[Federal Register: November 2, 1998 (Volume 63, Number 211)] [Rules and Regulations] [Page 58813-58862] From the Federal Register Online via GPO Access [wais.access.gpo.gov] [DOCID:fr02no98-16] [[Page 58813]] _______________________________________________________________________ Part II Department of Health and Human Services _______________________________________________________________________ Health Care Financing Administration _______________________________________________________________________ 42 CFR Part 405, et al. Medicare Program; Revisions to Payment Policies and Adjustments to the Relative Value Units Under the Physician Fee Schedule for Calendar Year 1999; Final Rule and Notice [[Page 58814]] DEPARTMENT OF HEALTH AND HUMAN SERVICES Health Care Financing Administration 42 CFR Parts 405, 410, 413, 414, 415, 424, and 485 [HCFA-1006-FC] RIN 0938-AI52 Medicare Program; Revisions to Payment Policies and Adjustments to the Relative Value Units Under the Physician Fee Schedule for Calendar Year 1999 AGENCY: Health Care Financing Administration (HCFA), HHS. ACTION: Final rule with comment period. ----------------------------------------------------------------------- SUMMARY: This final rule makes several policy changes affecting Medicare Part B payment. The changes that relate to physicians' services include: resource-based practice expense relative value units (RVUs), medical direction rules for anesthesia services, and payment for abnormal Pap smears. Also, we are rebasing the Medicare Economic Index from a 1989 base year to a 1996 base year. Under the law, we are required to develop a resource-based system for determining practice expense RVUs. The Balanced Budget Act of 1997 (BBA) delayed, for 1 year, implementation of the resource-based practice expense RVUs until January 1, 1999. Also, BBA revised our payment policy for nonphysician practitioners, for outpatient rehabilitation services, and for drugs and biologicals not paid on a cost or prospective payment basis. In addition, BBA permits certain physicians and practitioners to opt out of Medicare and furnish covered services to Medicare beneficiaries through private contracts and permits payment for professional consultations via interactive telecommunication systems. Furthermore, we are finalizing the 1998 interim RVUs and are issuing interim RVUs for new and revised codes for 1999. This final rule also announces the calendar year 1999 Medicare physician fee schedule conversion factor under the Medicare Supplementary Medical Insurance (Part B) program as required by section 1848(d) of the Social Security Act. The 1999 Medicare physician fee schedule conversion factor is $34.7315. DATES: Effective date: This rule this rule is effective January 1, 1999. Applicability date: Part 405 subpart D is applicable for private contract affidavits signed and private contracts entered into on or after January 1, 1999. This rule is a major rule as defined in Title 5, United States Code, section 804(2). Pursuant to 5 U.S.C. section 801(a)(1)(A), we are submitting a report to the Congress on this rule on October 30, 1998. Comment date: We will accept comments on interim RVUs for selected procedure codes identified in Addendum C and on interim practice expense RVUs for all codes as shown in Addendum B. Comments will be considered if we receive them at the appropriate address, as provided below, no later than 5 p.m. on January 4, 1999. ADDRESSES: Mail written comments (1 original and 3 copies) to the following address: Health Care Financing Administration, Department of Health and Human Services, Attention: HCFA-1006-FC, P.O. Box 26688, Baltimore, MD 21207-0488. If you prefer, you may deliver your written comments (1 original and 3 copies) to one of the following addresses: Room 443-G, Hubert H. Humphrey Building, 200 Independence Avenue, SW., Washington, DC 20201, or Room C5-14-03, 7500 Security Boulevard, Baltimore, MD 21244-1850. Because of staffing and resource limitations, we cannot accept comments by facsimile (FAX) transmission. In commenting, please refer to file code HCFA-1006-FC. Comments received timely will be available for public inspection as they are received, generally beginning approximately 3 weeks after publication of a document, in Room 443-G of the Department's offices at 200 Independence Avenue, SW., Washington, DC, on Monday through Friday of each week from 8:30 a.m. to 5 p.m. (phone: (202) 690-7890). FOR FURTHER INFORMATION CONTACT: Roberta Epps, (410) 786-4503 (for issues related to outpatient rehabilitation services). Stephen Heffler, (410) 786-1211 (for issues related to the Medicare Economic Index). Anita Heygster, (410) 786-4486 (for issues related to private contracts). Jim Menas, (410) 786-4507 (for issues related to Pap smears and medical direction for anesthesia services). Robert Niemann, (410) 786-4569 (for issues related to the drugs and biologicals policy). Regina Walker-Wren, (410) 786-9160 (for issues related to physician assistants, nurse practitioners, clinical nurse specialists, and certified nurse-midwives). Craig Dobyski, (410) 786-4584 (for issues related to teleconsultations). Stanley Weintraub, (410) 786-4498 (for issues related to practice expense relative value units and all other issues). SUPPLEMENTARY INFORMATION: Copies: To order copies of the Federal Register containing this document, send your request to: New Orders, Superintendent of Documents, P.O. Box 371954, Pittsburgh, PA 15250-7954. Please specify the date of the issue requested, and enclose a check or money order payable to the Superintendent of Documents, or enclose your Visa, Discover, or Master Card number and expiration date. Credit card orders can also be placed by calling the order desk at (202) 512-1800 (or toll free at 1-888-293-6498) or by faxing to (202) 512-2250. The cost for each copy is $8. As an alternative, you can view and photocopy the Federal Register document at most libraries designated as Federal Depository Libraries and at many other public and academic libraries throughout the country that receive the Federal Register. 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/index.html, 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). To assist readers in referencing sections contained in this preamble, we are providing the following table of contents. Some of the issues discussed in this preamble affect the payment policies but do not require changes to the regulations in the Code of Federal Regulations. Information on the regulation's impact appears throughout the preamble and not exclusively in part IX. Table of Contents I. Background A. Legislative History B. Published Changes to the Fee Schedule II. Specific Proposals for Calendar Year 1998; Response to Comments A. Resource-Based Practice Expense Relative Value Units 1. Resource-Based Practice Expense Legislation 2. Proposed Methodology for Computing Practice Expense Relative Value Units 3. Other Practice Expense Policies [[Page 58815]] 4. Refinement of Practice Expense Relative Value Units 5. Reductions in Practice Expense Relative Value Units for Multiple Procedures 6. Transition B. Medical Direction for Anesthesia Services C. Separate Payment for a Physician's Interpretation of an Abnormal Papanicolaou Smear D. Rebasing and Revising the Medicare Economic Index III. Implementation of the Balanced Budget Act A. Payment for Drugs and Biologicals B. Private Contracting with Medicare Beneficiaries C. Payment for Outpatient Rehabilitation Services 1. BBA 1997 Provisions Affecting Payment for Outpatient Rehabilitation Services a. Reasonable Cost-Based Payments b. Prospective Payment System for Outpatient Rehabilitation Services (1) Overview (2) Services Furnished by Skilled Nursing Facilities (3) Services Furnished by Home Health Agencies (4) Services Furnished by Comprehensive Outpatient Rehabilitation Facilities (5) Site-of-Service Differential (6) Mandatory Assignment 2. Uniform Procedure Codes for Outpatient Rehabilitation Services 3. Financial Limitation a. Overview b. Use of Modifiers to Track the Financial Limitation c. Treatment of Services Exceeding the Financial Limitation 4. Qualified Therapists 5. Plan of Treatment D. Payment for Services of Certain Nonphysician Practitioners and Services Furnished Incident to their Professional Services E. Payment for Teleconsultations in Rural Health Professional Shortage Areas IV. Refinement of Relative Value Units for Calendar Year 1999 and Responses to Public Comments on Interim Relative Value Units for 1998 A. Summary of Issues Discussed Related to the Adjustment of Relative Value Units B. Process for Establishing Work Relative Value Units for the 1999 Fee Schedule V. Physician Fee Schedule Update and Conversion Factor for Calendar Year 1999 VI. Provisions of the Final Rule VII. Collection of Information Requirements VIII. Regulatory Impact Analysis A. Regulatory Flexibility Act B. Resource-Based Practice Expense Relative Value Units C. Medical Direction for Anesthesia Services D. Separate Payment for a Physician's Interpretation of an Abnormal Papanicolaou Smear E. Rebasing and Revising the Medicare Economic Index F. Payment for Nurse Midwives' Services G. BBA Provisions Included in This Proposed Rule H. Impact on Beneficiaries Addendum A--Explanation and Use of Addenda B and C Addendum B--Relative Value Units (RVUs) and Related Information Addendum C--Codes with Interim RVUs In addition, because of the many organizations and terms to which we refer by acronym in this final rule, we are listing these acronyms and their corresponding terms in alphabetical order below: AANA: American Association of Nurse Anesthetists ABC: Activity based costing ABN: Advance Beneficiary Notice AHE: Average hourly earnings AMA: American Medical Association ANCC: American Nurses Credentialing Center ASA: American Society of Anesthesiologists ASOPA: American Society of Orthopedic Physician Assistants AWP: Average wholesale price BBA: Balanced Budget Act of 1997 BLS: Bureau of Labor Statistics CAAHEP: Commission on Accreditation of Allied Health Education Programs CF: Conversion factor CFR: Code of Federal Regulations CMSAs: Consolidated Metropolitan Statistical Areas CORF: Comprehensive outpatient rehabilitation facility CPEPs: Clinical Practice Expert Panels CPI: Consumer Price Index CPI-U: Consumer Price Index for All Urban Consumers CPS: Current Population Survey CPT: [Physicians'] Current Procedural Terminology CRNA: Certified Registered Nurse Anesthetist DME: Durable medical equipment DMEPOS: Durable medical equipment, prosthetics, orthotics, and supplies DRG: Diagnosis-related group EAC: Estimated acquisition cost ECI: Employment Cost Index ES-202 Data: Bureau of Labor Statistics from State unemployment insurance agencies ESRD: End-stage renal disease FDA: Food and Drug Administration FMR: Fair market rental FQHC: Federally qualified health center GAAP: Generally accepted accounting principles GAF: Geographic adjustment factor GPCI: Geographic practice cost index HCFA: Health Care Financing Administration HCPAC: Health Care Professionals Advisory Committee HCPCS: HCFA Common Procedure Coding System HHA: Home health agency HHS: [Department of] Health and Human Services HMO: Health maintenance organization HPSA: Health professional shortage area HRSA: Health Resources and Services Administration HUD: [Department of] Housing and Urban Development IPLs: Independent Physiologic Laboratories MedPAC: Medicare Payment Advisory Commission MEI: Medicare Economic Index MGMA: Medical Group Management Association MSA: Metropolitan Statistical Area MSA: Medicare Supplemental Insurance MVPS: Medicare volume performance standard NAIC: National Association of Insurance Commissioners NBCOPA: National Board on Certification for Orthopedic Physician Assistants NCCPA: National Council on Certification of Physician Assistants NPI: National provider identifier OBRA: Omnibus Budget Reconciliation Act OTIP: Occupational therapist in independent practice PC: Professional component PHS: Public Health Service PMSA: Primary Metropolitan Statistical Area PPI: Producer price index PPS: Prospective payment system PTIP: Physical therapist in independent practice RBRVS: Resource Based Relative Value Scale RHC: Rural health clinic RUC: [AMA's Specialty Society] Relative [Value] Update Committee RN: Registered nurse RVU: Relative value unit SMS: Socioeconomic Monitoring System SNF: Skilled nursing facility TC: Technical component TEFRA: Tax Equity and Fiscal Responsibility Act UPIN: Uniform provider identifier number I. Background A. Legislative History Since January 1, 1992, Medicare has paid for physicians' services under section 1848 of the Social Security Act (the Act), ``Payment for Physicians' Services.'' This section contains three major elements: (1) A fee schedule for the payment of physicians' services; (2) a sustainable growth rate for the rates of increase in Medicare expenditures for physicians' services; and (3) limits on the amounts that nonparticipating physicians can charge beneficiaries. The Act requires that payments under the fee schedule be based on national uniform relative value units (RVUs) based on the resources used in furnishing a service. Section 1848(c) of the Act requires that national RVUs be established for physician work, practice expense, and malpractice expense. Section 1848(c)(2)(B)(ii)(II) of the Act provides that adjustments in RVUs because of changes resulting from a review of those RVUs may not cause total physician fee schedule payments to differ by more than $20 million from what they would have been had the adjustments not been made. If this tolerance is exceeded, we must make adjustments to the conversion factors (CFs) to preserve budget neutrality. [[Page 58816]] B. Published Changes to the Fee Schedule In the June 5, 1998, proposed rule (63 FR 30820), we listed all of the final rules published through October 31, 1997 relating to the updates to the RVUs and revisions to payment policies under the physician fee schedule. In the June 5, 1998 proposed rule (63 FR 30818), we discussed several policy options affecting Medicare payment for physicians' services including resource-based practice expense RVUs, medical direction rules for anesthesia services, and payment for abnormal Pap smears. Also, we discussed the rebasing of the Medicare Economic Index from a 1989 base year to a 1996 base year. Further, based on BBA, we proposed revising our payment policy for nonphysician practitioners, for outpatient rehabilitation services, and for drugs and biologicals not paid on a cost or prospective payment basis. In addition, based on BBA, we discussed implementing new payment policies for certain physicians and practitioners who opt out of Medicare and furnish covered services to Medicare beneficiaries through private contracts. And finally, based on BBA, we discussed teleconsultation services. This final rule affects the regulations set forth at 42 CFR part 405, which consists of regulations on Federal health insurance for the aged and disabled; part 410, which consists of regulations on supplementary medical insurance benefits; part 414, which consists of regulations on the payment for Part B medical and other health services; part 415, which pertains to services furnished by physicians in providers, supervising physicians in teaching settings, and residents in certain settings; part 424, which pertains to the conditions for Medicare payment; and part 485, which pertains to conditions of participation: specialized providers. II. Specific Proposals for Calendar Year 1998; Response to Comments In response to the publication of the June 5, 1998 proposed rule, we received approximately 14,000 comments. We received comments