIPI.

For movable equipment vintage weights, the real annual capital purchase amounts for movable equipment derived from the AHA Panel Survey were used to capture the actual amount of the physical acquisition, net of price inflation. This real annual purchase amount for movable equipment was calculated by deflating the nominal annual purchase amount by the movable equipment price proxy, the PPI for Machinery and Equipment. Based on our determination that movable equipment has an expected life of 11 years, the vintage weights for movable equipment represent the average expenditure for movable equipment over an 11-year period. With real movable equipment purchase estimates available back to 1963, twenty-eight 11-year periods were averaged to determine the average vintage weights for movable equipment that are representative of average movable equipment purchase patterns over time. Vintage weights for each 11-year period are calculated by dividing the real movable capital purchase amount for any given year by the total amount of purchases in the 11-year period. This calculation was done for each year in the 11-year period, and for each of the twenty-eight 11-year periods. We used the average of each year across the twenty-eight 11-year periods to determine the average movable equipment vintage weights for the FY 2002-based CIPI.

For interest vintage weights, the nominal annual capital purchase amounts for total equipment (building and fixed, and movable) derived from the AHA Panel and Annual Surveys were used. Nominal annual purchase amounts were used to capture the value of the debt instrument. Because we have determined that hospital debt instruments have an expected life of 23 years, the vintage weights for interest are deemed to represent the average purchase pattern of total equipment over 23- year periods. With nominal total equipment purchase estimates available back to 1963, sixteen 23-year periods were averaged to determine the average vintage weights for interest that are representative of average capital purchase patterns over time. Vintage weights for each 23-year period are calculated by dividing the nominal total capital purchase amount for any given year by the total amount of purchases in the 23- year period. This calculation is done for each year in the 23-year period and for each of the sixteen 23-year periods. We used the average of each year across the sixteen 23-year periods to determine the average interest vintage weights for the FY 2002-based CIPI. The vintage weights for the FY 1997 CIPI and the FY 2002 CIPI are presented in Chart 17.

[[Page 47410]]

Chart 17.--FY 1997 and FY 2002 Vintage Weights for Capital-Related Price Proxies

Building and fixed equipment

Movable equipment

Interest

Year

FY 1997 23 FY 2002 23 FY 1997 11 FY 2002 11 FY 1997 23 FY 2002 23 years

years

years

years

years

years

1.......................................................

0.018

0.021

0.063

0.065

0.007

0.010 2.......................................................

0.021

0.022

0.068

0.071

0.009

0.012 3.......................................................

0.023

0.025

0.074

0.077

0.011

0.014 4.......................................................

0.025

0.027

0.080

0.082

0.012

0.016 5.......................................................

0.026

0.029

0.085

0.086

0.014

0.019 6.......................................................

0.028

0.031

0.091

0.091

0.016

0.023 7.......................................................

0.030

0.033

0.096

0.095

0.019

0.026 8.......................................................

0.032

0.035

0.101

0.100

0.022

0.029 9.......................................................

0.035

0.038

0.108

0.106

0.026

0.033 10......................................................

0.039

0.040

0.114

0.112

0.030

0.036 11......................................................

0.042

0.042

0.119

0.117

0.035

0.039 12......................................................

0.044

0.045 .............. ..............

0.039

0.043 13......................................................

0.047

0.047 .............. ..............

0.045

0.048 14......................................................

0.049

0.049 .............. ..............

0.049

0.053 15......................................................

0.051

0.051 .............. ..............

0.053

0.056 16......................................................

0.053

0.053 .............. ..............

0.059

0.059 17......................................................

0.057

0.056 .............. ..............

0.065

0.062 18......................................................

0.060

0.057 .............. ..............

0.072

0.064 19......................................................

0.062

0.058 .............. ..............

0.077

0.066 20......................................................

0.063

0.060 .............. ..............

0.081

0.070 21......................................................

0.065

0.060 .............. ..............

0.085

0.071 22......................................................

0.064

0.061 .............. ..............

0.087

0.074 23......................................................

0.065

0.061 .............. ..............

0.090

0.076

Total...............................................

1.000

1.000

1.000

1.000

1.000

1.000

After the capital cost category weights were computed, it was necessary to select appropriate price proxies to reflect the rate-of- increase for each expenditure category. Our price proxies for the FY 2002-based CIPI are the same as those used in the FY 1997-based CIPI. We still believe these are the most appropriate proxies for hospital capital costs that meet our selection criteria of relevance, timeliness, availability, and reliability. We ran the FY 2002-based index using the Moody's Aaa bonds average yield and then using the Moody's Baa bonds average yield as proxy for the for-profit interest cost category. There was no difference in the two sets of index percent changes either historically or forecasted. The rationale for selecting these price proxies is explained more fully in the August 30, 1996 final rule (61 FR 46196). The proxies are presented in Chart 18.

Chart 18.--Comparison of FY 1997-Based and FY 2002-Based Capital Input Price Index, Percent Change, FY 1998 Through FY 2007

CIPI, FY 1997- CIPI, FY 2002- Federal fiscal year

based

based

1998....................................

0.9

1.0 1999....................................

0.9

0.9 2000....................................

1.1

1.0 2001....................................

0.9

0.9 2002....................................

0.8

0.7 2003....................................

0.6

0.5 2004....................................

0.6

0.5 Forecast:

2005................................

0.6

0.5

2006................................

1.0

0.8

2007................................

1.0

0.9 Average:

FYs 1998-2004.......................

0.8

0.8

FYs 2005-2007.......................

0.9

0.7

Global Insight, Inc. forecasts a 0.8 percent increase in the FY 2002-based CIPI for 2006, as shown in Chart 17. This is the result of a 1.4 percent increase in projected depreciation prices (building and fixed equipment, and movable equipment) and a 3.3 percent increase in other capital expense prices, partially offset by a 2.3 percent decrease in vintage-weighted interest rates in FY 2006, as indicated in Chart 19. Accordingly, for FY 2006, we have adopted a 0.8 percent increase in the CIPI.

[[Page 47411]]

Chart 19.--CMS Capital Input Price Index Percent Changes, Total and Components, FYs 1995 Through 2007

Depreciation, Total building and Depreciation, Fiscal Year

Total depreciation

fixed

movable

Interest

Other equipment equipment

Weights FY 2002.........................................

1.000

0.7458

0.3623

0.3835

0.1986

0.0556

Vintage-Weighted Price Changes

1995....................................................

1.7

2.7

4.0

1.6

-1.2

2.5 1996....................................................

1.4

2.5

3.8

1.4

-1.8

2.6 1997....................................................

1.3

2.3

3.7

1.2

-2.0

2.8 1998....................................................

1.0

2.1

3.4

0.9

-2.6

3.2 1999....................................................

0.9

1.9

3.2

0.7

-2.6

3.2 2000....................................................

1.0

1.7

3.1

0.4

-1.7

3.4 2001....................................................

0.9

1.5

3.0

0.2

-2.2

4.3 2002....................................................

0.7

1.3

2.9

0.0

-2.4

4.3 2003....................................................

0.5

1.3

2.8

-0.2

-3.0

3.1 2004....................................................

0.5

1.3

2.8

-0.2

-3.3

2.7 Forecast:

2005................................................

0.5

1.3

2.8

-0.1

-3.7

3.0

2006................................................

0.8

1.4

2.7

0.0

-2.3

3.3

2007................................................

0.9

1.3

2.6

0.0

-2.0

3.2

Rebasing the CIPI from FY 1997 to FY 2002 decreased the percent change in the FY 2006 forecast by 0.2 percentage point, from 1.0 to 0.8, as shown in Chart 14. The difference is caused mostly by changes in the relationships between the cost category weights within depreciation and interest. The fixed depreciation cost weight relative to the movable depreciation cost weight and the nonprofit/government interest cost weight relative to the for-profit interest cost weight are both less in the FY 2002-based CIPI. The changes in these relationships have a small effect on the FY 2002-based CIPI percent changes. However, when added together, they are responsible for a negative two-tenths of a percentage point difference between the FY 2002-based CIPI and the FY 1997-based CIPI.

We did not receive any public comments on the CIPI.

V. Other Decisions and Changes to the IPPS for Operating Costs and GME Costs

A. Postacute Care Transfer Payment Policy (Sec. 412.4)

1. Background

Existing regulations at Sec. 412.4(a) define discharges under the IPPS as situations in which a patient is formally released from an acute care hospital or dies in the hospital. Section 412.4(b) defines transfers from one acute care hospital to another, and Sec. 412.4(c) defines transfers to certain postacute care providers. Our policy provides that, in transfer situations, full payment is made to the final discharging hospital and each transferring hospital is paid a per diem rate for each day of the stay, not to exceed the full DRG payment that would have been made if the patient had been discharged without being transferred.

The per diem rate paid to a transferring hospital is calculated by dividing the full DRG payment by the geometric mean length of stay for the DRG. Based on an analysis that showed that the first day of hospitalization is the most expensive (60 FR 45804), our policy provides for payment that is double the per diem amount for the first day (Sec. 412.4(f)(1)). Transfer cases are also eligible for outlier payments. The outlier threshold for transfer cases is equal to the fixed-loss outlier threshold for nontransfer cases, divided by the geometric mean length of stay for the DRG, multiplied by the length of stay for the case, plus one day. The purpose of the IPPS transfer payment policy is to avoid providing an incentive for a hospital to transfer patients to another hospital early in the patients' stay in order to minimize costs while still receiving the full DRG payment. The transfer policy adjusts the payments to approximate the reduced costs of transfer cases. 2. Changes to DRGs Subject to the Postacute Care Transfer Policy (Sec. Sec. 412.4(c) and (d))

Section 1886(d)(5)(J) of the Act provides that, effective for discharges on or after October 1, 1998, a ``qualified discharge'' from one of 10 DRGs selected by the Secretary to a postacute care provider would be treated as a transfer case. This section required the Secretary to define and pay as transfers all cases assigned to one of 10 DRGs selected by the Secretary, if the individuals are discharged to one of the following postacute care settings:

A hospital or hospital unit that is not a subsection 1886(d) hospital. (Section 1886(d)(1)(B) of the Act identifies the hospitals and hospital units that are excluded from the term ``subsection (d) hospital'' as psychiatric hospitals and units, rehabilitation hospitals and units, children's hospitals, long-term care hospitals, and cancer hospitals.)