from individual physicians, health care workers, and professional associations and societies. The majority of the comments addressed the proposal related to the resource-based practice expense policy. The proposed rule discussed policies that affect the number of RVUs on which payment for certain services would be based. Certain changes implemented through this final rule are subject to the $20 million limitation on annual adjustments contained in section 1848(c)(2)(B)(ii)(II) of the Act. After reviewing the comments and determining the policies we will implement, we have estimated the costs and savings of these policies and added those costs and savings to the estimated costs associated with any other changes in RVUs for 1999. We discuss in detail the effects of these changes in the Regulatory Impact Analysis (section IX). For the convenience of the reader, the headings for the policy issues in this section correspond to the headings used in the June 5, 1998 proposed rule. More detailed background information for each issue can be found in the June 5, 1998 proposed rule. A. Resource-Based Practice Expense Relative Value Units 1. Resource-Based Practice Expense Legislation Section 121 of the Social Security Act Amendments of 1994 (Public Law 103-432), enacted on October 31, 1994, required us to develop a methodology for determining resource-based practice expense RVUs for each physician's service that would be effective for services furnished in 1998. In developing the methodology, we were required to consider the staff, equipment, and supplies used in providing medical and surgical services in various settings. The legislation specifically required that, in implementing the new system of practice expense RVUs, we apply the same budget-neutrality provisions that we apply to other adjustments under the physician fee schedule. On August 5, 1997, the President signed the BBA into law. Section 4505(a) of BBA delayed the effective date of the resource-based practice expense RVU system until January 1, 1999. In addition, BBA provided for the following revisions in the requirements to change from a charge-based practice expense RVU system to a resource-based method. Instead of paying for all services entirely under a resource-based system in 1999, section 4505(b) of BBA provided for a 4-year transition period. The practice expense RVUs for the year 1999 will be the product of 75 percent of charge-based RVUs (1998) and 25 percent of the resource-based RVUs. For the year 2000, the percentages will be 50 percent charge-based and 50 percent resource-based. For the year 2001, the percentages will be 25 percent charge-based and 75 percent resource-based. For subsequent years, the RVUs will be totally resource-based. Section 4505(e) of BBA provided that, for 1998, the practice expense RVUs be adjusted for certain services in anticipation of the implementation of resource-based practice expenses beginning in 1999. Practice expense RVUs for office visits were increased. For other services whose practice expense RVUs (determined for 1998) exceeded 110 percent of the work RVUs and were provided less than 75 percent of the time in an office setting, the 1998 practice expense RVUs were reduced to a number equal to 110 percent of the work RVUs. This limitation did not apply to services that had a proposed resource- based practice expense RVU in the June 5, 1998 proposed rule that was an increase from its 1997 practice expense RVU. The total of the reductions under this provision was less than the statutory maximum of $390 million. The procedure codes affected and the final RVUs for 1998 were published in the October 31, 1997 final rule (62 FR 59103). Section 4505(d)(2) of BBA required that the Secretary transmit a report to the Congress by March 1, 1998, including a presentation of data to be used in developing the practice expense RVUs and an explanation of the methodology. A report was submitted to the Congress in early March 1998. Section 4505(d)(3) required that a proposed rule be published by May 1, 1998, with a 90-day comment period. For the transition to begin on January 1, 1999, a final rule must be published by October 30, 1998. BBA also required that we develop new resource-based practice expense RVUs. In developing these new practice expense RVUs, section 4505(d)(1) required us to-- Utilize, to the maximum extent practicable, generally accepted accounting principles that recognize all staff, equipment, supplies, and expenses, not just those that can be tied to specific procedures, and use actual data on equipment utilization and other key assumptions; Consult with organizations representing physicians regarding the methodology and data to be used; and Develop a refinement process to be used during each of the four years of the transition period. 2. Proposed Methodology for Computing Practice Expense Relative Value Units (See Addendum B in the June 5, 1998 proposed rule (63 FR 30888) for a detailed technical description of the proposed methodology.) In the June 5, 1998 proposed rule (63 FR 30827), we proposed a methodology [[Page 58817]] for computing resource-based practice expense RVUs that uses the two significant sources of actual practice expense data we have available: the Clinical Practice Expert Panel (CPEP) data and the American Medical Association's (AMA's) Socioeconomic Monitoring System (SMS) data. This methodology is based on an assumption that current aggregate specialty practice costs are a reasonable way to establish initial estimates of relative resource costs of physicians' services across specialties. It then allocates these aggregate specialty practice costs to specific procedures and, thus, can be seen as a ``top-down'' approach. Practice Expense Cost Pools We used actual practice expense data by specialty, derived from the 1995 through 1997 SMS survey data, to create six cost pools: administrative labor, clinical labor, medical supplies, medical equipment, office supplies, and all other expenses. There were three steps in the creation of the cost pools. Step 1: We used the AMA's SMS survey of actual cost data to determine practice expenses per hour by cost category. The practice expenses per hour for each physician respondent's practice was calculated as the practice expenses for the practice divided by the total number of hours spent in patient care activities by the physicians in the practice. The practice expenses per hour for the specialty are an average of the practice expenses per hour for the respondent physicians in that specialty. Step 2: We determined the total number of physician hours, by specialty, spent treating Medicare patients. This was calculated from physician time data for each procedure code and the Medicare claims data. The primary sources for the physician time data were surveys submitted to the AMA's Specialty Society Relative Value Update Committee (RUC) and surveys done by Harvard for the initial establishment of the work RVUs. Step 3: We then calculated the practice expense pools by specialty and by cost category by multiplying the practice expenses per hour for each category by the total physician hours. Cost Allocation Methodology For each specialty, we separated the six practice expense pools into two groups and used a different allocation basis for each group. For group one, which includes clinical labor, medical supplies, and medical equipment, we used the CPEP data as the allocation basis. The CPEP data for clinical labor, medical supplies, and medical equipment were used to allocate the clinical labor, medical supplies, and medical equipment cost pools, respectively. For group two, which includes administrative labor, office expenses, and all other expenses, a combination of the group one cost allocations and the physician fee schedule work RVUs were used to allocate the cost pools. For procedures performed by more than one specialty, the final procedure code allocation was a weighted average of allocations for the specialties that perform the procedure, with the weights being the frequency with which each specialty performs the procedure on Medicare patients. Other Methodological Issues Professional and Technical Component Services Using the methodology described above, the professional and technical components of the resource-based practice expense RVUs do not necessarily sum to the global resource-based practice expense RVUs since specialties with different practice expenses per hour provide the components of these services in different proportions. We made two adjustments to the methodology, depending on the specific HCFA Common Procedure Coding System (HCPCS) code, so that the professional and technical component practice expense RVUs for a service sum to the global practice expense RVUs. Practice Expenses per Hour Adjustments and Specialty Crosswalks Since many specialties identified in our claims data did not correspond exactly to the specialties included in the practice expenses tables from the SMS survey data, it was necessary to crosswalk these specialties to the most appropriate SMS specialty category. (See Table 3 in the June 5, 1998 proposed rule (63 FR 30833) for a listing of all proposed crosswalks.) We also made the following adjustments to the practice expense per hour data: We set the medical materials and supplies practice expenses per hour for the specialties of ``Oncology'' and ``Allergy and Immunology'' equal to the medical materials and supplies practice expenses per hour for ``All Physicians,'' stating that we make separate payment for the drugs furnished by these specialties. We based the administrative payroll, office, and other practice expenses per hour for the specialties of ``Physical Therapy'' and ``Occupational Therapy'' on data used to develop the salary equivalency guidelines for these specialties. We set the remaining practice expense per hour categories equal to the ``All Physicians'' practice expenses per hour from the SMS survey data. Due to uncertainty concerning the appropriate crosswalk and time data for the nonphysician specialty ``Audiologist,'' we derived the resource-based practice expense RVUs for codes performed by audiologists from the practice expenses per hour of the other specialties that perform these codes. Because we believed that the use of the average practice expenses per hour should create the appropriate practice expense pool for radiology, we did not attempt to differentiate the practice expenses per hour for radiologists according to who owned the equipment. Time Associated With the Work Relative Value Units The time data resulting from the refinement of the work RVUs have been, on the average, 25 percent greater than the time data obtained by the Harvard study for the same services. We increased the Harvard time data in order to ensure consistency between these data sources. For services such as radiology, dialysis, and physical therapy, and for many procedures performed by independent physiological laboratories and the nonphysician specialties of clinical psychologist and psychologist (independent billing), we calculated estimated total physician times for these services based on work RVUs, maximum clinical staff time for each service as shown in the CPEP data, or the judgment of our clinical staff. We calculated the time for Current Procedural Terminology (CPT) codes 00100 through 01996 using the base and time units from the anesthesia fee schedule and the Medicare allowed claims data. We received the following comments on our proposed methodology to calculate resource-based practice expense RVUs: Top-Down Methodology Comment: Most of the physician specialty societies commenting on our proposed general methodology supported the use of the top-down approach as the most reasonable methodology for developing resource- based practice expense RVUs, and the most responsive approach to the requirements of BBA. This was echoed by comments from several nonphysician organizations, the Association of American Medical Colleges, and the Medical Group Management [[Page 58818]] Association, as well as several hundred individual commenters. These commenters supported the top-down method for a variety of reasons: It reflects the relative values of physicians' actual practice expenses. It uses the best available sources of aggregate practice expense data. It recognizes specialty-specific indirect costs. It does not rely upon arbitrary, distorting data adjustments such as ``linking'' and ``scaling.'' It is conducive to refinement. MedPAC also agreed that this approach is necessary, because of limitations in the CPEP process and because the top-down approach assures that all practice costs are reflected in the RVUs. However, several organizations, mainly representing primary care physicians and supported by comments from individual physicians, opposed the use of a top-down methodology to develop practice expense RVUs. They argued that the top-down approach is not resource-based but, rather, rewards higher paid physicians who have spent more in the past, regardless of the extent to which these expenditures contributed to patient care. Thus, the commenters claimed that the top-down approach perpetuates the inequities in the current charge-based practice expense RVUs that the implementation of a resource-based practice expense system was supposed to correct. One commenter also claimed that the top-down approach is not responsive to the requirements of BBA, as the methodology is not based on generally accepted accounting principles. Further, the commenter argued that this new proposal is not more responsive to the concerns of the medical community in general but, rather, only benefits those specialties whose income was projected to decline under the bottom-up approach. A specialty society representing clinical oncology opposed the top- down methodology because-- It does not actually measure appropriate input resource costs and thus pays for inefficiencies; It overpays hospital-based and underpays office-based services; and The RVUs for individual codes cannot be refined because of the use of macro-specialty per hour costs. There were several comments that expressed concern about the more specific impacts of the methodology. A major primary care organization pointed out that, under the 1997 proposed rule, an internist would have had to provide only 15 midlevel established patient office visits to obtain the practice expense reimbursement of a single coronary triple- bypass graft, compared to 40 visits under our current proposal. One organization opposed the use of the top-down approach because of the estimated reduction in payments to radiology and radiation oncology. Another commenter, representing pathologists, expressed concern that because pathology received small gains under the bottom-up method, but a 10 percent reduction under the top-down, there are possible flaws in the top-down methodology. A few of the above comments specifically recommended that we adopt a new bottom-up approach that is responsive to the BBA, the General Accounting Office (GAO), and the concerns of the medical community. Another organization commented that both top-down and bottom-up methodologies are inherently flawed, and that we should consider an entirely new payment algorithm using type of practice. One of the major primary care organizations concluded that the top-down methodology is only a reasonable starting point that will need to be improved during refinement in order to meet the original intent of improving practice- expense payments for undervalued primary care and other office-based services. Response: As we stated in our proposed rule, BBA requires us to ``utilize, to the maximum extent practicable, generally accepted cost accounting principles which recognize all staff, equipment, supplies, and expenses, not just those which can be tied to specific procedures****'' We still believe that the top-down methodology is more responsive to this BBA requirement. By using aggregate specialty practice costs as the basis for establishing the practice expense pools, the top-down method recognizes all of a specialty's costs, not just those linked to specific procedures. We also believe that the other reasons outlined in the proposed rule for preferring the top-down method are still valid. It answers many of the criticisms and questions from the medical community and the GAO regarding the bottom-up method's indirect practice expense allocation method, treatment of administrative costs, and use of caps and linking. However, we agree that a possible weakness of the top-down approach is that it may perpetuate historical inequities in the current charge- based practice expense RVUs. More highly paid physicians would presumably have more revenues that could subsequently be spent on their practices. We believe this issue should be discussed during the refinement process. Comment: One major organization commented that we will need to develop an alternative method for new and revised codes that are not included in the SMS data because having multiple methods would lead to questionable validity. Response: It will not be necessary to develop an alternate methodology for refinement of new and revised codes. Once direct inputs are assigned to the new and revised codes, allocation to these codes will follow the same methodology used for all other services. (See Section II.A.4, Refinement of Practice Expense RVUs.) Comment: Two major primary care organizations expressed concern that we did not consult with the physician community about our intention to abandon, rather than refine, our originally proposed bottom-up approach, since they had assumed we would only be modifying our original methodology. They commented that this is of greater concern in light of BBA's requirement that we consult with physicians regarding our methodology and of GAO's recommendation that we refine, with no mention of replacing, the bottom-up method. One of the comments stated, that as the GAO found the bottom-up method acceptable, their society would like the GAO's assurance that the new method is sound. Response: We believe we carried out the BBA requirement to consult with physician organizations. There were extensive consultations with physicians, including the validation panels, the cross specialty panel, and the indirect cost symposium. During the course of each of these meetings, physicians and others pointed out serious problems with the bottom-up methodology. We have had two multispecialty meetings this year to explain our proposed methodology and have also had numerous meetings and discussions with many specialty societies. During all these meetings we carefully listened to all points of view and to suggestions for developing the new proposal. Following this lengthy consultation process, we published our new proposal with a 90-day comment period. This provided further opportunities for all interested groups to review and comment on this proposal. It is true that the GAO did not recommend that we totally replace our bottom-up approach. It is our understanding that the GAO was not asked to review alternative methods. In any case, their report did not recommend against adopting a new methodology. Their report did point out [[Page 58819]] several significant weaknesses in our original approach that we believed were better responded to by adopting a top-down methodology. Comment: One organization urged that we publish the practice- expense RVUs three ways, using a top-down, a bottom-up, and a hybrid approach that uses SMS data for indirect costs and CPEP data for direct costs. The bottom-up and hybrid approaches should reflect the recommendations previously received relating to scaling, linking, and the treatment of administrative costs. This could provide a basis for developing comments that compare the interim practice expense RVUs with those derived from a modified bottom-up approach. The commenter stated that we should be open to considering arguments for a change in the interim practice expense RVUs based on a group's determination that the values under the bottom-up approach were more accurate. Response: We believe that we proposed the methodology for developing resource-based practice expense RVUs that best responds to the requirements of the Social Security Act Amendments of 1994 and BBA. From a practical standpoint, it would be very difficult to deal with the inconsistencies between RVUs for various services that have been derived from totally different methodologies. SMS Data Comment: Almost all specialty society commenters, and many individual commenters, raised questions concerning shortcomings in the SMS data, though several commented that SMS is the most appropriate data source to use in developing specialty-specific practice expense RVUs. As we noted in the proposed rule, the AMA itself pointed out that the survey had not been designed to support the development of practice expense RVUs. The AMA also stated that the sample size, the response rate, and the fact that data was collected on the physician level, rather than the practice level, raised methodological issues. Many commenters echoed these concerns, and many raised what they saw as further general methodological problems: MedPAC expressed concern about three types of potential errors in the SMS data: the sampling error and nonresponse error originally identified in our proposed rule and measurement error. Some of this measurement error could occur because the survey measures physician-level rather than practice-level costs, as noted above. In addition, there could be measurement error by using a self-reported survey if no mechanism exists to verify the information provided. MedPAC suggested that we could reduce these errors through additional data collection, perhaps implementing a subsample of SMS survey participants, through an analysis of nonresponse error that compares respondents with nonrespondents, through AMA's plans to do a practice-level survey every other year, and through considering methods, other than actual audits, to verify survey responses. Several of the smaller specialties, such as maxillofacial, pediatric, vascular and thoracic surgeons, cardiology and gynecology subspecialties, geriatricians, and pulmonologists expressed concern with the validity and reliability of SMS data for those specialty and subspecialty groups not adequately represented in the SMS survey. A commenter also stated that academic and hospital-based specialties, such as critical care and neonatology, were not appropriately represented. Many specialty societies requested that we consider practice expense data obtained by under-represented specialty and subspecialty groups. Several nonphysician specialties, though supporting the use of SMS data, raised the need to modify the survey to include nonphysicians in the future. A commenter stated that, because nonphysicians were not represented in the SMS survey, we have been forced to make an educated guess about which specialties they most resemble. Another commenter pointed out that the SMS data contains no information about osteopathic physicians. Several specialties, regardless of their overall sample size, expressed concerns about the combining together of subspecialties with differing practice costs. For example, organizations representing cardiologists commented that it is not known how many in their sample were providing evaluation and management services, as opposed to performing equipment intensive procedures that have much higher costs. Two specialty societies representing nuclear physicians, along with several hundred individual commenters, objected to the small sample of this subspecialty, with its high costs related to the use of radiopharmaceuticals, being combined with radiologists into a single practice expense pool. The comments recommended that we increase nuclear medicine's practice expense RVUs by 20 percent. Similarly, a vascular surgery organization objected to being combined with cardiothoracic surgeons, who made up 75 percent of the sample and whose practice style differs substantially from vascular surgeons. An organization representing pediatrics expressed concern that pediatric subspecialties were grouped together with their adult counterparts, such as gastroenterology. The AMA commented on this point that it plans refinements for future surveys to enhance the utility of the data. Several commenters noted that the survey consisted of physician-owned practices, despite the trend toward more physicians working as employees, resulting in a possible bias toward solo or small group practices. For example, one commenter stated that the majority of emergency room physicians now work as employees or under contract. Another commenter asserted that the majority of pediatricians list their status as ``employed.'' The AMA commented, in this regard, that a key refinement to the SMS survey will be the development of a practice- level survey to complement the current process. One commenter questioned our assumption that physician respondents to SMS share practice expenses equally with all other physician owners in the practice, since there is no data to show that this is the prevalent method. An organization representing nurses commented that issues related to changes in acuity and case mix in ambulatory care are not being addressed, particularly as they pertain to the increased professionalization of clinical staff types. The organization argued that there is a need to incorporate into the survey process a clearer distinction between the types of clinical staff that are employed based on specialty practice. Concerns were raised by some commenters that the SMS data did not always include the actual costs of a given specialty. Several organizations representing radiologists, radiation oncologists, and cardiologists commented that the methodology employed by the SMS survey consistently underestimated the actual costs of equipment. Organizations representing emergency room physicians, supported by the comment from the AMA, argued that the significant costs of both stand- by time and uncompensated care are not reflected in the SMS data and that these costs need to be recognized. A gastroenterology specialty society asserted that the SMS data grossly understated actual expenses when compared to its own study. Two commenters stated that costs for home visits, such as travel expenses and insurance, are not adequately represented in the data. One organization commented that the SMS [[Page 58820]] data fails to adequately incorporate resources, including billing, nursing time, and transportation costs for audiologists utilized in settings such as skilled nursing facilities. One commenter stated that the added costs for compliance with federal initiatives, such as anti-fraud and abuse efforts and the new evaluation and management documentation guidelines, are not yet reflected in the SMS data. These costs should be recognized during the refinement process and included in future surveys. On the other hand, several commenters argued that costs were included in the SMS data that should be excluded because they are paid for separately from the physician fee schedule. One commenter pointed to separately reimbursable supplies and drugs, and another to the costs of taking physician staff into the hospital, as examples of costs included in SMS that could lead to a double payment by Medicare. A society representing vascular surgeons commented that the technical component of noninvasive vascular laboratory testing falls into this ``gray zone.'' A national specialty society commented that the AMA analysis of the ``zero'' responses by specialty by cost categories (that is, those cost categories where respondents indicated there were no costs) shows that a significant percentage of pathologists' responses for direct cost categories are zero as compared to the ``zero'' response rates for all physicians. The comment requested that the SMS pathology data be cleared of all ``zero'' responses for all cost categories, not just for the total cost category, prior to the calculation of mean costs. For the purpose of calculating practice expense per hour for pathology, the society said, we should only use data from pathologists who incur a particular cost. There were a number of comments concerning the SMS data on the specialty-specific physician patient care hours, which is one of the variables used to compute the practice expense per hour for each specialty: Many specialty societies stated their concern that in the calculation of the specialty-specific practice expense per hour, specialties working the longest hours are disadvantaged. One commenter pointed out that practice expense is not uniformly distributed over the course of a given day; there are less costs when patient care takes place after, rather than during, office hours. Another commenter argued that our approach assumes that all of the patient care hours in the SMS survey are reflected in our claims data. However, the commenter stated, much time spent in patient care activities is not billable, such as the involvement of transplant surgeons in patient care after the initial assessments but prior to the actual transplants. One specialty society stated that hospital-based physicians' hours of work are probably overstated, as they will include total time spent in the facility and not just hours of providing patient services. One commenter questioned both the accuracy of the SMS data on hours worked per week, as well as our assumption that the level of practice expense incurred increases proportionally with the hours spent in patient care. An organization stated that physician reports of number of hours are less reliable than the reports of costs and are prone to overstatement. For these reasons, five specialty societies recommended using a standardized work week, usually a 40-hour week, for all specialties. Many other specialty groups argued equally vehemently against any standardization of the patient care hours. One group commented that subjective adjustments to the SMS data, especially those which reallocate practice expenses among specialties, should be avoided. The comment added that suggestions that a standardized 40-hour work week be imposed on the data should be rejected because the proposal is driven by an arbitrary, subjective presumption that cross- specialty practice expense variations are ``too large.'' Another group argued that, as many physicians work more than a 40- hour week, such an adjustment would introduce additional error into the data and distort the relationship between different specialties' practice expenses per hour. Three organizations were concerned about the advantage given to specialties that use nonphysician practitioners who are not reimbursable. In such cases, the physician would incur practice expense costs, but the time of practitioners would not be included in the physician patient care hours in the denominator of the practice expense per hour calculation. On the other hand, another commenter stated that we should not adjust the SMS data for midlevel practitioners, such as optometrists or audiologists, as physician practices employing midlevel practitioners are likely to be more complex than a physician-only operation. One specialty society commented that the demographics of the SMS survey are not clear, as there are no assurances that the sample is not biased towards