A SNF (as defined at section 1819(a) of the Act).

Home health services provided by a home health agency, if the services relate to the condition or diagnosis for which the individual received inpatient hospital services, and if the home health services are provided within an appropriate period (as determined by the Secretary).

In the FY 1999 IPPS final rule (63 FR 40975 through 40976), we specified that a patient discharged to home would be considered transferred to postacute care if the patient received home health services within 3 days after the date of discharge. In addition, in the FY 1999 IPPS final rule, we did not include patients transferred to a swing-bed for skilled nursing care in the definition of postacute care transfer cases (63 FR 40977).

Section 1886(d)(5)(J) of the Act directed the Secretary to select 10 DRGs based upon a high volume of discharges to postacute care and a disproportionate use of postacute care services. As discussed in the FY 1999 IPPS final rule, these 10 DRGs were selected in

[[Page 47412]]

1998 based on the MedPAR data from FY 1996. Using that information, we identified and selected the first 20 DRGs that had the largest proportion of discharges to postacute care (and at least 14,000 such transfer cases). In order to select 10 DRGs from the 20 DRGs on our list, we considered the volume and percentage of discharges to postacute care that occurred before the mean length of stay and whether the discharges occurring early in the stay were more likely to receive postacute care. We identified 10 DRGs to be subject to the postacute care transfer rule starting in FY 1999.

Section 1886(d)(5)(J)(iv) of the Act authorizes the Secretary to expand the postacute care transfer policy for FY 2001 or subsequent fiscal years to additional DRGs based on a high volume of discharges to postacute care facilities and a disproportionate use of postacute care services. In the FY 2004 IPPS final rule (68 FR 45412), we expanded the postacute care transfer policy to include additional DRGs. We established the following criteria that a DRG must meet, for both of the 2 most recent years for which data are available, in order to be included under the postacute care transfer policy:

At least 14,000 postacute care transfer cases;

At least 10 percent of its postacute care transfers occurring before the geometric mean length of stay;

A geometric mean length of stay of at least 3 days; and

If a DRG is not already included in the policy, a decline in its geometric mean length of stay during the most recent 5-year period of at least 7 percent.

In the FY 2004 IPPS final rule, we identified 21 new DRGs that met these criteria. We also determined that one DRG from the original group of 10 DRGs (DRG 263) no longer met the volume criterion of 14,000 transfer cases. Therefore, we removed DRGs 263 and 264 (DRG 264 is paired with DRG 263) from the policy and expanded the postacute care transfer policy to include payments for transfer cases in the new 21 DRGs, effective October 1, 2003. As a result, a total of 29 DRGs were subject to the postacute care transfer policy in FY 2004. In the FY 2004 IPPS final rule, we indicated that we would review and update this list periodically to assess whether additional DRGs should be added or existing DRGs should be removed (68 FR 45413).

For FY 2005, we analyzed the available data from the FY 2003 MedPAR file. For the 2 most recent years of available data (FY 2002 and FY 2003), we found that no additional DRGs qualified under the four criteria set forth in the IPPS final rule for FY 2004. We also analyzed the DRGs included under the policy for FY 2004 to determine if they still met the criteria to remain under the policy. In addition, we analyzed the special circumstances arising from a change to one of the DRGs included under the policy in FY 2004.

In the FY 2005 IPPS final rule (69 FR 48942), we deleted DRG 483 (Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth, and Neck Diagnosis) and established the following new DRGs as replacements: DRG 541 (Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth and Neck Diagnoses With Major O.R. Procedure) and DRG 542 (Tracheostomy With Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth and Neck Diagnoses Without Major O.R. Procedure). Cases in the existing DRG 483 were assigned to the new DRGs 541 and 542 based on the presence or absence of a major O.R. procedure, in addition to the tracheostomy code that was previously required for assignment to DRG 483. Specifically, if the patient's case involves a major O.R. procedure (a procedure whose code is included on the list that is assigned to DRG 468 (Extensive O.R. Procedure Unrelated to Principal Diagnosis), except for tracheostomy codes 31.21 and 31.29), the case is assigned to the DRG 541. If the patient does not have an additional major O.R. procedure (that is, if there is only a tracheostomy code assigned to the case), the case is assigned to DRG 542.

Based on data analysis, we determined that neither DRG 541 nor DRG 542 would have enough cases to meet the existing threshold of 14,000 transfer cases for inclusion in the postacute care transfer policy. Nevertheless, we believed the cases that would be incorporated into these two DRGs remained appropriate candidates for application of the postacute care transfer policy and that the subdivision of DRG 483 should not change the original application of the postacute care transfer policy to the cases once included in that DRG. Therefore, for FY 2005, we proposed alternate criteria to be applied in cases where DRGs do not satisfy the existing criteria, for discharges occurring on or after October 1, 2004 (69 FR 28273 and 28374). The proposed new criteria were designed to address situations such as those posed by the split of DRG 483, where there remain substantial grounds for inclusion of cases within the postacute care transfer policy, although one or more of the original criteria may no longer apply. Under the proposed alternate criteria, DRGs 430, 541, and 542 would have qualified for inclusion in the postacute care transfer policy.

In the response to comments on our FY 2005 proposal, we decided not to adopt the proposed alternate criteria for including DRGs under the postacute care transfer policy in the FY 2005 IPPS final rule. Instead we adopted the policy of simply grandfathering, for a period of 2 years, any cases that were previously included within a DRG that has split, when the split DRG qualified for inclusion in the postacute care transfer policy for both of the previous 2 years. Under this policy, the cases that were previously assigned to DRG 483 and that now fall into DRGs 541 and 542 continue to be subject to the policy. Therefore, effective for discharges on or after October 1, 2004, 30 DRGs, including new DRGs 541 and 542, are subject to the postacute care transfer policy. We indicated that we would monitor the frequency with which these cases are transferred to postacute care settings and the percentage of these cases that are short-stay transfer cases. Because we did not adopt the proposed alternate criteria for DRG inclusion in the postacute care transfer policy, DRG 430 (Psychoses) did not meet the criteria for inclusion and has not been subject to the postacute care transfer policy for FY 2005. We also invited comments on how to treat the cases formerly included in a split DRG after the grandfathering period.

We noted that some commenters also suggested that, in place of the proposed alternate criteria, we should adopt a policy of permanently applying the postacute care transfer policy to a DRG once it has initially qualified for inclusion in the policy. These commenters noted that removing DRGs from the postacute care transfer policy makes the payment system less stable and results in inconsistent incentives over time. They also argued that ``a drop in the number of transfers to postacute care settings is to be expected after the transfer policy is applied to a DRG, but the frequency of transfers may well rise again if the DRG is removed from the policy.'' We indicated that we would consider adopting this general policy once we had evaluated the experience with the specific cases that are subject to the grandfathering policy for FY 2005 and FY 2006.

In the FY 2005 IPPS proposed rule, we also called attention to the data concerning DRG 263, which was subject to the postacute care transfer policy until FY 2004. We removed DRG 263

[[Page 47413]]

from the postacute care transfer policy for FY 2004 because it did not have the minimum number of cases (14,000) transferred to postacute care (13,588 transfer cases in FY 2002, with more than 50 percent of transfer cases being short-stay transfers). The FY 2003 MedPAR data show that there were 15,602 transfer cases in the DRG in FY 2003, of which 46 percent were short-stay transfers. Because we removed the DRG from the postacute care transfer policy in FY 2004, it was required to meet all of the criteria to be included under the policy in subsequent fiscal years. The geometric mean length of stay for DRG 263 showed only a 6-percent decrease since 1999. As a result, DRG 263 did not qualify to be included in the policy for FY 2005 under the criteria that were applied in last year's final rule. DRG 263 would have qualified under the volume threshold and percent of short-stay transfer cases under the proposed new alternate criteria contained in the FY 2005 proposed rule. However, it still would not have met the proposed required decline in length of stay to qualify to be added to the policy for FY 2005. We indicated that we would continue to monitor the experience with DRG 263, especially in light of the comment that recommended a general policy of grandfathering cases that qualify under the criteria for inclusion in the postacute care transfer policy.

The table below displays the 30 DRGs that are included in the postacute care transfer policy, effective for discharges occurring on or after October 1, 2004.

DRG

DRG Title

12................................... Degenerative Nervous System Disorders. 14................................... Intracranial Hemorrhage and Stroke with Infarction. 24................................... Seizure and Headache Age >17 With CC. 25................................... Seizure and Headache Age >17 Without CC. 88................................... Chronic Obstructive Pulmonary Disease. 89................................... Simple Pneumonia and Pleurisy Age > 17 With CC. 90................................... Simple Pneumonia and Pleurisy Age >17 Without CC. 113.................................. Amputation for Circulatory System Disorders Except Upper Limb and Toe. 121.................................. Circulatory Disorders With AMI and Major Complication, Discharged Alive. 122.................................. Circulatory Disorders With AMI Without Major Complications Discharged Alive. 127.................................. Heart Failure & Shock. 130.................................. Peripheral Vascular Disorders With CC. 131.................................. Peripheral Vascular Disorders Without CC. 209.................................. Major Joint and Limb Reattachment Procedures of Lower Extremity. 210.................................. Hip and Femur Procedures Except Major Joint Age >17 With CC. 211.................................. Hip and Femur Procedures Except Major Joint Age >17 Without CC. 236.................................. Fractures of Hip and Pelvis. 239.................................. Pathological Fractures and Musculoskeletal and Connective Tissue Malignancy. 277.................................. Cellulitis Age >17 With CC. 278.................................. Cellulitis Age >17 Without CC. 294.................................. Diabetes Age>35. 296.................................. Nutritional and Miscellaneous Metabolic Disorders Age >17 With CC. 297.................................. Nutritional and Miscellaneous Metabolic Disorders Age >17 Without CC. 320.................................. Kidney and Urinary Tract Infections Age >17 With CC. 321.................................. Kidney and Urinary Tract Infections Age >17 Without CC. 395.................................. Red Blood Cell Disorders Age >17. 429.................................. Organic Disturbances and Mental Retardation. 468.................................. Extensive O.R. Procedure Unrelated to Principal Diagnosis. 541 (formerly 483)................... Tracheostomy with Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth and Neck Diagnoses With Major O.R. Procedure. 542 (formerly 483)................... Tracheostomy with Mechanical Ventilation 96+ Hours or Principal Diagnosis Except Face, Mouth and Neck Diagnoses Without Major O.R. Procedure.