one particular area of the country and does not exclude some areas. Response: We believe that most of the above comments identified important areas for needed future improvement in our data collection efforts on aggregate specialty-specific practice expense. However, although the SMS survey was not initially intended to be used to develop practice expense RVUs, we believe it is the best available source of data on actual multispecialty practice costs that allows us to recognize all staff, equipment, supplies, and expenses, not just those that can be tied to specific procedures. Many specialties supported this. For example, a specialty society commented, ``As with any complex database, the AMA SMS database is not perfect. It is, however, the best available source of data for aggregate practice expenses.'' The Medical Group Management Association (MGMA) stated in its comment that, ``The SMS survey data is the most appropriate and only primary data set in existence to determine specialty specific costs pools.'' We also need to point out that many of the weaknesses in the SMS data could well be found in any other survey, whether undertaken by us, some other national group, or a medical specialty society. Problems with sample size and response rate have plagued other previous attempts to gather reliable data on practice expenses. Problems with measurement error may be a serious impediment for survey data that is collected with the purpose of influencing the level of a given specialty's practice expense pool. In fact, we believe one advantage of the current SMS data is that they were collected before the 1997 and 1998 proposed rules were published. We recognize that some specialties are under-represented or not appropriately represented in the SMS data and some are not included at all. We also acknowledge that additional data may need to be obtained and some adjustments made. One of our most important tasks during the immediate refinement period will be to work with the AMA and the medical community to consider possible ways to improve the representativeness of the aggregate specialty-specific data so that sampling error is decreased. As part of the refinement, we will also need to develop strategies to eliminate as many sources of nonresponse and measurement error as possible. (For further information on our refinement efforts to improve the accuracy of our [[Page 58821]] data, see Section II.A.4, Refinement of Practice Expense RVUs.) As indicated earlier, we believe an advantage of the SMS data we used is that it was collected prior to the proposed rule. In fact, it was collected prior to the original proposal in 1997 that was delayed by BBA and that would have resulted in large redistributions among specialties. We are very concerned, though, about the potential biases that may exist in any subsequent survey data collected by the SMS process or other surveys. We especially believe there is a problem in using data collected and submitted to us by individual specialties. We believe it is more appropriate to use data collected at the same time by an independent surveyor for a wide variety of specialties that both gain and lose under the proposal. Further, now that it is widely known how these survey data are being used, every specialty has an incentive to ensure that their data are as high as possible in future surveys. We agree with MedPAC that it may not be possible for Medicare to audit these data and that it is essential that alternatives be established by SMS and others. Perhaps specialty data that significantly changes in a future survey should be selectively audited by SMS through an independent auditor or other appropriate entity before being considered for use by us. We will consult with physician groups and others about this during the refinement process. Comment: One national organization suggested the use of MGMA survey data either as a supplement or alternative to SMS in the future. Response: We do not believe that the MGMA survey could currently be used as an alternative to SMS. As we noted in our proposed rule, due to selective sampling and low response rate, this survey is not representative of the population of physicians and cannot be used to derive code-specific RVUs. This view is based on consultations with MGMA representatives. However, we do believe that this survey data can be used as one way to validate the general accuracy of the SMS data. We have analyzed the MGMA data and have concluded that, in general, it supports the relative specialty-specific ranking of the practice expense per hour data derived from the SMS survey. Comment: One specialty society recommended using median, instead of mean, values to calculate each specialty's practice expense per hour. This comment argued that the use of medians would eliminate outliers and is statistically more appropriate. However, three other organizations specifically commented supporting our decision to use mean SMS data rather than median data. These comments asserted that, particularly with a small sample, use of the median would obscure any major differences in practice costs within a specialty. Response: We will continue to calculate the practice expenses per hour by using the mean values for each specialty, at least for the purposes of this final rule. This is another issue that can be revisited during the refinement period. Comment: Organizations representing emergency room physicians, as well as several hundred individual commenters, claimed that the SMS data seriously under-represented the true practice costs of emergency care. The commenters stated that the SMS data, as noted above, did not include costs of uncompensated care, much of it mandated under the Federal Emergency Medical Treatment and Active Labor Act (Public Law 99-272), nor stand-by expenses. In addition, the comments argued, the SMS data failed to capture a representative cross-section of their types of practice arrangements; the SMS survey focused on physician owners, but the majority of emergency room physicians work as employees or under contract. Therefore, one commenter asserted, SMS did not include the largest single expense for most emergency physicians: the costs associated with employment by practice management firms, which can total between 30-40 percent of the physician's fee. One of the specialty societies included with its comments the results of a study it commissioned, which showed that the mean practice expense per hour for emergency physicians was $27.33, more than double the $13 per hour based on SMS, even without including uncompensated care. If we are not willing at this time to substitute this survey data for that from the SMS, the organization recommended, with support from a comment from the AMA, that we crosswalk emergency medicine to the practice expense per hour for ``All Physicians,'' which is $67.50. Response: Though many specialties must deal with the issue of uncompensated care, we do agree that it may pose a particular problem for emergency physicians, who are obligated under law to treat any patient regardless of the patient's ability or willingness to pay for treatment. Therefore, the amount of patient care hours spent on uncompensated care could be significantly higher for emergency medicine than for any other specialty. These issues require further examination. In the meantime, we will make an adjustment in our calculation of the practice expense per hour for emergency medicine by using the ``All Physicians'' practice expense per hour to calculate the administrative labor and other expenses cost pool. We will continue to calculate the clinical labor, supply, equipment, and office cost pools using the SMS- derived data, as it seems unlikely that, as a hospital-based specialty, emergency medicine's costs for these categories would approximate those of the average physician. Comment: Many commenters argued that the reductions published in the June 5, 1998, NPRM for services without work RVUs were inappropriate. The commenters represented a wide spectrum of specialties including radiology, radiation oncology, cardiology, independent physiological and other laboratories, psychology, audiology, dermatology, and others. These comments focused on the fact that AMA does not survey some of the entities that provide these services. They argued that the CPEP data are flawed and the indirect allocation methodology is biased. Response: Although it is true that the AMA does not survey the entities that provide some of these services, this does not necessarily mean that these services are inadequately represented in the SMS data. If these services (or in the case of technical component services, the associated global services) are provided in the practices of physician owners surveyed by the SMS in the same proportion as they are reflected in our claims data, the practice expense per hour calculations and the practice expense pools are reasonable. If the CPEP data accurately contain the direct cost inputs for these services, then the direct practice expense pool is being allocated appropriately. With regard to the indirect allocation methodology, we are modifying it to increase the weight of the direct costs in the allocation, as discussed elsewhere. However, the possibility exists that inaccuracies in the CPEP data for these services are causing the substantial reductions seen in the NPRM. Therefore, because we are not altering the CPEP at this time, as an interim solution until the CPEP data for these services have been validated, we have created a practice expense pool for all services without work RVUs regardless of the specialty that provides them. We allocated this practice expense pool to procedure codes using the current practice expense relative value units. [[Page 58822]] While we are not convinced by the comments that were received to date regarding a bias in the SMS survey data against these services, we acknowledge those concerns and will examine this issue during the refinement process. Comment: The College of American Pathologists (CAP) requested that patient care time included in the SMS data that is spent in autopsies and supervision of technicians and paraprofessionals be excluded from the patient care hours used to calculate the practice expense per hour for pathology services. The commenter stated that these are Part A services for which pathologists rarely incur any direct costs. The AMA supported these adjustments and estimated the percentage of total pathology patient care hours attributable to autopsy and supervision services at 6 and 15 percent, respectively. CAP also asked that some portion of the patient care hours category of ``personally performing nonsurgical laboratory procedures including reports'' be eliminated for 1999 when determining pathologists' total patient care hours, as the SMS data includes both Part A and Part B services. CAP stated that we should work with the CAP and the AMA to determine the appropriate adjustment. Response: Since pathologists have more Part A reimbursement than any other specialty, we will decrease the number of patient care hours by 6 percent for autopsies and 15 percent for supervision services. However, until we have more information about the appropriate adjustment for ``personally performing non-surgical laboratory procedures including reports,'' the hours for those services cannot be eliminated from our calculations. This point, as well as the general issue of nonbillable hours, should be revisited during refinement. Comment: Many specialty societies have commented on specific problems with the SMS data that affect their own specialty and have requested that we supplement or replace the SMS data with data provided with their comments. Response: There is not sufficient time before publication of the final rule to begin to validate either the methodology or findings of the submitted data. Since changes in any specialty's practice expense per hour would have an impact on other specialties, we do not believe it would be equitable to make any sweeping changes without the adequate review that the refinement process can achieve. In addition, we stated in our proposed rule that, for those larger specialties included in the SMS survey, ``we are unlikely to make any changes in the final rule****'' Therefore, we will continue to use the SMS-derived practice expense per hour for these specialties, but will ensure that all of the submitted data will be considered during the refinement process. CPEP Data Comment: Though one major specialty society commented that the CPEP data, in general, is relatively sound, many comments pointed out problems with the CPEP process and with the data derived from that process: One group commented that the CPEPs did not have adequate representation from practice managers; that there was no uniform policy dealing with issues such as duplication of time or efficiencies that might result from performing more than one task at a time; and that there was inadequate time allotted for CPEPs to meet. Several subspecialties pointed out that they were not included in the CPEP process and that this could have led to the undervaluing of their services. Several commenters recommended that we use the CPEP data as validated and refined by the validation panels. One organization commented that the CPEP data are flawed since only 200 codes were reviewed by validation panels. One primary care group argued that we should not abandon edits and modifications to raw CPEP data, as many codes are performed by more than one specialty, and inaccuracies in the CPEP data can affect several specialties. Two organizations commented that the CPEPs used what is now obsolete salary and benefits data, at least for sonographers and vascular technologists. One of these comments pointed out that for some codes, a different cost was computed for the same equipment. Another specialty society recommended that a review of prices and quantities for supplies and equipment be included as part of the refinement process. Two commenters were concerned that the CPEP data include expenses that can be billed separately. A primary care specialty society argued that we should edit out all direct inputs for services to hospital patients. The comment mentioned that since these services are paid for outside of the practice expense RVUs, failure to exclude these inputs can distort relativity across categories of services such as surgical services and office visits. One commenter clarified that the costs of therapy aides are a part of practice expense and should be reflected in the CPEP data, while the services of therapy assistants are included in the work RVUs. Response: We are aware that the raw CPEP data we have used in our proposed methodology need further review. We also share many of the concerns raised by those commenting on the issue. However, we believe that the CPEP resource estimates, which were developed by practitioners representing all the major specialties, are the best procedure level data available at this time. Under our top-down methodology, the CPEP inputs are used solely to allocate each specialty's practice expense pool to the procedures performed by that specialty. We have always believed that the relative input estimates within families of codes for each specialty's CPEP data were generally appropriate. In addition, the most contentious CPEP values were the varying estimates for the administrative staff times, and these values are not utilized in our top-down approach. We chose not to apply the edits, caps, or linking that had originally been proposed in our 1997 proposed rule as part of our bottom-up methodology. These edits had met with severe criticism from the medical community and were questioned by the GAO. We also did not use the revised inputs from the validation panels we held in October 1997, as these panels only came to consensus on about 200 codes, and we were not convinced that all of the revised values were correct. However, we know that there is much needed improvement in the CPEP data, and the identification and correction of any CPEP errors whether in staff times, supplies, equipment, or pricing will be a major focus of our refinement process. Comment: One specialty society commented that we erred in not incorporating increases in staff time recommended by validation panels. Partly as a result, the practice expense RVUs for gastroenterologists' out-of-office billing, scheduling, and record keeping are inadequate. Another commenter stated that there were discrepancies in the administrative data for skilled nursing facility services, with subsequent visit codes being assigned only half of the billing time of initial visits. A third commenter requested that we standardize the administrative staff types according to the validation panels' recommendations. Three commenters stated that we do not account for the costs of maintaining an office full-time when the physician is providing services out of the office. [[Page 58823]] Response: As stated above, under our proposed methodology, CPEP administrative staff times have no effect on the practice expense RVUs calculated for any code. The costs of maintaining an office while the physician is providing services in a facility should be captured in the SMS cost data and, thus, are a part of each specialty's practice expense pool. As these would be indirect costs, they would be included in the practice expense for each service by use of our allocation methodology, which utilizes both directs costs and the physician work RVUs. Comment: Almost 30 specialty societies submitted specific CPT code- level changes for the CPEP input data for clinical and administrative labor time, supplies, and equipment for just under 3000 CPT codes. In addition, many commenters included lists of codes with practice expense RVUs that were considered anomalous, either within a code family, or in relation to comparable codes. We also received comments from several organizations with recommendations for revised crosswalks for those codes not valued by the CPEPs, as well as recommended in-office inputs for some codes that are now being done in the office, but were only given practice expense RVUs for the facility setting. Response: We had intended to make the CPEP revisions requested by a given specialty as part of the final rule if the recommendations appeared reasonable and if there would be no significant impact on any other specialty. However, given the huge volume of recommended revisions--over a third of the codes in the fee schedule would be affected--acceptance of the recommended changes across the board would almost certainly have a spill-over impact on many subspecialties and between sites-of-service. We believe it would be more responsible and fair to allow the medical specialties to participate collectively in the needed revisions as part of the refinement process. The deferral of the CPEP revisions is in no way a reflection on the effort and thought that the commenters obviously expended in arriving at their recommendations. All the code- specific comments referred to above will be considered at the start of the refinement period. (See Section II.A.4, Refinement of Practice Expense RVUs) Comment: Many organizations, representing both surgical and primary care specialties, expressed concern that we averaged CPEP data for the same procedures valued by more than one CPEP. Different rationales were offered for this concern: Averaging could have disturbed the relative rankings of codes within CPEPs. Straight averaging significantly overstated the costs of evaluation and management services. Averaging CPEP costs altered practice expense relationships within the evaluation and management family of services, particularly with respect to emergency department evaluation and management codes. The inclusion of estimates from those not performing the procedures, including nonphysicians, could have distorted the values for those services. Likewise, different solutions were offered to answer the concerns: One specialty society recommended that we link the CPEP data rather than relying on straight averages. Two organizations recommended using frequency-weighted averages. Five groups recommended that the CPEP costs for redundant codes be based on the inputs from the dominant specialty's CPEP panel. Response: As we are making no other changes in the CPEP data for this final rule, we will continue to use straight averaging for the redundant CPEP codes for the purposes of this final rule. This issue will be considered further during refinement. Comment: Two commenters requested the inclusion in practice expense of the procedure-related supplies which are brought into a skilled nursing facility (SNF). One of these commenters made the same request for home visits. Response: Home visits are to be paid using the non-facility RVUs. Therefore, any supplies that would be used are already included in the payment. As for the SNF setting, this is an issue for refinement. We would need more information about the supplies and why the SNF is not responsible for providing them. Comment: The American College of Surgeons sent a list of new crosswalked codes where CPEP data had inadvertently been duplicated in our database. Response: We thank the commenter for pointing out this discrepancy, and these duplications have been deleted. Physician Time Comment: One major specialty society recommended that efforts be undertaken to move toward greater consistency in physician time data. The commenter was concerned that since these data are derived from eight different sources using different methodologies, our inflation of the Harvard time data raises even more concern about consistency. Three major organizations, two representing primary care and the other a surgical specialty, recommended that we use the unadjusted Harvard and RUC survey data. One reason given was the implication for the work RVUs of any proposed revisions to the time data. The RUC commented that, while the RUC physician time data may be greater than Harvard time data for the same codes, it may be incorrect to assume that all Harvard time data should be increased. The RUC and several other organizations requested that we provide a description of the methodology we used to make adjustments to the data in both the RUC and Harvard physician time databases so they can comment on the validity of the changes. Response: The physician time data used for the development of the practice expense pools are based on the Harvard resource-based RVUs study and RUC survey data that were developed as part of the refinement of the work RVUs. Both sets of data were based on physician surveys. However, the RUC data, gathered in the process of refining the work values of many CPT codes, are more current and, on average, exceeded the original Harvard values by 25 percent. As a matter of consistency and fairness to those services not yet refined by the RUC, we increased the Harvard time data in proportion to the increases for related services. A detailed description of the methodology we employed to make all adjustments in physician time will be placed on the HCFA Homepage. We still believe this adjustment is appropriate and we will continue to use the adjusted values in our calculations for this final rule. However, as the time values attributed to each procedure play an important role in the determination of each specialty's practice expense pool, we believe that ensuring the increased accuracy and consistency of physician time data should be addressed as part of the refinement of the practice expense RVUs. Comment: Three surgical specialty societies commented that evaluation and management times have been artificially inflated due to rounding. A small increase in time would disproportionately inflate high volume procedures that take little time. Response: In our proposed rule, we expressed concern that imprecision in the time estimates for any high volume services that have relatively little time associated with them may potentially bias the practice expense methodology in favor of the specialties that perform these services. We stated at that time that this issue should be examined as [[Page 58824]] part of the refinement of the resource-based practice expense RVUs. Comment: There were several other comments regarding the accuracy of the physician time data: The RUC acknowledged that some of the RUC physician time data may not be absolutely precise. One specialty society, as well as the AMA, pointed out that there are some problems with the accuracy of the physician time data for psychotherapy services. For example, the times assigned to psychotherapy codes that include evaluation and management services are equal to and, in some cases, less than the psychotherapy codes that do not include these services. One commenter stated that the physician time data, as computed in the Harvard studies, are not current and are likely to be inappropriate for use in computing practice expense RVUs. The American College of Surgeons commented that physician time for pediatric surgery codes is based on erroneously low physician time data from the original Harvard study, rather than the time data from the special study of pediatric services performed by the same Harvard study team for the American Pediatric Surgical Association in 1992. The latter data were used as the basis for the work RVUs assigned to 48 pediatric surgical services. A surgical specialty society commented that the physician time does not compensate its members for longer hours and cited examples of nonbillable time, such as standby time for cardiac catheterization and supervision of residents and interns. The society suggested that this be considered during refinement. One commenter stated that travel time for home visits is not included in either the work or practice expense RVUs. The commenter suggested that travel time for house calls should be equal to the work equivalent of the lowest office service times 3, for an average of 15 minutes. Further, a modifier should be used to cover instances where travel exceeds the average. The American Society of Transplant Surgeons identified physician times for several services that it believes are inaccurate and recommended adjusted times for these services. Response: As stated above, we will ensure that all identified anomalies and inaccuracies in the physician time data are considered as part of the refinement process. Comment: The American College of Radiology commented that for our top-down approach we had used a level three office visit (99213) as a benchmark for estimating physician time for radiology codes. They suggested that it would be more appropriate to use the intravenous pyelography procedure (CPT 74400) instead of the office visit used in our methodology. Response: Although we agree that 99213 may be an inappropriate benchmark since it is not often performed by radiologists, we are not convinced that the average work per unit time of codes on the radiology fee schedule is equivalent to CPT 74400. Instead, we are using the weighted average work per unit time for CPT 71010 and 71020 as the benchmark. These two services represent over approximately one-third of the total allowed services in the radiology fee schedule, while CPT 74400 represents less than two-tenths of one percent. We will work with the medical community to develop time estimates for radiology procedures that will make the imputation of time from the work estimates unnecessary. Comments: The American Occupational Therapy Association commented that the practice expense pool for occupational therapy codes was understated because the time values of 15 minutes that we arbitrarily assigned were too low. They included a list of time values we should use for each code. The American Hospital Association also objected to the reductions in times for outpatient rehabilitation codes and urged the use of the actual surveyed times for all procedure codes in the range 97001 through 97770. Response: We believe that the time of 15 minutes we assigned to these codes is appropriate and does not lead to an underestimation of the practice expense pool for outpatient rehabilitation services. The outpatient rehabilitation codes in this range are timed codes and are billed in 15 minute increments. Also, we have been told by some physical therapy associations that at times, some of the 15 minute period time may be performed by therapy aides or assistants. (Note: We plan to review this issue during a future five-year review of work RVUs.) Finally, it is common for these timed codes to be billed in multiple units during one therapy session. Thus, any therapist's work prior to or after the visit is spread across more than one unit, rather than applied to each unit. Crosswalk Issues Comment: The American Academy of Maxillofacial Prosthetics (AAMP) and the American College of Prosthodontists commented that crosswalking is not valid for maxillofacial prosthetic codes since this specialty does not correspond to any other medical specialty included in the SMS data and its practice expense values are much higher than other medical specialties in the SMS survey. AAMP submitted several studies from its own organization and from the American Dental Association, as well as two studies published in professional journals that report the results of polls of prosthodontic practitioners, including information on overhead expenses. The AAMP recommended that this data be used to calculate its practice expense per hour. Response: We agree that maxillofacial prosthetics does not correspond closely with any other medical specialty. It also is not a separately-identified specialty in either the SMS survey or the Medicare claims database. Though the AAMP submitted survey data compiled by both its own organization and the American Dental Association, the format, definitions, and methodology of these surveys were not consistent with those of the SMS survey. For example, the 1993 AAMP survey did not survey practice expense, but rather the ``percent overhead of gross collections for 1992.'' The American Dental Association surveys counted dentist shareholder and employee dentist income as practice expense in many tabulations. Because of these methodological differences from the SMS data, we are not able at this time to use the information in the submitted surveys to calculate a comparable practice expense per hour for maxillofacial prosthetics. For this final rule we will create a practice expense pool for the maxillofacial prosthetic codes (CPT 21076 through 21087) and crosswalk this pool to the practice expense per hour for ``All Physicians.'' We had imputed physician times for these services in our proposed rule. However, we are now using the physician times utilized in calculating the work RVUs for the same services. In addition, until the CPEP data for these codes can be validated, we will allocate the practice expense pool to the specific services using the current RVUs. We hope to work with the specialty society as part of the refinement process in order to develop a reliable method of deriving accurate practice expense RVUs for maxillofacial prosthetics. Comment: The American Optometric Association (AOA) disagreed with our crosswalk of optometry to the average practice expense per hour for ``All Physicians,'' that results in a practice expense per hour of $67.50. The commenter stated that AOA