For the FY 2006 IPPS proposed rule, we conducted an extensive analysis of the FY 2003 and FY 2004 MedPAR data to monitor the effects of the postacute care transfer policy. We also conducted an overall assessment of the postacute care transfer policy since its inception in FY 1999. Specifically, we examined the relationship between rates of postacute care utilization and the geometric mean length of stay and the relationship between a high volume and a high proportion of postacute care transfers within a DRG considering our experience under the current policy. We also examined whether a decline in the geometric mean length of stay is associated with an increase in the volume and proportion of total cases in a DRG that are discharges to postacute care. We analyzed these data as part of determining whether to retain the criteria that a DRG must have a decline in the geometric mean length of stay of at least 7 percent in the previous 5-year period to be included under the postacute care transfer policy.

Our current criteria for inclusion in the postacute care transfer policy include a requirement that, if a DRG is not already included in the policy, there must be a decline of at least 7 percent in the DRG's geometric mean length of stay during the most recent 5-year period. It has come to our attention that not all DRGs that experience an increase in postacute care utilization also experience a decrease in geometric mean length of stay. In fact, some DRGs with increases in postacute care utilization during the past several years have also experienced an increase in the geometric mean length of stay. The table below lists a number of DRGs that experienced increases in postacute care utilization and increases in the geometric mean length of stay from FY 2002 through FY 2004:

[[Page 47414]]

Percent change Percent change in geometric in postacute DRG

DRG Title

mean length of care stay

utilization

1................ Craniotomy Age >17

5.26

2.70 With CC. 6................ Carpal Tunnel Release

4.76

56.92 15............... Nonspecific CVA and

30.00

27.75 Precerebral Occlusion Without Infarction. 40............... Extraocular

12.50

15.47 Procedures Except Orbit Age >17. 42............... Intraocular

12.75

6.71 Procedures Except Retina, Iris, and Lens. 51............... Salivary Gland

5.56

20.00 Procedures Except Sioloadenectomy. 55............... Miscellaneous Ear,

11.11

22.22 Nose, Mouth, and Throat Procedures. 113.............. Amputation for

2.04

21.25 Circulatory System disorders Except Upper Limb and Toe. 118.............. Cardiac Pacemaker

11.11

30.29 Device Replacement. 223.............. Major Shoulder/Elbow

4.76

36.17 Procedure or Other Upper Extremity Procedure With CC. 317.............. Admittance for Renal

20.00

80.84 Dialysis. 319.............. Kidney and Urinary

4.76

24.49 Tract Neoplasms Without CC. 345.............. Other Male

11.11

94.34 Reproductive System O.R. Procedure Except for Malignancy. 447.............. Allergic Reactions

5.56

16.81 Age >17. 494.............. Laparoscopic

5.26

26.39 Cholecystectomy Without C.D.E. Without CC.

Our current criteria also include a requirement that a DRG have at least 14,000 total postacute care transfer cases in order to be included in the policy. We have examined the data on the numbers of transfers and the percentage of postacute care transfer cases across DRGs. Among the 30 DRGs currently included within the postacute care transfer policy, we found that the percentage of postacute care transfer cases ranges from a low of 15 percent to a high of 76 percent. Among DRGs that are not currently included within the policy, many had a relatively high percentage of postacute care transfer cases in proportion to the total volume of cases for the DRG or a relatively high volume of discharges to postacute care facilities, or both. For this reason, we reviewed the data for all DRGs before we proposed a change to the postacute care transfer payment policy in the FY 2006 proposed rule. As part of this review, we found that:

Of 550 DRGs, 26 have been deactivated and 17 have no cases in the FY 2004 MedPAR files. We did not propose any changes for these DRGs because application of the postacute care transfer policy to them would have no effect.

Of the remaining 507 DRGs, 220 have geometric mean lengths of stay that are less than 3.0 days. Because the transfer payment policy provides 2 times the per diem rate for the first day of care (due to the large proportion of charges incurred on the first day of a patient's treatment), including these DRGs in the transfer policy would be relatively meaningless as they would all receive a full DRG payment. For this reason, we did not propose any changes to the postacute care transfer policy for these DRGs for FY 2006.

Of the remaining 287 DRGs, 64 have fewer than 100 short- stay transfer cases. In addition, 39 of these 64 DRGs have fewer than 50 short-stay transfer cases. Consistent with the statutory guidance, we did not propose any change to how we would apply the postacute care transfer payment policy to these DRGs because we believe that these DRGs do not have a high volume of discharges to postacute care facilities or involve a disproportionate use of postacute care services.

Once we eliminated the DRGs cited above from consideration for the postacute care transfer policy, we examined the characteristics of the remaining 231 DRGs. In the proposed rule, we stated that 223 DRGs were included in this analysis, but subsequently posted a change to the number of eligible DRGs on our Web site. This change reflected that we had inadvertently excluded 8 DRGs. Of these 231 DRGs, we found that these DRGs had three common characteristics:

The DRG had at least 2,000 total postacute care transfer cases.

At least 20 percent of all cases in the DRG were discharged to postacute care settings.

10 percent of all discharges to postacute care were prior to the geometric mean length of stay for the DRG.

Consistent with the statutory guidance giving the Secretary the authority to make a DRG subject to the postacute care transfer policy based on a high volume of discharges to postacute care facilities and a disproportionate use of postacute care services, in the FY 2006 proposed rule, we indicated that we believed these DRGs have characteristics that make them appropriate for inclusion in the postacute care transfer policy. We also indicated that we believed it is appropriate to consider major revisions to the criteria for including a DRG within the postacute care transfer policy. First, our analysis called into question the requirement that a DRG experience a decline in the geometric mean length of stay over the most recent 5- year period. Our findings that some DRGs with increases in postacute care utilization during the past several years have also experienced increases in geometric mean length of stay indicated that this criterion is no longer effective to identify those DRGs that should be subject to the postacute care transfer policy. In addition, our findings about the number of DRGs with relatively high volumes (at least 2,000 cases) and relatively high proportions (at least 20 percent) of postacute care utilization suggested that we should revise the requirement that a DRG have at least 14,000 total postacute care transfer cases to be included within the postacute care transfer policy.

Our analysis did confirm that it is appropriate to maintain the requirement that a DRG must have a geometric mean length of stay of at least 3.0 days in order to be included within the postacute care transfer policy. We believe that this policy should be retained because, under the transfer payment methodology, hospitals receive the entire payment for cases in these DRGs in the first 2 days of the stay. Lowering the limit below 3.0 days would, therefore, have no effect on payment for DRGs with geometric mean lengths-of-stay in this range. For the reasons discussed in the FY 2004 IPPS proposed rule (68 FR 27199) and because it is a common characteristic of DRGs with a large number of cases discharged to postacute care, we also indicated that we would retain the criterion that at least 10 percent of all cases that are transferred to postacute care should be short-stay cases where the patient is transferred before the

[[Page 47415]]

geometric mean length of stay for the DRG. We also continue to believe that both DRGs in a CC/non-CC pair should be subject to the postacute care transfer policy if one of the DRGs meets the criteria for inclusion. By including both DRGs in a CC/non-CC pair, our policy precludes an incentive for hospitals to code cases in ways designed to avoid triggering the application of the policy, for example, by excluding codes that would identify a complicating or comorbid condition in order to assign a case to a non-CC DRG that is not subject to the policy.

Therefore, for the FY 2006 IPPS proposed rule, we considered substantial revisions to the four criteria that are currently used to determine whether a DRG qualifies for inclusion in the postacute care transfer policy. The current criteria provide that, in order to be included within the policy, a DRG must have, for both of the 2 most recent years for which data are available:

At least 14,000 total postacute care transfer cases;

At least 10 percent of its postacute care transfers occurring before the geometric mean length of stay;

A geometric mean length of stay of at least 3 days;

If a DRG is not already included in the policy, a decline in its geometric mean length of stay during the most recent 5-year period of at least 7 percent; and

If the DRG is one of a paired set of DRGs based on the presence or absence of a comorbidity or complication, both paired DRGs are included if either one meets the first three criteria above.

As we indicated in the FY 2006 IPPS proposed rule, as a result of our analysis, we considered two options for revising the current criteria. Option 1 was to include all DRGs within the postacute care transfer policy. This option has the advantage of providing consistent treatment of all DRGs. However, as we discussed in the proposed rule and above in this final rule, our analysis tends to indicate that, at a minimum, it may be appropriate to maintain the requirement that a DRG must have a geometric mean length of stay of at least 3.0 days because, under the transfer payment methodology, hospitals receive the entire payment for these DRGs in the first 2 days of the stay. Therefore, lowering the limit below 3.0 days would have little or no effect on payment for DRGs with geometric mean lengths of stay in this range.

The second option that we considered in the FY 2006 IPPS proposed rule was to expand the application of the postacute care transfer policy by applying the policy to any DRG that meets the following criteria:

The DRG has at least 2,000 postacute care transfer cases;

At least 20 percent of the cases in the DRG are discharged to postacute care;

Out of the cases discharged to postacute care, at least 10 percent occur before the geometric mean length of stay for the DRG;.

The DRG has a geometric mean length of stay of at least 3.0 days;

If the DRG is one of a paired set of DRGs based on the presence or absence of a comorbidity or complication, both paired DRGs are included if either one meets the first three criteria above.

As explained above, option 2 would expand the application of the postacute care transfer policy to 231 DRGs (rather than 223 DRGs as stated in the proposed rule) that have both a relatively high volume and a relatively high proportion of postacute care utilization. We proposed this change to avoid applying the postacute care transfer policy to DRGs with only a small number or proportion of cases transferred to postacute care. We believe that the analysis that we conducted suggests that substantial revisions to the criteria for including a DRG within the postacute care transfer policy are warranted. Therefore, in the FY 2006 IPPS proposed rule, we formally proposed Option 2 as presented above. However, we invited comments on both of the options and on the analysis that we had presented.