understands that the crosswalk decision [[Page 58825]] was based, at least in part, on the 1997 survey conducted by AOA which had been provided to us. This survey has been conducted regularly since 1990 and was included with the comment, along with a study commissioned by the AOA entitled ``Results of the First National Census of Optometrists.'' Using data from this survey and study, AOA computed an $89.53 practice expense per hour for optometry, significantly higher than the average for ``All Physicians.'' Response: As in the above request, the data submitted by AOA are not easily comparable to the SMS data. For example, the AOA calculation used medians rather than means, and retirement and fringe benefits were not counted as median net income, but rather as practice expense. It is therefore not possible, without further information, consultation, and analysis, for us to calculate a practice expense per hour that would be comparable with that of other specialties. During the refinement period we will be working with specialties not represented in the SMS survey to identify the data needed to enable us to determine accurate practice expense RVUs for their services. Comment: Although generally supporting the crosswalk to General Internal Medicine, the American Chiropractic Association (ACA) submitted data from the 1997 survey results of ACA's biannual survey of the chiropractic profession. This survey shows considerably lower direct patient care hours than SMS shows for General Internists. Therefore, the ACA requested that we use its data to calculate the practice expense per hour for Doctors of Chiropractic, stating that we should accept specialty societies' data over SMS data if they were collected in a comparable manner. Response: The survey submitted by the commenter indicated that the patient care hours worked by chiropractors are significantly lower than those of general internists to whom chiropractors' practice expense per hour is crosswalked. However, the hours of direct patient care a week shown in the survey were defined more narrowly than in the SMS data. For example, the 29 hours of patient care a week calculated in the submitted survey did not include the hours spent for documentation, administration, and billing, activities that we have considered to be included in the direct patient care hours for other specialties. In addition, there are insufficient details in the survey for us to determine its comparability to the SMS data and we will maintain the crosswalk for chiropractors for this final rule. We do intend, however, to revisit this issue during the refinement process. Comment: The American Podiatric Medical Association, Inc. (APMA) objected to its crosswalk to general surgery because it believes that there is little similarity between the two specialties based on site- of-service and types of services provided. General surgery services are typically performed in the facility setting, while the high volume podiatry services are almost entirely done in the office. In addition, the comment stated that podiatrists work fewer hours than general surgeons. The comment also included the results from APMA's 1996 and 1998 surveys of podiatric practice, as well copies of the surveys themselves. According to the comment, these surveys show that the actual practice expense per hour for podiatry is $91.50 and APMA recommends that we use this data in place of our proposed crosswalk. The American Academy of Orthopaedic Surgeons also disagreed with the crosswalk for podiatry, but recommended that podiatry be crosswalked to orthopaedic surgery in the short run, as 70 percent of the codes billed by podiatrists are those that are shared with orthopaedic surgery. Response: Because of significant methodological differences between the submitted surveys and the SMS data (for example, only gross and net incomes are surveyed) we are not able at this time to calculate a practice expense per hour in total, let alone for each of the different cost pools. However, we are persuaded that the crosswalk to general surgery is not appropriate for the reasons cited in the comment, and we are changing the crosswalk to ``All Physicians.'' We will be working with all specialties not represented in the SMS data to ensure that we obtain comparable information to calculate their practice expenses per hour. Comment: The Joint Council of Allergy, Asthma, and Immunology stated that, in calculating the allergists' practice expense per hour, we reduced the supply category practice expense per hour to that of ``All Physicians,'' because we believed that we made a separate payment for the drugs used. However, this is not true for immunotherapy drugs provided by allergists, as the cost of these drugs is included in the practice expense RVUs. Therefore an adjustment needs to be made. Response: The commenter is correct and the adjustment has been made to the medical supplies practice expense per hour. Comment: The American Society of Clinical Oncology commented that since the SMS supply cost data for chemotherapy codes included the costs of expensive chemotherapy drugs, which are paid for separately, we used the lower supply costs for ``All Physicians'' for their supply cost pool. The commenter argued that this fails to recognize that, in addition to the cost of the drugs, chemotherapy administration has extra supply costs in excess of that for ``All Physicians.'' Also, although chemotherapy drugs are generally among the costliest drugs, the cost of drugs was probably included in other specialties' supply costs as well, and all specialties should be treated in the same manner. The Association of Community Cancer Centers, the Society of Gynecologic Oncologists, and the American Society of Hematology also disagreed with our adjustment for drug costs, as did the AMA, which called our method of correcting for the double counting of drugs inequitable and imprecise. The American Society of Hematology recommended increasing the supply per hour costs to 125 percent of the ``All Physicians'' level. Response: It is true that other specialties may have some drug costs included in their SMS supply cost data, but we believe that the total costs for chemotherapy drugs are far greater than are the drug costs included for any other specialty. Failure to make an adjustment for these high drug costs would lead to a gross distortion in the supply cost pool for oncology. We also are not convinced that the other supply costs for oncologists would necessarily exceed that of ``All Physicians,'' and we will continue to crosswalk oncology's supply costs to that category's practice expense per hour. We do agree that during refinement we need to consider development of a methodology for removing separately billable supplies and services from the SMS data so that the Medicare program avoids making duplicate payments. We also will work with the oncology specialty to ensure that their practice expense per hour for the supply category adequately reflects the actual costs of other oncology supplies. Comment: The American Association of Oral and Maxillofacial Surgeons objected to the crosswalk of oral surgery and maxillofacial surgery to the practice expense per hour of ``All Physicians.'' They recommended a crosswalk to either otolaryngology or plastic surgery, as most of the medical procedures billed [[Page 58826]] by oral and maxillofacial surgeons can be crosswalked to these two specialties. The commenter argued that because of their significantly higher practice expenses, oral and maxillofacial surgery should not be in the same practice expense pool as manipulative therapists and optometrists, as this dilutes the practice expenses for these surgical services. In addition, the 1996 Harvard Study grouped oral and maxillofacial surgery under otolaryngology and plastic surgery. Response: We do not currently have sufficient data to make such a change in our crosswalk. This is an issue that can be addressed during the refinement period. Comment: The American College of Cardiology and the American Society of Echocardiography disagreed with the crosswalk of Independent Physiologic Laboratories (IPLs) to ``All Physicians.'' The comment recommended that IPLs' practice expense per hour be crosswalked to cardiologists, as 60 percent of IPL billings are in the 93000 series and for the 13 highest volume IPL codes, cardiologists account for 40 percent of claims. The Society of Vascular Technology/Society of Diagnostic Medical Sonographers also expressed concern that our crosswalk of IPLs did not adequately recognize their costs and recommended that we use the figure of $176 per hour based on the studies cited in the comment. Response: As discussed above, we will be creating a separate practice expense pool for all services without physician work, which will include those technical component services done by IPLs and by cardiologists. Comment: The Society of Gynecologic Oncologists requested that we consider using multiple crosswalks to determine practice expense per hour for specialties that provide interdisciplinary care. The comment stated that the true reflection of practice expense per hour for a gynecologic oncologist is a hybrid of the practice expense per hour for the specialties of obstetrics and gynecology and oncology. Response: It is not clear whether this is desirable or what data would be used to weight such a split between more than one specialty. Many physicians belong to more than one specialty or subspecialty. This is another issue that can be discussed during the refinement period. Comment: The American Geriatrics Society disagreed with our crosswalk of geriatrics to the General Internal Medicine practice expense per hour. The comment stated that geriatricians typically have higher costs than internists because of the need for more office space and more health care professionals on staff. Since many geriatricians are family physicians, geriatrics should be cross-walked to family practice. Response: We believe that geriatricians are typically more like internists than family practitioners, so for the final rule we will not change the crosswalk. However, we are open to receiving data that would demonstrate that a crosswalk to family practice would be more appropriate. However, we would note that geriatrics is a relatively small specialty and the services performed by them are frequently done by other specialties. Thus, changes in the practice expense per hour data for geriatricians would not likely have a significant impact on the RVUs for services they perform. Comment: One commenter made recommendations for revisions or additions to our proposed crosswalks for several nursing subspecialties. Another specialty society commented that under the physician fee schedule we have chosen to pay nonphysician practitioners a percentage of the physician reimbursement, and crosswalking to specialties with higher practice expense per hour rates than general internal medicine or general surgery is not logical or reasonable. Another organization also recommended that data from nurse practitioners and physician assistants be excluded from the practice expense pool calculations. Response: We will further consider appropriate crosswalks for nursing subspecialties during the refinement period. Comment: The American Hospital Association and the American Occupational Therapy Association recommended that we crosswalk all of the practice expense pools for outpatient rehabilitation services to the ``All Physicians'' practice expense category, rather than using the salary equivalency guidelines for the administrative, office, and other pool. Response: We believe that using the ``All Physicians'' practice expense per hour for the administrative, office, and other pool would considerably overstate the actual practice expense for occupational therapy. We have carefully examined outpatient therapy practice costs for the development of the salary equivalency guidelines, and believe that these better approximate the actual expenses for this cost pool. We will continue to use the salary equivalency guidelines to calculate this portion of the practice expense pool for occupational therapy for this final rule. Comment: The American Speech-Language Hearing Association commented that it is not appropriate to use the practice expense per hour data from physicians that perform audiology tests and it submitted a 1993 survey, ``Audiology Services--Scale of Relative Work,'' as part of its comments. Response: As we stated above, we are creating a single practice expense pool for all services, such as audiology, that have no work RVUs. This practice expense pool, created by using the average clinical staff time per procedure from the CPEP data and the ``All Physicians'' practice expense per hour, raises practice expense RVUs for audiology services relative to those previously proposed. However, during the refinement process we will be considering all data submitted on any of these services, including the study submitted with the above comment. Calculation of Practice Expense Pools--Other Issues Comment: Several organizations commented on potential problems with the Medicare claims data, which are used as one component of the specialty-specific practice expense pool calculation. Many commenters were concerned about reliance on Medicare claims data to determine the size of each specialty's practice expense pool. The comments claimed that to the extent that the Medicare population is not representative of the general population, there is a bias against specialties whose patient population does not match Medicare's. Several organizations, representing the gamut of medical specialties, urged us to work during the refinement period with organizations for whom we have no, or inadequate, historical claims utilization information and to acquire nationally representative claims data that include Medicare, Medicaid, and private payer data. One of these commenters recommended that, if this is not feasible, we should conduct sensitivity analyses to explore the influence Medicare service utilization patterns may have on private payers. The specialty-specific utilization data are crucial for the final step of volume-weighted averaging that brings the individual specialty scales onto one scale, particularly when involving services performed very frequently by specialties that see relatively few Medicare patients. For example, the comment argued, to the extent that the cost estimates for evaluation and management (E&M) services provided by obstetricians and gynecologists and pediatricians differ [[Page 58827]] significantly from those of specialties that account for the bulk of E&M services provided to Medicare patients, the use of an all-payer claims database would probably yield different RVUs for E&M services. Several surgical specialties urged that we clean the Medicare claims data to eliminate obvious errors, such as data showing a sometimes significant number of nonsurgeons or physician assistants performing complex surgeries that can only be performed by surgical specialties. This misreporting can decrease a specialty's practice expense pool and should either be reassigned or excluded during refinement. One of the commenters recommended that Medicare claims data be reviewed for the existence of a second listed surgical specialty identifier. In addition, physician assistants' claims should use the - AS modifier, and calculations should use only the time that is assigned to the intraoperative period. Three specialty organizations commented that many physicians' self-designated specialties are incorrectly classified in our claims data. For example, many cardiologists and geriatricians may bill as