Comment: Many commenters expressed opposition to the postacute care transfer policy in general. Some of these commenters argued that the policy is contrary to the premise of the DRG system, which is to pay the average of all cases in a DRG, regardless of cost and length of stay. Some commenters also contended that the transfer policy is based on a false assumption of gaming by providers, and that it punishes providers for providing the appropriate level of care in the most appropriate setting. Other commenters argued that the rationale for the policy no longer exists because most of the providers of postacute care services in question have transitioned from cost-based reimbursement to PPSs themselves (SNFs as of October 1, 1998; HHAs as of October 1, 2000; IRFs as of January 1, 2002; LTCHs as of October 1, 2002; and IPFs as of January 1, 2005). Further, commenters noted that each of these postacute care payment systems have admission criteria to ensure that patients are not discharged prematurely to a lower level of care.

Other commenters contended that the policy creates a geographic bias against regions that have access to greater capital resources and postacute care facilities and that traditionally have had shorter lengths of stay for their patients than other regions of the country. Some commenters argued that the provision creates a perverse incentive for hospitals to keep their patients longer and to deny them the appropriate care in postacute care facilities when it is needed. Commenters continued to argue that this policy undermines the incentive for hospitals to reduce lengths of stay. Several commenters pointed to the tremendous administrative burden placed on hospitals with the expansion of the policy, particularly with regard to transfers to HHAs. Other commenters noted the administrative burdens of time, resource utilization, and delay of payments already associated with these types of transfers and subsequent claims corrections and reprocessing with the existing 30 DRGs.

Response: We do not agree that the postacute care transfer policy is inappropriate or contrary to the principles of prospective payment. The policy is fully consistent with the principles of prospective payment because the operative averaging principle in such systems assumes that the full extent of care is consistently provided in an acute care hospital. The averaging principle would be undermined if the system did not provide for adjustments in cases where a large proportion of the patient's care is provided by another entity. Thus, the statute appropriately provides for treating discharges to postacute care from certain DRGs as transfer cases. The statute also gives the Secretary the discretion to select appropriate DRGs to which this policy should be applied on the basis of a high volume of discharges to postacute care and a disproportionate use of post discharge services. Although it is true that many postacute settings to which the policy applies are now subject to a prospective payment methodology, this fact in no way undermines the appropriateness of a postacute transfer policy. Rather, such a policy serves to ensure that Medicare does not make full payments under two different payment systems when a patient's full course of treatment is divided between two facilities. It is just as inappropriate for Medicare to pay for the treatment of patients in these cases, at the full DRG amount at the IPPS hospital and under either a per discharge or per diem prospective payment in the postacute care setting as it is to pay the full DRG payment twice

[[Page 47416]]

when a patient is transferred from one acute care hospital to another. Therefore, because the majority of short-stay transfer cases receive the majority of their care at postacute care facilities (except for those DRGs that we have identified as having high costs on the first day and that are paid under a special payment methodology), we continue to believe that full payment to those facilities and reduced payment to acute facilities for these cases are merited. Numerous studies of the postacute care transfer policy by MedPAC, the Office of Inspector General, and other health-related entities continue to support the need for the policy, and some studies have supported expansion of the policy to additional DRGs where appropriate.

Comment: Most commenters objected to the proposed alternate criteria for DRGs to be included in the postacute care transfer policy. Some commenters objected to our proposing changes in the qualifying criteria for the postacute care transfer policy for the third consecutive year. These commenters argued that such frequent changes in policy gives the appearance of a contrived policy to suit CMS' desires and makes the regulatory process unpredictable and unfair. Many commenters asserted that there was little analytical support for changing the criteria, and in particular that CMS had presented little analytical support for the proposed thresholds of 2,000 and 20 percent of cases transferred to postacute care. Some commenters also contended that the proposed criteria appeared contrived to ensure that the proposal would meet specific budgetary goals. Many commenters expressed dismay that CMS would lower the limit so drastically from 14,000 postacute care transfer cases to 2,000, a ``dramatic drop of 86 percent.'' Many commenters also believed that the proposed alternate criteria did not meet the standards established in the statute. Specifically, these commenters indicated that the proposed threshold of 2,000 transfer cases does not constitute a ``high volume of discharges'' under the statute. Similarly, many commenters stated that a threshold of 20 percent of postacute transfer cases does not constitute a ``disproportionate use of post-discharge services.'' These commenters argued that, by definition, disproportionate use of postacute care should be well above the norm. One commenter added that it ``is a statistical impossibility for half of the universe of DRGs to have `disproportionate use of post-discharge services.' '' Some commenters suggested that CMS consider using alternatives to the newly proposed criteria. One commenter proposed that CMS establish thresholds at least one standard deviation above the average to determine when DRGs meet a disproportionate use of postacute care. Another commenter noted that thresholds of one standard deviation are employed elsewhere in Medicare policy. One commenter noted that, under the original implementation of the policy, the 10 DRGs that were included had a postacute care utilization rate of at least 45.3 percent (not including pairs) and when the policy was expanded to 30 DRGs, the lowest percentage of postacute care utilization (not including pairs) was 34.86. Therefore, this commenter contended that a reasonable figure that might represent a disproportionate use of postacute care utilization would be no less than 34.0 percent.

Response: We do not agree that the proposed thresholds were inappropriate or without analytical support. In particular, we do not agree that the threshold of 2,000 discharges to postacute care falls short of the statutory standard that DRGs included within the policy must have a ``high volume of discharges.'' In analyzing the total number of discharges to postacute care in each DRG, we found that the median DRG had approximately 1,600 discharges to postacute care. Thus, our proposed criteria of 2,000 discharges to postacute care is well above the median DRG's number of discharges to postacute care and can easily be argued to meet the statutory criteria of a ``high volume of discharges.'' Nevertheless, in response to the many comments on the proposed new thresholds, we have reexamined the data concerning the volume and the proportions of discharges to postacute care across DRGs. Our goal was to select thresholds that are appropriate to the purposes of the postacute care transfer policy and that clearly meet the statutory standards cited by the commenters.

We began by considering the suggestion of several commenters that it might be appropriate to establish thresholds at levels of one standard deviation above the average to determine high volume and disproportionate use of postacute care services. As one commenter pointed out, we have used such a standard for similar purposes in other areas of the Medicare program. However, our examination of the DRG data indicated that the average, or mean, is not the most appropriate measure of central tendency in these cases. The distributions of discharge volume and postacute care usage across DRGs are positively skewed. As a result, a relatively small number of DRGs with very high volume and rates of postacute care utilization have a disproportionate impact on the average or mean. Therefore, a better measure of central tendency in these cases is the median, or 50th percentile in the rankings of discharges and rates of postacute care utilization from highest to lowest. However, employing the median rather than the mean makes it impossible to employ the standard deviation in setting an appropriate threshold. In lieu of using the mean and standard deviation as suggested by the commenter, it is possible to select a percentile ranking in each array as an appropriate measure of ``high volume'' and ``disproportionate use.'' By definition, any volume of discharges above the 50th percentile can be considered a high volume in the context of the ranking from highest to lowest. Similarly, any rate of postacute care utilization above the 50th percentile can also be considered disproportionate use of such services. However, we agree with those commenters who recommended thresholds based on standard deviations above the mean, that it is appropriate to establish levels somewhat above the measures of central tendency as thresholds for high volume and disproportionate use. Therefore, we have determined that the 55th percentile is an appropriate level at which to establish these thresholds.

In the course of examining the relevant data, we also considered several alternatives to the ratio of postacute care discharges to total discharges as the most appropriate measure of the rate of postacute care utilization across DRGs. We came to the conclusion that a more appropriate measure of postacute care utilization is the proportion of discharges to postacute care that occur prior to the geometric mean length of stay for a DRG. We believe that the proportion of such short- stay discharges is a more appropriate measure in this context than the overall proportion of discharges to postacute care because only these discharges are affected by the postacute care transfer policy. Specifically, under the formula employed to determine payments for transfer cases, discharges that occur at or after the mean length of stay receive payments that equal the full DRG payment. Furthermore, we believe a focus on short-stay discharges to postacute care is more consistent with the statutory criteria of ``disproportionate use of post-discharge services.'' These short-stay cases are atypical in that they are discharged

[[Page 47417]]

before the geometric mean length of stay and result in the majority of care being provided at postacute care facilities.

Therefore, we examined the percentile rankings of DRGs inVersion 23.0 of the DRG Definitions Manual (FY 2006) in relation to the volume of discharges to postacute care, and the ratio of short-stay discharges to postacute care. We employed the March 2005 update of FY 2004 MedPAR data, the most recent data available to us. We determined that the median number of discharges to postacute care across all DRGs was 1,619, and the 55th percentile was 2,050. The median proportion of short-stay discharges to postacute care was 4.8 percent, and the 55th percentile was 5.5 percent. Therefore, in place of the first two criteria that we proposed in the FY 2006 IPPS proposed rule, we are establishing the following two criteria in this final rule, effective October 1, 2005:

The DRG has at least 2,050 postacute care transfer cases;

At least 5.5 percent of the cases in the DRG are discharged to postacute care prior to the geometric mean length of stay for the DRG.

In response to the comments suggesting that we provided little data or analytic support for our proposal, we provided detailed analysis of our findings on these issues in both the FY 2006 IPPS proposed rule and this final rule. The data underlying our analysis are publicly available through the CMS Web site at: http://www.cms.hhs.gov/data/order/default.asp .

Comment: Many commenters also objected to our proposal to eliminate the requirement that a DRG experience a decline in length of stay. These commenters contended that there was no evidence provided that hospitals are changing their behavior, transferring patients earlier, or taking advantage of the payment system. Another commenter argued that removal of the requirement that DRGs experience a decline in length of stay was contrary to the intent of the statute. This commenter argued that the objective of the policy was ``to adjust inpatient PPS payments to account for reduced hospital lengths of stay due to a discharge to another setting.'' Therefore, the commenter argued, if the MedPAR data demonstrates that postacute care utilization for a DRG does not contribute to a significant decrease in the geometric mean length of stay, the DRG should not be subject to the policy. In general, commenters recommend a different approach to further expansions of the postacute care transfer policy that they assert would more accurately reflect the costs of patient care provided in acute care hospitals.

Response: The statute does not establish any requirement that we consider declining length of stay as a standard in selecting appropriate DRGs for inclusion under the postacute care transfer policy. We originally adopted such a standard because we found a relationship between declining lengths of stay and increasing use of postacute care services. As we discussed in the proposed rule, and again above, our more recent analysis has called into question the basis for the requirement that a DRG experience a decline in the geometric mean length of stay over the most recent 5-year period. Our finding that some DRGs with increases in postacute care utilization during the past several years have also experienced increases in geometric mean length of stay indicates that this criterion is no longer effective to identify those DRGs that should be subject to the postacute care transfer policy. Therefore, we are finalizing our proposal to discontinue the current criterion for inclusion in the policy that requires a DRG to experience a decline of at least 7 percent over the last 5 years in the geometric mean length of stay.

Comment: Some commenters objected to our current criterion that 10 percent of the postacute care transfer cases within a DRG must be short-stay cases in order for the DRG to be included in the policy. Some of these commenters argued that this would effectively mean that up to 90 percent of all discharges within a DRG are not short-stay discharges, and therefore, these DRGs would not meet the disproportionate use requirement as provided in the statute.

Response: We do not agree with the commenters that inclusion of this criterion in the policy was inappropriate. To the contrary, for the reasons we have discussed above and in previous rules, we believe that some consideration of the proportion of short-stay discharges to postacute care--the discharges actually affected by the application of the policy--is an appropriate component of the criteria employed to determine the scope of the policy. However, we have decided not to retain that specific criterion under the revised policy that we are adopting in this final rule. This criterion is unnecessary because we decided to adopt the criterion that at least 5.5 percent of cases in the DRG must be discharged to postacute care prior to the geometric mean length of stay for the DRG. By including this criterion as a measure of disproportionate use of postacute care services, we believe that it becomes redundant to retain another measure that uses short- stay transfer cases.

Comment: Many commenters also did not support the criterion of including paired DRGs in the policy, citing that most hospitals have switched to a coding system that interfaces with the coder electronically, thereby reducing the probability that a coder would remove a CC code in order to change the payment for a case that was transferred to postacute care. Further, some commenters noted that it might be inappropriate to include paired DRGs in the special payment methodology, as the transfer payment for the first day for many of the CC DRGs in the CC/non-CC pair is typically higher than the full DRG payment for the non-CC pair. As a result, these commenters contended that coders would not have any incentive to exclude a CC from the hospital's bill. Therefore, these commenters suggested that CMS consider adopting a policy that excludes ``the non-CC of a paired DRG when the transfer weight of the CC DRG would be greater than the full DRG payment of the non-CC DRG.'' They noted that, by following this recommendation, the policy would agree with CMS' rationale for the inclusion of paired DRGs and also exclude those DRGs that do not meet the qualifying criteria.

Response: It has been our practice to include paired DRGs since the inception of the policy in 1998. This practice is in compliance with Sec. 412.4(d)(1)(iv) of the regulations. While it is possible that technical advances have resulted in electronic systems and more automated coding, the selection of codes to include on the bill remain within the responsibility and authority of the hospital and its staff. Thus, we believe the coder will have the ability to select whether to include or exclude a CC secondary diagnosis code on the hospital's bill when a patient is transferred to postacute care. We include both DRGs from a paired-DRG combination because if we were to include only the more complex DRG (that is, the ``with CC'' DRG from a ``with/without CC'' DRG combination) in the transfer policy, there might be an incentive for hospitals not to include any code that would identify a complicating or cormorbid condition. In our analysis of the included pairs in our data, we have not found support for the commenter's assertion that, in some instances, the transfer adjusted payment for a ``CC'' DRG is greater than the full payment for the non-transfer adjusted ``without CC'' DRG. In cases where a ``CC:'' DRG is transferred after a one day length of stay, the estimated transfer adjusted payments for the ``CC'' DRGs

[[Page 47418]]

are less than the full payments for the ``without CC'' DRGs. As this could introduce improper coding incentives, we continue to believe our approach of identifying either DRG from a paired-DRG combination individually for inclusion in the policy is appropriate.

Comment: Some commenters argued that including a transfer-adjusted case weight in the DRG relative weight calculation has the effect of maintaining the DRG weight at an artificially high level. Other commenters indicated that, in the absence of this adjustment, the lower costs of short-stay postacute care transfers will be reflected in lower DRG case weights, making a postacute care transfer payment policy unnecessary. Another commenter stated that the cost savings realized through shorter lengths of stay, including those from transfer cases, have already been considered and accounted for by Congress each year when it sets the market basket update.

Response: We agree with the commenters that a high proportion of short-stay to total cases in DRGs that are not subject to the postacute care transfer policy will likely result in lower weights for these DRGs. However, we believe these commenters actually support our argument for expanding the postacute care transfer policy to more DRGs where there is disproportionate use of postacute care services. While including all cases in the relative weight calculation without any adjustment would likely result in a lower DRG weight and payment for a short-stay transfer case, it would also result in lower payments for all of the remaining cases in the DRG where the hospital used more resources to care for the patient. To the extent that there is disproportionate use of postacute care services, hospitals would be disadvantaged in the relative weight calculation and their payments when the patient is not discharged early if we were to make no adjustment for a transfer case when setting the DRG relative weight. By reducing the impact that short-stay cases have on the DRG relative weight, we believe our payment will more accurately reflect all of the resources provided by a hospital during a typical stay. Thus, the payment will better reflect all of the costs a hospital expends for the stay when a full course of treatment is provided and our postacute care transfer policy will appropriately provide less payment for a transfer case in recognition of the lower cost of an abbreviated hospital stay.

Comment: Some commenters objected to the method by which CMS proposed the change in the criteria for DRGs to qualify to be included in the postacute care transfer policy. They argued that CMS should have proposed the criteria, accepted comment on the alternate criteria, and made appropriate changes based on those comments before applying them to any additional DRGs. Instead, commenters contended that CMS seemingly arbitrarily created the alternate set of criteria and applied them to new DRGs in the same rule.

Response: We are making the change to our postacute care transfer policy through a notice and comment rulemaking procedure before applying the new policy to any DRGs. The implication in these comments that we have already expanded the policy to additional DRGs is incorrect. We will be applying the revised postacute care transfer policy for discharges occurring on or after October 1, 2006, after having provided notice of our proposal to revise the policy in the proposed rule; allowing for a 60-day public comment period; and making changes to the policy in response to public comment.

Comment: Some commenters objected to the implication that early discharges to postacute care are done for economic reasons instead of patient need. Other commenters believed hospitals may keep patients in the hospital longer to avoid the reduced IPPS payment. These commenters indicated that the policy would increase, not reduce, Medicare spending to treat the same patients. Other commenters encouraged CMS to complete its analysis of the MedPAC recommendation to adopt severity DRGs before expanding the postacute care transfer policy. These commenters argued that CMS should apply the postacute care transfer policy to DRGs consistent with the goal of ``aligning patient severity with payment.'' These commenters argued that, if severity DRGs were implemented, there would be no need for a postacute care transfer policy because the system would recognize higher payment for more resource intensive patients.

Response: Our proposal to expand the postacute care transfer policy was not intended to imply that hospitals will prematurely discharge patients early to postacute care for financial reasons. Rather, our policy recognizes that hospitals expend fewer resources for patients who are discharged prior to the geometric mean length of stay and Medicare's payment should be less. We do note that some of the commenters themselves imply that hospitals will react to the financial incentives of the revised postacute care transfer policy by keeping patients in the hospital longer to avoid payment reductions that will occur if patients are discharged early to postacute care. If true, it is hard to understand what the hospitals would accomplish because even though they would receive the full DRG payment, they would also have costs associated with retaining patients who would be more appropriately discharged to another setting in the hospital.

It is not clear to us why an analysis of MedPAC's recommendation that we adopt severity DRGs is relevant to the postacute care transfer policy. To our knowledge, such a change to the DRG system would be intended to result in better recognition of severity levels in making DRG assignments, but would not involve any direct consideration of whether a hospital provides the full course of treatment to a patient. We are unaware that a severity DRG system, such as the APR-DRGs, would use length of stay and early discharge to postacute care as a basis for making a DRG assignment. Nevertheless, we will consider this issue as we study the MedPAC recommendation.

Comment: Commenters argued that we should not further expand the postacute care transfer payment policy until a full analysis of last year's changes to the policy is completed. According to the commenters, we should analyze whether the postacute care transfer policy has led to unnecessarily extended hospital stays in order to avoid the adjustment and affected quality of care. Commenters also noted that studies show that the majority of patients who use postacute care have longer (7.51 days), not shorter (4.93 days), hospital stays. These commenters argued that CMS should focus its efforts on improving quality of care, not on further expanding the postacute care transfer provision.

Response: In the FY 2005 IPPS final rule (69 FR 49073), we established a policy for how to apply the criteria for the postacute care transfer policy to cases that were previously assigned to a DRG that has split, when the split DRG qualified for inclusion in the postacute care transfer policy. This policy was a rather limited change to our postacute care transfer policy that has little bearing on the changes that we are making for FY 2006. Thus, we do not believe further analysis of this change is necessary before undertaking the changes we are adopting in this final rule.

We believe the point made by the commenter provides further grounds to expand the postacute care transfer policy. The policy only applies to patients that are discharged from the hospital at least one day before the geometric mean length of stay. The policy does not apply to the longer stay patients that are, according to the

[[Page 47419]]

commenter, more resource intensive. Thus, we make a reduced payment only for those short-stay patients transferred to postacute care that are, following the logic of the comment, less costly to the hospital.

Comment: Some commenters argued that studies have shown that many rural areas now have the same types of postacute care facilities as urban hospitals and expanding the postacute care transfer policy will harm rural areas by reducing payments to rural hospitals. Many commenters suggested that, if CMS is determined to make an expansion to the policy without providing analysis supporting the changes, any changes should be made in a budget neutral manner. Other commenters suggested that we should implement the policy expansion over 3 years to lessen the financial impact in the first year. Commenters also disputed our savings estimates indicating that once the effects of IME, disproportionate share, capital and outlier payments are taken into consideration, the total annual reduction would be closer to $894 million. They argued that hospitals can ill-afford this kind of reduction in payments at a time when they are already experiencing nursing shortages, incurring losses for treating Medicare beneficiaries, and expecting tremendous increases in costs associated with the aging baby boom generation.