internists, which may affect the respective practice expense pools. Until these data become more accurate, one of the commenters recommended that the specialty practice expense pools be recalculated on an annual basis. An organization representing transplant surgeons commented that, as transplant surgery is not a designated specialty in the Medicare claims database, many transplant surgeons designate themselves as general surgeons, who have the lowest practice expense per hour of any surgical specialty. The comment argued that this has led to a significant underestimation of the costs associated with transplant surgery. Response: We would be interested in receiving any reliable national utilization data on the procedure code level though, to date, we are not aware of the existence of such a data source. We plan during the refinement period to work with the medical community in order to pinpoint problems in the Medicare claims data, to develop strategies to improve their accuracy, and, if possible, to find reliable supplemental data for those specialties not appropriately represented in the Medicare database. Comment: One organization commented that the Medicare frequency numbers for occupational therapy codes will be understated because BBA requires that all outpatient therapy services be paid under the Medicare Physician Fee Schedule beginning January 1, 1999. Response: We disagree. We have not included estimates for frequencies of expected services of outpatient therapy services in computing the practice expense RVUs. BBA specified that we pay for these services using the physician fee schedule. BBA did not incorporate these services into the fee schedule. Comment: Many organizations representing radiation oncology, as well as numerous individual commenters, argued that we erroneously combined the SMS radiation oncology survey data with that of radiology. The commenters argued that these two specialties should be dealt with separately, as radiation oncology utilizes different codes and has considerably higher costs than radiology. Response: We had combined radiation oncology and radiology together into one practice expense pool because of the small sample of radiation oncologists in the SMS data. However, we now agree with the commenters that these are two different specialties with differing practice costs. Therefore, we have separated them into two separate practice expense cost pools in order to calculate the practice expense per hour for each of the specialties. For radiology, excluding radiation oncology, the total practice expense per hour is $55.90. This is comprised of $17.90 for nonphysician payroll per hour ($9.70 for clerical payroll), $12.80 for office expense, $4.50 for supply expenses, $7.70 for equipment expense, and $12.90 for other expenses. For radiation oncology, the total practice expense per hour is $68.30. This is comprised of $23.70 for nonphysician payroll per hour ($9.20 for clerical payroll), $11.30 for office expense, $6.20 for supplies expense, $11.00 for equipment expense, and $16.20 for other expenses. Allocation of Practice Expense Pools to Codes Comment: Several organizations commented on our use of work RVUs as part of the allocation formula for indirect practice expense costs: A primary care specialty group stated that we should not allocate the indirect practice expenses using the work RVUs, since there is no reason to believe that the costs of providing the service, such as the cost of utilities, would vary by the intensity, where the costs would vary by time. We should, therefore, use time rather than work in our indirect allocation. Another primary care organization commented that using work as one allocator for indirect expenses inappropriately gives surgical procedures with higher work RVUs substantially higher administrative costs for billing activities than is given to evaluation and management services. We should develop a standardized method to address administrative staff costs. Five other organizations argued that allocating indirect costs based on a combination of direct costs and physician work RVUs is inappropriate and treats unfairly chemotherapy and radiation oncology services as well as other technical component services, since they typically are assigned no work RVUs. Various recommendations were made by these commenters to rectify what they see as discrimination against these technical component services: + Indirect costs should be based on direct costs. + Physician time or clinical staff time should be used instead of work. + We could allocate 50 percent of the indirect costs based on direct costs and 50 percent based on physician work or time. + As an alternative for chemotherapy services, work could be imputed by using the work to time ratio for other hematology or evaluation and management services. One commenter recommended that we vary the indirect cost allocation methodology in recognition of the practice patterns of particular specialties. One accounting organization commented that the use of work REUS is arbitrary and argued for the use of total dollars actually spent to perform the procedures, not indirect splits, suggesting the use of Activity Based Costing as a preferable methodology. Response: In this final rule, we will use an allocation method for the final rule that is basically similar to our proposed allocation method. It is widely recognized by accountants and others that there is no single best method of allocating indirect expenses to individual services. If we used physician time as an allocator of indirect expenses, we would be using the same values, whose accuracy have already been questioned by some commenters, both to create the practice expense pools and to allocate these pools to individual services. If we used only direct costs, we would be giving full weight to CPEP values that have not yet been refined. We agree that the use of physician work as an allocator is not preferable in the long term. It likely provides maximum advantage to hospital-based services in which the [[Page 58828]] physician incurs relatively few direct costs. For this final rule, we are making a technical change to the allocation method for indirect costs by using direct costs and the work REUS scaled using the Medicare conversion factor instead of a factor calculated using the physician time data. Because of questions raised by commenters concerning the time data adjustments, we believe that it is more appropriate to convert the work REUS into dollars using the Medicare conversion factor (expressed in 1995 dollars, consistent with the AMA SMS survey data). This will give somewhat less weight to work while, at the same time, avoiding a major methodological change until it has been examined further. We intend to work with the medical community during refinement so that we ensure that our allocation methodology is both appropriate and equitable. Comment: Many major specialty societies, both primary care and surgical, commented that we should not apply a different methodology for allocating the practice expense pools to the radiology codes than we do to all other codes. One commenter argued that multiplying the current charge-based practice expense RVUs for radiology codes by some percentage cannot yield a resource-based system. Organizations representing urologists, pulmonologists, cardiologists, and ophthalmologists commented that the uniform reductions made in the radiology codes to maintain relative values assumed that all radiology services are done only by radiologists, when many of these procedures are performed by these other specialties. A commenter stated that decisions regarding the practice expense values for radiology codes done predominantly by other specialists should not be made by one specialty. These organizations recommended that the practice expense RVUs for their codes be established using the allocation methodology used for all other services. One specialty society, representing diagnostic vascular testing, commented that the use of the existing radiology relatives to allocate practice expense to the code level results in significantly larger decreases in the technical component than in the professional component of their services. The commenter recommended that if we continue to use the radiology relatives, then we should reduce the professional components of the codes more than the technical components because practice expenses are greater for the technical component than for the professional component. The AMA supported the use of the radiology relative values for actual radiology services, but recommended that this methodology should be applied only to services that are performed predominantly by radiologists. The American College of Radiology endorsed the radiology relativity of the radiology RVUs without exception, and they would oppose the exclusion of individual radiology procedures since this is inconsistent with the concept of radiology relative values. They argued that maintaining the relativity of the radiology fee schedule-- Is consistent with generally accepted accounting principles because it is based on surveys and physician panels; Is widely accepted; Solves rank order anomalies caused by raw CPEP data; Simplifies the derivation of the professional component, technical component, and global practice expense RVUs; Is mandated by law, as the Omnibus Budget Reconciliation Act of 1989 stated that for radiology services ``the Secretary shall base the relative values on the relative values developed under section 1395m(b)(1)(A)****''; and They also argue that we have recognized and honored the statutory obligation to maintain the relationships in the radiology relative value scale. Another national organization representing diagnostic imaging services also suggested keeping the radiology fee schedule as the allocator for radiology, rather than the direct costs from the CPEP data, as there would be even greater reductions on codes we allocated using the CPEP relatives. Response: Because the majority of specialties that perform radiology services object to the use of the current practice expense RVUs for radiology services, we cannot continue to use these RVUs. However, since we are not making changes to the CPEP data for this final rule and since the American College of Radiology has not had sufficient opportunity to comment on the CPEP data because of our proposed use of the current radiology RVUs, we are using the current radiology RVUs to allocate the direct cost pools of the specialty radiology until such time as the CPEP data for radiology services have been validated. We will not use the current radiology RVUs for any other specialty. It should be noted that radiology services or components of radiology services that lack work RVUs are handled as described in the section on services without work RVUs. This alters the impact of using the current radiology RVUs for the specialty radiology since we set the global portion of a radiology service equal to the sum of the technical and professional components. Comment: One specialty society commented that, for one important high volume pathology service, the proposed total professional component practice expense RVU payment would be $11.37, approximately $2 short of the administrative labor costs alone. The commenter wanted more information on how our method splits administrative costs between the professional and technical components. The commenter requested that we provide a data set of the RVUs for administrative labor, office expenses, and other expenses that result from our allocation method, with a break-out of the professional and technical component RVUs for services that have both components, so that the appropriateness of the allocation method can be evaluated. Response: Our methodology was described in the proposal, and we also provided additional detailed data files that we used to develop the proposed values. We will try to make additional data available if the request is further specified. Comment: The American College of Cardiology expressed concern that, though it might be necessary to weight average the allocation to codes according to the practice expense per hour of the different specialties performing the service, this defeats the intent of Congress to recognize actual costs and could also lead to negative incentives. The commenter suggested that this is an issue that we and the specialties should pursue. The American Society of Echocardiography more specifically commented that we should not include in the calculations for cardiovascular diagnostic tests the even more unrepresentative data for internists coding for these procedures. The society maintained that because of the low equipment costs for internists, this blend dilutes the RVUs allocated to these codes. Response: The statute is very specific that Medicare is not to pay specialty differentials. Therefore, weight averaging of the CPEP inputs among specialties that do a service seems appropriate. Other Issues Comment: Many commenters, representing a broad spectrum of specialties, expressed concern that reductions in payment for specific services could have a negative impact on access to care. Many of these [[Page 58829]] commenters recommended that we monitor access and quality of care issues that may arise as a result of the implementation of a resource- based practice expense system. Response: Maintaining access to high quality health care for Medicare beneficiaries is, and will continue to be, a high priority, and we will monitor available relevant data. However, we do not anticipate that the implementation of resource-based practice expense RVUs should lead to any major impediments to access to care. Any impacts of this new system are being transitioned in over a 4-year period, during which we will be refining both the practice expense per hour data and the direct cost inputs. We will be working closely with the medical community during this refinement period, and we are confident that we will achieve a resource-based practice expense system that will maintain our beneficiaries' access to the best possible medical care. Comment: One commenter was concerned about how the monthly capitated payment for end-stage renal disease (ESRD) services was handled under the top-down approach. The commenter argued that, though the ``building block'' process used for the work RVUs for these services does not translate perfectly for practice expense values, this approach should still be utilized to calculate the practice expense RVUs. In addition, the commenter questioned our choice of CPT 99213, a mid-level office visit, to calculate physician time for ESRD services. Response: We allocated the practice expense pool to ESRD services using the CPEP inputs, as we did for almost all other services. We also believe that the intensity of an average evaluation and management service provides a reasonable estimate of physician time. These issues can be further analyzed during refinement. Comment: Two commenters noted that costs associated with the supervision of diagnostic tests were not included in the technical component amounts. Response: In separate carrier manual instructions, we are revising the level of physician supervision required for many diagnostic services. For example, we are changing the requirements for most ultrasound procedures from personal or direct supervision to general supervision. We believe the required supervision for any remaining services that are at the personal supervision level are generally already reflected in the work RVUs. Therefore, we do not believe that there are additional costs for physician supervision. Comment: One commenter indicated that there will be a marked