Commenters also indicated that the policy should not apply in situations where a patient is living in a SNF. In these cases, the commenters argued that an early discharge of the patient to a SNF is really a discharge to the patient's home and the policy should not be applied.

Response: We do not believe that the law permits us to distinguish between urban and rural areas when applying this policy. Furthermore, we do not believe there is a policy basis for such a distinction because the principle of making lower payments to the acute care hospital based on the majority of care being provided in a postacute care setting would apply equally to urban and rural hospitals. The law does not require or authorize us to make these changes over a transitional period or in a budget neutral manner as suggested by the commenters. For this reason, we are implementing the policy as we have described. We note that our savings estimates have been updated to reflect the policies we are adopting in this final rule. With respect to a discharge to a SNF, we note that section 1886(d)(5)(J)(ii)(II) of the Act makes clear that the postacute care transfer policy must apply in this situation.

The impact section in Appendix A of this final rule discusses our findings on the effects of adopting our final rule policy. The DRG relative weights in Tables 5 and 7 of the Addendum to this final rule also include the effect of changing the postacute care transfer policy. We note that we will follow procedures similar to those that are currently followed for treating cases identified as transfers in the DRG recalibration process. That is, as described in the discussion of DRG recalibration in section II.C. of the preamble to this final rule, additional transfer cases will be counted as a fraction of a case based on the ratio of a hospital's transfer payment under the per diem payment methodology to the full DRG payment for nontransfer cases.

In summary, after consideration of the comments received, in this final rule, we have revised the criteria that we proposed for determining which DRGs qualify for postacute care transfer payments. The final policy, which we are incorporating into the regulations at Sec. 412.4, specifies that, effective October 1, 2005, we are making a DRG subject to the postacute care transfer policy if, based on the Version 23.0 GROUPER (FY 2006), using data from FY 2004, the DRG meets the following criteria:

The DRG must have a geometric mean length of stay of at least 3 days;

The DRG must have at least 2,050 postacute care transfer cases;

At least 5.5 percent of the cases in the DRG are discharged to postacute care prior to the geometric mean length of stay for the DRG; and

If the DRG is one of a paired set of DRGs based on the presence or absence of a comorbidity or complication, both paired DRGs are included if either one meets the three criteria above.

If a DRG meets the above criteria based on the Version 23.0 GROUPER and FY 2004 MedPAR data, we are making the DRG subject to the postacute care transfer policy. We will not revise the list of DRGs subject to the postacute care transfer policy annually unless we are making a change to a specific DRG. Using the version of the Medicare GROUPER for the year when a new or revised DRG first becomes effective, we will make the DRG subject to the postacute care transfer policy if its total number of discharges and proportion of short-stay discharges to postacute care exceed the 55th percentile for all DRGs. We are establishing this policy to promote certainty and stability in the postacute care transfer payment policy. Annual reviews of the list of DRGs subject to the policy would likely lead to great volatility in the payment methodology with certain DRGs qualifying for the policy in one year, deleted the next year, only to be readded the following year. However, over time, as treatment practices change it is possible that some DRGs that currently qualify for the policy will no longer exhibit a disproportionate use of postacute care. Similarly, there may be other DRGs that currently have a low rate of discharges to postacute care, but which will have very high rates in the future. For these reasons, we expect to periodically review the criteria that are used to make a DRG subject to the postacute transfer policy. At this time, we have not decided on how frequently to perform this review but are considering undertaking this analysis every 5 years. We welcome public comments on this issue.

Section 1886(d)(5)(J)(i) of the Act recognizes that, in some cases, a substantial portion of the cost of care is incurred in the early days of the inpatient stay. Similar to the policy for transfers between two acute care hospitals, transferring hospitals receive twice the per diem rate for the first day of treatment and the per diem rate for each following day of the stay before the transfer, up to the full DRG payment, for cases discharged to postacute care. However, in the past, three of the DRGs subject to the postacute care transfer policy have exhibited an even higher share of costs very early in the hospital stay in postacute care transfer situations. For these DRGs, hospitals receive 50 percent of the full DRG payment plus the single per diem (rather than double the per diem) for the first day of the stay and 50 percent of the per diem for the remaining days of the stay, up to the full DRG payment.

Comment: Commenters indicated there was not a clear explanation for when a DRG would be subject to the special payment methodology. For example, commenters indicated that DRGs 107 (Coronary Bypass with PTCA), 108 (Coronary Bypass with Cardiac Catheterization) and 109 (Coronary Bypass without PTCA or Cardiac Catheterization) are all related, but only DRG 109 is paid using the special payment methodology. The commenters argued that resource utilization for all three of these surgical DRGs would be similar, and therefore, all three DRGs should be paid using the special payment methodology.

Response: To identify DRGs that are subject to the special payment methodology, we compare the average charges for all cases with a length of stay of 1 day to the average charges of all cases in a particular DRG. To qualify for the alternative methodology, the average charges of the 1-day discharge

[[Page 47420]]

cases must be at least 50 percent of the average charges for all cases in the DRG. We only apply this methodology to those DRGs that have a mean length of stay that is greater than 4 days because cases with a shorter average length of stay will receive the full DRG payment for the case on the second day of the stay regardless of the payment methodology used. In addition, if a DRG in a paired set of DRGs based on the presence or absence of a comorbidity or complication meets the criteria for being included in the postacute care transfer policy and qualifies for the special payment methodology, we include both DRGs in the special payment methodology in order to eliminate any incentive to code incorrectly to receive a higher payment for a case. We have identified those additional DRGs that are subject to the special payment methodology in Table 5 of the Addendum to this final rule.

B. Reporting of Hospital Quality Data for Annual Hospital Payment Update (Sec. 412.64(d)(2))

1. Background

Section 1886(b)(3)(B)(vii) of the Act, as added by section 501(b) of Pub. L. 108-173 revised the mechanism used to update the standardized amount of payment for inpatient hospital operating costs. Specifically, the statute provides for a reduction of 0.4 percentage points to the update percentage increase (also known as the market basket update) for each of FYs 2005 through 2007 for any ``subsection (d) hospital'' that does not submit data on a set of 10 quality indicators established by the Secretary as of November 1, 2003. The statute also provides that any reduction will apply only to the fiscal year involved, and will not be taken into account in computing the applicable percentage increase for a subsequent fiscal year. This measure establishes an incentive for IPPS hospitals to submit data on the quality measures established by the Secretary.

We initially implemented section 1886(b)(3)(B)(vii) of the Act in the FY 2005 IPPS final rule (August 11, 2004, 69 FR 49078) in continuity with the Department's Hospital Quality Initiative as described at the CMS Web site: http://www.cms.hhs.gov/quality/hospitals. At a

press conference on December 12, 2002, the Secretary of the Department of Health and Human Services (HHS) announced a series of steps that HHS and its collaborators were taking to promote public reporting of hospital quality information. These collaborators include the American Hospital Association, the Federation of American Hospitals, the Association of American Medical Colleges, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), the National Quality Forum, the American Medical Association, the Consumer-Purchaser Disclosure Project, the American Association of Retired Persons, the American Federation of Labor-Congress of Industrial Organizations, the Agency for Healthcare Research and Quality, as well as CMS, Quality Improvement Organizations (QIOs), and others.

In July 2003, CMS began the National Voluntary Hospital Reporting Initiative (NVHRI), now known as the Hospital Quality Alliance (HQA): Improving Care through Information. Data from this initiative have been used to populate a professional Web site providing data to healthcare professionals. The Hospital Compare Web site has also been developed to provide information appropriate for Medicare beneficiaries. The consumer Web site is intended to be an important tool for individuals to use in making decisions about health care options. The information in this Web site assists beneficiaries by providing comparison information for consumers who need to select a hospital. It also serves as a way to encourage accountability of hospitals for the care they provide to patients.

The 10 measures that are employed in this voluntary initiative as of November 1, 2003, are:

Heart Attack (Acute Myocardial Infarction)

Was aspirin given to the patient upon arrival to the hospital?

Was aspirin prescribed when the patient was discharged?

Was a beta-blocker given to the patient upon arrival to the hospital?

Was a beta-blocker prescribed when the patient was discharged?

Was an ACE inhibitor given for the patient with heart failure?

Heart failure

Did the patient get an assessment of his or her heart function?

Was an ACE inhibitor given to the patient?

Pneumonia

Was an antibiotic given to the patient in a timely way?

Had a patient received a pneumococcal vaccination?

Was the patient's oxygen level assessed?

These measures have been endorsed by the National Quality Forum (NQF) and are a subset of the same measures currently collected for the JCAHO by its accredited hospitals. The Secretary chose these 10 quality measures in order to collect data to: (1) Provide useful and valid information about hospital quality to the public; (2) provide hospitals with a sense of predictability about public reporting expectations; (3) begin to standardize data and data collection mechanisms; and (4) foster hospital quality improvement. Many hospitals have participated in the National Voluntary Hospital Reporting Initiative (NVHRI), and are continuing to submit data to the QIO Clinical Warehouse for that purpose.

Over the next several years, hospitals are encouraged to take steps toward the adoption of electronic medical records (EMRs) that will allow for reporting of clinical quality data from the electronic record directly to a CMS data repository. CMS intends to begin working toward creating measures specifications and a system or mechanism, or both, that will accept the data directly without requiring the transfer of the raw data into an XML file as is currently done. The Department is presently working cooperatively with other Federal agencies in the development of Federal health architecture data standards. CMS encourages hospitals that are developing systems to conform them to both industry standards and, when developed, the Federal health architecture data standards, and to ensure that the data necessary for quality measures are captured. Ideally, such systems will also provide point-of-care decision support that enables high levels of performance on the measures. Hospitals using EMRs to produce data on quality measures will be held to the same performance expectations as hospitals not using EMRs. In the FY 2006 IPPS proposed rule, we indicated that we were exploring requirements and other options to encourage the submission of electronically produced data, and invited comments on such requirements and options.

Comment: One commenter expressed support for the creation of a system to move information from electronic health records to a CMS data repository.

Response: We agree, and this is one of the reasons why we proposed the question in the preamble of the Notice. We appreciate this commenter's support and will strive to minimize data burdens while improving hospital quality by moving to an industry-accepted system of electronic health records.

Comment: One commenter supported the use of a single national database of quality measures that could be used by all stakeholders. However, this commenter believed that the business case for the investment in electronic medical records is not clear.

Response: CMS strives to minimize data reporting burdens, while working with providers to improve hospital

[[Page 47421]]

quality. We will study and assess cost, burden, and benefits of moving to an industry-accepted system of electronic health records.

Comment: Several commenters suggested that CMS should provide financial support and appropriate technical assistance to hospitals prior to, or in conjunction with any requirements for hospitals to implement electronic medical records and submit the data directly to the CMS data warehouse. The commenter added that, eventually, using electronic medical records to submit data would add additional burdens to the hospital, such as cost and the need for additional resources.

Response: We do not currently have the authority to pay for electronic data submission. However, we do appreciate the challenges and work that lie ahead to achieve the vision of using electronic medical records to submit data. We will keep these issues in mind as we move forward pursuing electronic data submission.

This method of collecting data, if well designed, should expedite the submission of quality data. We found in the Surgical Care Improvement Project (SCIP) that hospitals with electronic records were able to abstract SCIP data in as little as 10 minutes. It may require designing a report specifically for the Medicare measures, but after that work is complete, we would expect no increase in resources in hospitals with electronic records. At worst, hospitals with EMRs should only have the additional expense of printing the patient record. After that is done, the abstraction cost would be no greater burden on the hospital.

2. Requirements for Hospital Reporting of Quality Data

The procedures for participating in the Reporting Hospital Quality Data for the Annual Payment Update (RHQDAPU) program created in accordance with section 501(b) of Pub. L. 108-173 can be found on the QualityNet Exchange at the Web site: http://www.qnetexchange.org in the

``Reporting Hospital Quality Data for Annual Payment Update Reference Checklist'' section of the Web site. This checklist also contains all of the forms to be completed by hospitals participating in the program. In order to participate in the hospital reporting initiative, hospitals must follow these steps:

The hospital must identify a QualityNet Exchange Administrator who follows the registration process and submits the information through the QIO. This must be done regardless of whether the hospital uses a vendor for transmission of data.

All participants must first register with the QualityNet Exchange, regardless of the method used for data submission. If a hospital participated in the voluntary reporting initiative, re- registration on QualityNet Exchange is unnecessary. However, the hospital must complete the ``Reporting Hospital Quality Data for Annual Payment Update Notice of Participation'' form. All hospitals must send this form to their QIO.

The hospital must collect data for all 10 measures and submit the data to the QIO Clinical Warehouse either using the CMS Abstraction & Reporting Tool (CART), the JCAHO Oryx Core Measures Performance Measurement System (PMS), or another third-party vendor tool that has met the measurement specification requirements for data transmission to QualityNet Exchange. The QIO Clinical Warehouse will submit the data to CMS on behalf of the hospitals. The submission will be done through QualityNet Exchange, which is a secure site that voluntarily meets or exceeds all current Health Insurance Portability and Accountability Act (HIPAA) requirements, while maintaining QIO confidentiality as required under the relevant regulations and statutes. The information in the Clinical Warehouse is considered QIO information and, therefore, is subject to the stringent QIO confidentiality regulations in 42 CFR Part 480.

For the first year of the program, FY 2005, hospitals were required to begin the submission of data by July 1, 2004, under the provisions of section 1886(b)(3)(B)(vii)(II) of the Act, as added by section 501(b) of Pub. L. 108-173. Because section 501(b) of Pub. L. 108-173 granted a 30-day grace period for submission of data for purposes of the FY 2005 update, hospitals were given until August 1, 2004, to begin submissions into the QIO Clinical Warehouse. Hospitals were required to submit data for the first calendar quarter of 2004. We received data from over 98 percent of the eligible hospitals.

We proposed in the FY 2006 IPPS proposed rule, and are adopting as final policy in this rule, that, for FY 2006, hospitals must continuously submit the required 10 measures each quarter according to the schedule found on the Web site at http://www.qnetexchange.org. New

facilities must submit the data using the same schedule, as dictated by the quarter they begin discharging patients. We expect that all hospitals will have submitted data to the QIO Clinical Warehouse for discharges through the fourth quarter of calendar year 2004 (October to December 2004). Hospitals had 4\1/2\ months from the end of the fourth quarter until the closing of the warehouse (from December 31, 2004, until May 15, 2005) to make sure there were no errors in the submitted data. The warehouse was closed at that time in order to draw the validation sample and to begin preparing the public file for the Hospital Compare public reporting Web site. Data from fourth quarter 2004 discharges (October through December 2004) are the last quarter of data with a submission deadline (May 15, 2005) preceding our deadline for certifying the hospitals' eligible to receive the full update for FY 2006. As we required for FY 2005, the data for each quarter must be submitted on time and pass all of the edits and consistency checks required in the clinical warehouse. Hospitals that do not treat a condition or have very few discharges will not be penalized, and will receive the full annual payment update if they submit all the data on the 10 measures.

New hospitals should begin collecting and reporting data immediately and complete the registration requirements for the RHQDAPU. New hospitals will be held to the same standard as established facilities when determining the expected number of discharges for the calendar quarters covered for each fiscal year. The full annual payment updates will be based on the successful submission of data to CMS via the QIO Clinical Warehouse by the established deadlines.

For FY 2005, hospitals could have withdrawn from RHQDAPU at any time up to August 1, 2004. Hospitals withdrawing from the program did not receive the full market basket update and, instead, received a reduction of 0.4 percentage points in their update. By law, a hospital's actions each year will not affect its update in a subsequent year. Therefore, a hospital must meet the requirements for RHQDAPU each year the program is in effect to qualify for the full update each year.

For the first year, FY 2005, there were no chart-audit validation criteria in place. Based upon our experience from the FY 2005 submissions, and upon our requirement for reliable and valid data, we proposed to place the following additional requirements on hospitals for the data for the FY 2006 payment update. We are finalizing the proposed additional requirements in this rule. These requirements, as well as additional information on validation requirements, are being placed on QualityNet Exchange.

The hospital must pass our validation requirement of a minimum of

[[Page 47422]]

80 percent reliability, based upon our chart-audit validation process, for the third quarter data of calendar year 2004. These data were due to the clinical warehouse by February 15, 2005. We use appropriate confidence intervals as explained in the proposed rule to determine if a hospital has achieved an 80-percent reliability. The use of confidence intervals allows us to establish an appropriate range below the 80-percent reliability threshold that demonstrates a sufficient level of validity to allow the data to still be considered valid. We estimate the percent reliability based upon a review of five charts, and then calculate the upper 95 percent confidence limit for that estimate. If this upper limit is above the required 80 percent reliability, the hospital data are considered validated. As proposed, we are using the design specific estimate of the variance for the confidence interval calculation, which, in this case, is a single stage cluster sample, with unequal cluster sizes. (For reference, see Cochran, William G, (1977) Sampling Techniques, John Wiley & Sons, New York, chapter 3, section 3.12.)

We use a two-step process to determine if a hospital is submitting valid data. In the first step, we calculate the percent agreement for all of the variables submitted in all of the charts. If a hospital falls below the 80 percent cutoff, we then restrict the comparison to those variables associated with the 10 measures required under section 501(b) of Pub. L. 108-173. We recalculate the percent agreement and the estimated 95 percent confidence interval and again compare to the 80 percent cutoff point. If a hospital passes under this restricted set of variables, the hospital is considered to be submitting valid data for purposes of the RHQDAPU.

Under the standard appeal process, all hospitals are given the detailed results of the Clinical Data Abstraction Center (CDAC) reabstraction along with their estimated percent reliability and the upper bound of the 95 percent confidence interval. If a hospital does not meet the required 80 percent threshold, the hospital has 10 days to appeal these results to their QIO. The QIO will review the appeal with the hospital and make a final determination on the appeal. If the QIO does not agree with the hospital's appeal, then the original results stand. The new results will be provided to the hospital through the usual processes, and the validation described previously will be repeated. This process is described in detail at the following Web site: http://www.qnetexchange.org. Hospitals that fail to receive the

required 80-percent reliability after the standard appeals process may ask that CMS accept the fourth quarter of calendar year 2004 validation results as a final attempt to present evidence of reliability. However, in order to process the fourth quarter data in time to meet our internal deadlines, these hospitals needed to submit the charts requested for reabstraction by no later than August 1, 2005, in order for us to guarantee consideration of this information. Hospitals that make the early submission of these data and pass the 80-percent reliability minimum level will satisfy this requirement. In reviewing the data for these hospitals, we plan to combine the 5 cases from the third quarter and the 5 cases from the fourth quarter into a single sample to determine whether the 80-percent reliability level is met. This gives us the greatest accuracy when estimating the reliability level. The confidence interval approach accounts for the variation in coding among the 5 charts pulled each quarter and for the entire year around the overall hospital mean score (on all individual data elements compared). The closer each case's reliability score is to the hospital mean score, the tighter the confidence interval established for that hospital. A hospital may code each chart equally inaccurately, achieve a tight confidence interval, and not pass even though its overall score is just below the passing threshold (75 percent, for example). A hospital with more variation among charts will achieve a broader confidence interval, which may allow it to pass even though some charts score very low and others score very high.

We believe we have adopted the most suitable statistical tests for the hospital data we are trying to validate. However, in the FY 2006 IPPS proposed rule, we invited public comments on this and any other approaches. We expressed particular interest in comments from hospitals on the initial starting points for the passing threshold, the confidence interval established, and the sampling approach. Because we will be receiving data each quarter from hospitals, our information on the sampling methodology will improve with each quarter's submissions. We have indicated that we will analyze this information to determine if any changes in our methodology are required. We will make any necessary revisions to the sampling methodology and the statistical approach through manual issuances and other guidance to hospitals.

Comment: Several commenters requested that we provide additional time for the hospitals to appeal their validation determinations. Many commenters stated that the current timeframe of 10 days is not sufficient time to decide to appeal the results. Commenters also asked CMS to specify if the 10-day time period is measured in calendar days or business days.