increase in the volume of services paid under the physician fee schedule as a result of BBA changes in payment for outpatient therapy services. The commenter maintained that this increase should not adversely affect future budget neutrality adjustments. Response: Although payment for these outpatient therapy services are based on payment amounts contained in the physician fee schedule, these services are not included as part of the fee schedule pool for budget neutrality calculations. Comment: One commenter argued that the budget neutrality adjustment is inappropriately applied because it does not recognize the savings provided by the elimination of the facility payments for endoscopic procedures that will move to the office setting. Response: The statute specifies that there shall be budget neutrality for physician fee schedule services. The budget neutrality adjustment does not take into account payments to facilities. Comment: Two commenters suggested that any fiscal adjustments made to comply with BBA should be reflected in the conversion factor, or other ratio, rather than be included in the calculation of the practice expense RVUs, so that other payer reimbursement would not be affected. Response: We do not completely understand these comments, but we believe the request is consistent with our practice of making budget- neutrality adjustments on the conversion factor. Comment: Several commenters requested additional impact analyses such as-- Comparison of actual practice expense by specialty with expected practice expense payments, both by amount and by percent, for both our proposed practice expense payments and the current fee schedule practice expense RVUs; Comparison of impacts by geographic area, including rural and urban impacts; Analysis of impacts on hospital, academic, and community- based physicians; Analysis of total Medicare and non-Medicare impact using national claims case mix data; and An analysis that would demonstrate to other payers the degree to which our proposed payment rates are less than actual practice costs. Response: We lack the data to provide some of the requested analyses. For example, we do not have national claims case mix data and are unaware of the existence of such data. With regard to rural and urban impacts, in the June 5, 1998 proposed rule we discussed the limitations of such analyses given the structure of the Medicare payment localities. We are unsure what the commenters are specifically requesting on the issue of actual costs since we have based the resource-based practice expense RVUs on the best available source of multi-specialty actual cost data: the SMS survey. Cost analyses at the individual practice level are problematic since, for example, we do not have physician cost reports, but we are open to concrete suggestions on how to perform such analyses. We also note that the Medicare public use files are an excellent source of data for commenters who wish to perform additional analyses that they believe are possible with the data sources available to us. Comment: One commenter requested that we make clear to Medicare contractors that hospital-based pathologists who incur technical component costs for nonhospital patients can be paid for both the technical and professional components. Response: This is a long-standing policy, and we are not aware of any general problems in this regard. However, we would be willing to discuss the issue with individual carriers if the commenter provides more specific information. Comment: One commenter recommended that we recalibrate the allocation of RVUs to the pools for physician work, practice expense and malpractice, as this allocation has remained constant since the resource-based relative value scale was implemented in 1992. Response: We are recalibrating the allocation this year to match the Medicare Economic Index (MEI) weights. For example, work goes from 54.2 percent of the total to 54.5 percent, the practice expense portion goes from 41.0 percent to 42.3 percent, and the malpractice portion goes from 4.8 percent to 3.2 percent. (See Section II.D, ``Rebasing and Revising the Medicare Economic Index.'') In order to prevent the work RVUs from changing as a result of this, we are altering only the practice expense and malpractice RVUs. The changes to the practice expense and malpractice RVUs due to this are offset by an adjustment to the conversion factor. Comment: One commenter recommended that we should limit the magnitude of the changes in physician payments resulting from the shift to resource-based payment for practice [[Page 58830]] expenses by imposing some reasonable limit on payment increases and decreases for individual services. The commenter maintains that section 1848(c)(4) of the Act, which authorizes the Secretary of Health and Human Services to, ``establish ancillary policies, as may be necessary to implement this section,'' provides statutory authority on which to base such a policy. The comment pointed out that we invoked this section in 1991 with reference to the transition to resource-based payment for physician work. Response: We believe that Congress intended the transition period to be the mechanism by which we would mitigate the impacts of any changes in payment brought about by the shift to resource-based practice expense. Therefore, we believe it would be inappropriate for us to impose further limits on payment increases or decreases. Comment: One commenter maintained that the proposal violates both the Regulatory Flexibility Act and the Paperwork Reduction Act of 1980 because the adequate filings required in both of these Acts did not accompany the proposal. Additionally, the commenter stated that we did not cite any evidence to support its contention that a Regulatory Impact Statement is not required. Response: We had included a Paperwork Reduction Act (PRA) section in HCFA-1006-P that meets the requirements of the PRA of 1980. One commenter stated that we do not cite any evidence in either of our proposals to support our contention that no regulatory impact statement is required. There may be some confusion about the purpose of an impact statement and the difference between a regulatory impact statement and a regulatory impact analysis (RIA). A regulatory impact statement is a brief rational on why an analysis was not conducted. An RIA is a complete analysis based on recent available data and is more extensive. An RIA was conducted in the proposed rule of June 5, 1998 (63 FR 30866). Absent this analysis, we would be required to furnish an impact statement. Therefore, there is no violation of either the RIA or Regulatory Flexibility Act requirements. 3. Other Practice Expense Policies Site-of-Service Payment Differential As part of the resource-based practice expense initiative, we are replacing the current policy that systematically reduces the practice expense RVU by 50 percent for certain procedures performed in facilities with a policy that would generally identify two different levels (facility and nonfacility) of practice expense RVUs for each procedure code depending on the site-of-service. Some services, by the nature of their codes, are performed only in certain settings and will have only one level of practice expense RVU per code. Many of these are evaluation and management codes with code descriptions specific as to the site of service. Other services, such as most major surgical services with a 90-day global period, are performed entirely or almost entirely in the hospital, and we are generally providing a practice expense RVU only for the out-of-office or facility setting. In the majority of cases, however, we will provide both facility and nonfacility practice expense RVUs. The higher nonfacility practice expense RVUs are generally used to calculate payments for services performed in a physician's office and for services furnished to a patient in the patient's home, or facility or institution other than a hospital, skilled nursing facility (SNF), or ambulatory surgical center (ASC). For these services, the physician typically bears the cost of resources, such as labor, medical supplies, and medical equipment associated with the physician's service. The lower facility practice expense RVUs generally are used to calculate payments for physicians' services furnished to hospital, SNF, and ASC patients. The costs for nonphysicians' services and other items, including medical equipment and supplies, are typically borne by the hospital, by the SNF, or the ASC. We received the following comments on our site-of-service payment differential proposal. Comment: We received several comments concerning the appropriateness of our site-of-service proposal: Several specialty groups commented that they agreed with eliminating the site-of-service differential and replacing it with two levels of payment. A national specialty society representing gastroenterologists, as well as several hundred individual commenters, strongly opposed the elimination of the current site-of-service differential and replacement of it with the facility and nonfacility resource-based practice expense RVUs. The comments argued that we should not have established different practice expense RVUs for facility and nonfacility settings for gastrointestinal endoscopy codes 43234 through 45385 because: It is unsafe to do these procedures in the office and will thus jeopardize patient safety; It creates an incentive to provide care in the inappropriate office setting; and It is not authorized by legislation, is against the intent of BBA to have different payment levels for different settings, and is likely to result in legal challenge. The commenter recommended that we drop the office and out-of-office differential in practice expense payment. One organization commented that our site-of-service proposal will exacerbate the ability to subsidize uncompensated care and suggested exempting teaching physicians from the new site-of- service provision. It also suggested that HCFA should also monitor the effects of the site-of-service policy. The AMA, the American Hospital Association, and three other organizations commented that payment differentials should not provide an incentive for physicians and patients to choose one site over another. Some physician groups are concerned that the differential will accelerate the shift of some services from facility to nonfacility settings at the expense of patient safety. They asserted that claims data on changes in place of service should be made available and this issue should be one focus of refinement efforts. Response: We believe that, to the extent that the differing RVUs for in-office and out-of-office services reflect the relative differences in practice costs for performing those services, we have not created incentives to provide services in inappropriate settings. We are required by both the Social Security Act Amendments of 1994 and BBA to develop resource-based practice expense RVUs, based on physicians' actual costs. All of our data indicate that physicians' practice expenses are higher in the office, where the physician must incur all the costs of staff, equipment, and supplies, than in a facility that provides and is paid separately for these resources. As the facility and nonfacility costs to the physician can vary by a considerable amount, we believe that adopting a single average payment for both sites would consistently underpay in-office procedures, and overpay those performed in a facility and would thus be inherently inequitable, not resource-based, and contrary to the intent of the law. Furthermore, we are not aware of any studies showing that codes 43234 through 45385 are being unsafely performed in offices. We have complete [[Page 58831]] confidence that physicians will continue to exercise their best clinical judgment as to the most appropriate setting for their patients. Comment: One specialty society stated its support for the proposed change in the site-of-service payment, as long as it does not result in nonpayment for services actually provided. For example, there are no practice expense RVUs for emergency intubation in the nonfacility setting, though this service may occasionally have to be performed in the office. Response: If a service for which there are only facility RVUs is performed in the office, the facility rate will be paid. Comment: The American Urological Association commented that certain codes--50590, 52234, 52235, 52240, 52276, and 52317 were inappropriately assigned nonfacility PERVUs, as it is not safe to perform these services in the office. Response: We would need more data to demonstrate that performing these services in the office is not appropriate before we would eliminate the nonfacility RVUs. We are willing to review such information during the refinement process. Such information should be submitted to HCFA, Office of Clinical Standards and Quality. Comment: Two societies representing pulmonologists commented that critical care is listed with facility and nonfacility practice expense RVUs, although it is nearly always performed in an inpatient setting. One organization representing psychiatrists noted that CPT codes 90816 through 90829 are restricted to the inpatient hospital and partial hospital and residential care settings, and that CPT code 90870, electroconvulsive therapy, would not generally be performed in an office setting. The commenter recommended that the final rule list RVUs for only the facility setting. Response: We are not deleting RVUs proposed for the nonfacility setting in this final rule, but will be considering this issue during refinement. We would note, however, that services performed in the residential care setting would be paid by using the nonfacility RVUs. Comment: One commenter pointed out that in our proposed rule we list the services that, by nature of their codes, would only have one level of practice expense; this list includes codes 99321 through 99333 and 99341 through 99350. However, in Addendum C, both facility and nonfacility values are given and the facility values are higher than the nonfacility values for most of these codes. These inconsistencies should be corrected. Another commenter submitted a list of some codes where the facility practice expense RVUs are higher than the in-office values. Response: We thank the commenters for pointing out these discrepancies. The instances of higher facility RVUs are an artifact of our indirect methodology and reflect the differing mix of specialties performing a service in each setting. We will look at this more closely during the refinement process. Comment: One specialty society commented that the dual energy x-ray absorptiometry codes have the same practice expense RVUs for both the in-office and out-of-office setting. The comment recommended that the in-office RVUs be adjusted to reflect the high costs of equipment for the office-based physician. Response: More specific data will be needed on the actual costs of the equipment so that we can address any changes to the CPEP data during the refinement process. Comment: Three organizations representing outpatient therapy services commented that, though outpatient rehabilitation providers will be paid the nonfacility rate, there are higher costs for providing rehabilitation services in an SNF or hospital than in a doctor's office. These costs are not reflected in the CPEP data and are grossly underestimated in the practice expense RVUs. There should be a special higher site-of-service differential to be applied when outpatient therapy services are furnished in provider settings. Response: The site-of-service differential is intended to ensure that the Medicare program avoids making duplicate payme