Response: The 10 days are business days. This timeframe is designed to provide sufficient time for hospitals to gather relevant information. Hospital will not need to produce more information in deciding whether or not to appeal the results of the abstraction. Hospitals are required only to submit their request for appeal form within the 10-day time period. We believe 10 business days is sufficient time for a hospital to decide whether or not it wants the contractor to review its original abstraction. However, it does expedite the final determination and minimizes data lag for public reporting and payment determination.

Comment: Four commenters requested that CMS allow more time for the hospital to produce the medical record and submit the record for the validation review.

Response: We believe the timeframe provides sufficient time for hospitals to gather the medical record and copy and forward it to the contractor. After the warehouse closes for quarterly submissions, a sample of five charts is selected for validation. The CDAC requests the charts from the hospital. Hospitals are provided 30 days, as stated in the Hospital Validation Flow Chart which can be found on QNet Exchange Web site. Upon completion of validation, the hospital receives a submission report that states whether the five charts meet the validation criteria. If the hospital fails validation, the hospital is provided 10 business days to notify the QIO that it wishes to appeal the validation decision. This timeframe helps expedite the final determination and minimizes data lag for public reporting and payment determination.

Comment: Four commenters requested that CMS delay hospital reporting until we have aligned our definitions and abstraction guidelines with JCAHO.

Response: The third quarter 2004 definitions and abstraction guidelines are better aligned to JCAHO than previous quarters, and with these third quarter 2004 definitions and abstraction guidelines, we believe we have made great strides in the long-term alignment process with JCAHO. Although CMS and JCAHO will not be fully aligned with third quarter and fourth quarter discharges, validation results for these periods are calculated from only those

[[Page 47423]]

aligned data elements. We anticipate full alignment with first quarter 2005 discharges.

Comment: Many of the comments requested requiring only submission of hospital reporting data in order to be eligible for the full annual payment update, and separating the process of validation from eligibility for the market basket update. Commenters frequently cited difficulties with the data infrastructure, specifically the communication of validation results to the hospital that was causing confusion for the hospitals. The commenters also cited technical difficulties with data submission to the warehouse.

Response: A production problem occurred while releasing the first set of third quarter 2004 validation results. A CMS contractor had forwarded individual validation results with the wrong data to a small number of hospitals. The run was discontinued immediately upon discovery. All hospitals involved were notified of the error and have verified the destruction of the files. In addition, hospitals also encountered abstraction and processing issues in this process. CMS and its contractors readily resolved these issues and there has not been a negative impact on hospitals or their patients. Ninety-eight percent of the hospitals that submitted data for the third quarter of 2004 that are eligible for the market basket update will receive the full update based on validation results. The production problem did not contribute to the 2 percent of the hospitals whose data did not validate. We believe that it is important for the data in the clinical warehouse on which full payment is determined to be reliable and valid.

Comment: Three commenters stated that five charts per hospital for validation is not a sufficient number to judge the quality of the care delivered in the hospital.

Response: CMS factored cost, burden, and precision of the validation results when deciding to implement the current validation sampling methodology. The goal of the chart audit validation process is to ensure that the hospital is abstracting and submitting accurate data. In order to calculate quality measures, which are used to determine the standard of care, we need to have complete and accurate data. Errors of omission and transcription errors contribute to the overall errors in calculating quality measures. We agree that it is important to differentiate between these errors in order to provide feedback to hospitals. The process we have in place to provide this feedback gives each hospital the detailed abstraction results from the CDAC reabstraction so that hospital staff may determine the types of errors and take appropriate action.

The five sampled charts usually yield 100 data elements that are used to determine the validation rate. This sample of data elements is sufficient to produce reliable validation rate estimates. Analysis of previous quarters' submitted data indicates that the clustering effect caused by the five chart sample boosts sampling variability by a relatively small proportion. Despite this increase in sampling variability, the sample still produces reliable validation rate estimates. The relative sampling variability is largely determined by the number of data elements abstracted, while incorporating the increased variability caused by the number of records. Analysis of previous quarters' submitted data indicates that the sampling variability is increased by a relatively small proportion.

Comment: Seven commenters requested that we use a test process for our data submission and our validation parameters.

Response: We agree that there should be a test process. In order to address this concern, we encourage hospitals to submit data continuously throughout the quarter; thereby data submission problems can be addressed and corrected early. Also, CMS, JCAHO, and the Hospital Reporting QIOSC conduct National calls once a month with vendors to provide further assistance. The calls give vendors the opportunity to ask questions and get timely feedback to make necessary changes to the data file prior to submission.

Hospitals have continued access to view and change their own data in the warehouse up to the time the warehouse is closed. The hospitals can pull the validation sample and begin preparing the file used for public reporting on the Hospital Compare Web site at http://www.HospitalCompare.hhs.gov. CMS encourages hospitals to test their

data submission processing during this time by submitting quality data into the QIO clinical warehouse before the deadline and reviewing their submission reports to ensure that all data were successfully submitted into the warehouse.

The validation parameters for the CART software are extensively tested through internal quality assurance and independent validation and verification. The CMS contractor uses an internal quality control process to verify that all applications and data processes produce the results outlined in the specifications. CMS provides further quality assurance in some areas using an Independent Validation & Verification (IV&V) process by another contractor. CMS will extend IV&V to all areas involving the annual payment update. In addition, a pre production check has been implemented and will be enhanced to review any output prior to production release. Finally, CMS has a QualityNet Help Desk that can assist providers with questions or concerns.

Comment: One commenter suggested that CMS resolve all of the vendor upload issues prior to increasing the reimbursement for pay-for- performance programs.

Response: The hospital-to-vendor relationship is external to CMS. Therefore, hospitals are responsible for selecting and ensuring that vendors submit valid data into the QIO clinical warehouse. CMS does not have contractual agreements with vendors. Communication with vendors is the hospitals' responsibility. CMS holds hospitals responsible for submitting accurate data. Therefore, hospitals that have a contractual agreement with a vendor must collaborate with the vendor to ensure the data file is submitted accurately. When the data are uploaded to the QIO Clinical Warehouse, we encourage providers to access their Data Submission report. To access this report, log in to QNet Exchange and click on the QIO Clinical Warehouse Feedback Reports. The Data Submission report will give the provider a detailed summary of the cases that entered the clinical warehouse.

Comment: Sixteen commenters recommended that CMS state submission and validation parameters clearly and document them. They also recommended that CMS provide 120-day notice prior to any changes to the parameters. The commenters added that there should be less frequent changes to the requirements.

Response: CMS and its contractors strive to give providers sufficient time to incorporate changes to submission and validation parameters. However, processing and logistical issues sometimes require more expedited implementation of these changes, because measure and policy changes frequently occur. To address this issue, CMS and JCAHO released an aligned manual on January 1, 2005. This release occurred 108 days prior to implementation of any of the provisions in the manual. Since that time, CMS and JCAHO have agreed to release documents at a minimum of 120 days prior to implementation. All manuals contain data file submission requirements and programming formats for each quarter.

Comment: Eight commenters requested that CMS be consistent when releasing any communications related to

[[Page 47424]]

hospital reporting, and that there should be one central point for all of these communications.

Response: Hospitals are required to establish relationships with the Quality Improvement Organizations for their States and, furthermore, must establish a formal relationship with the QIO Clinical Data Warehouse and its Web site, http://www.QNetexchange.org. All

policies and procedures concerning hospital reporting are communicated to the hospital community through these two channels. CMS communicates information about hospital reporting directly to the QIOs through the QIO Hospital Public Reporting contact for each State, using a formal system of memoranda (``SDPS memos'') which can be viewed at http://qionet.sdps.org. The QIOs are then responsible for dissemination of the

information to the appropriate hospital staff in each State. Responses to specific questions are addressed through the Quest system; CMS monitors responses and clarifications that are published on Quest. QIOs are expected to provide technical assistance, as well as provide e-mail blasts to all hospitals on any important topics and developments pertinent to reporting hospitals. Hospitals also receive direct communication or can seek assistance from the QIO Clinical Data Warehouse, by Internet (through QNet Exchange).

CMS and the JCAHO have formally agreed to work together to maintain common performance measures and to ensure that any communication concerning these measures is coordinated and consistent.

In addition to this formal system of communication, hospitals can obtain information or seek answers to specific questions on the monthly Hospital Open Door Forum (see http://www.cms.hhs.gov/opendoor/schedule.asp for schedule) on the hospital quality initiative.

Hospitals can also monitor CMS's activities to promote quality of care in hospitals by checking http://www.cms.hhs.gov/quality/hospital/. This

site includes information about CMS's involvement in the Hospital Quality Alliance, a public-private partnership to promote hospital public reporting (see http://www.hospitalcompare.hhs.gov).

Comment: A few commenters suggested that the only requirement to receive the full market basket update should be submission of data to the warehouse. These commenters stated the intent of the law was to limit the requirement to data submission, and not require validation. In addition, there were comments that the validation process is flawed and any link to payment should be delayed until data infrastructure and processes are improved.

Response: We disagree with the comments indicating that section 501(b) of Pub. L. 108-173 only requires the submission of data. The commenters stated that additional requirements were not contemplated by Congress. However, the validation process does not contradict Pub. L. 108-173. Section 501 (b) also states the submission of the data is to be in the ``form and manner specified by the Secretary''. We believe that validation requirements fall under this broad authority. This requirement does not appear to be to stringent based on validation results showing 98 percent of providers that submitted data for the third quarter 2004 are eligible for the full market basket update. While hospitals did encounter abstraction and processing issues, these problems were immediately resolved. CMS' policy on validation requirements are very lenient, and offer hospitals several opportunities to validate their data in order to receive the full update.

Comment: Two commenters recommended using the first quarter 2005 as the first quarter in which the validation process is used for calculating the full payment to occur in 2007.

Response: We appreciate the comment and will incorporate this comment into the decisionmaking process for the FY 2007 payment determination. It has been our intention to use continuous quarters of data, but CMS and JCAHO measures differed in several substantial areas (pre-alignment) prior to the third quarter 2004 calendar year. Based largely on these differences in measures, we chose to use validation results from third and fourth quarter 2004 calendar year discharges only using aligned measures to provide the highest possibility for validation for hospitals. The CMS and JCAHO measures were approximately 95 percent aligned for the third quarter 2004 calendar year discharges. Our validation results for this period w