Medical Technology and Costs of THE MEDICARE PROGRAM OTA Reports are the principal documentation of formal assessment projects. These projects are approved in advance by the Technology Assessment Board. At the con- clusion of a project, the Board has the opportunity to review the report, but its release does not necessarily imply endorsement of the results by the Board or its individual members. Recommended Citation: Medical Technology and Costs of the Medicare Program (Washington, D.C.: U.S. Congress, Office of Technology Assessment, OTA-H-227, July 1984). Library of Congress Catalog Card Number 84-601085 For sale by the Superintendent of Documents U.S. Government Printing Office, Washington, D.C. 20402 Foreword = — U.S. health care costs have escalated rapidly over the past 15 years, and medical technology is a primary cause of the increase. A major focus of Federal policymakers’ concerns about rising health care expenditures is the Medicare program, which pays for hospital and other acute-care health services for over 30 million elderly and disabled Americans. To aid in congressional efforts to contain Medicare costs, the House Committee on Energy and Commerce and its Subcommittee on Health and the Environment and the Senate Committee on Finance, Subcommittee on Health asked OTA to assess a broad range of mechanisms to limit or reduce Medicare costs related to medical technology without sacrificing the quality of health care delivered. In addition, the committees requested a study of the proposed use of Diagnosis Related Groups (DRGs) as Medicare's inpatient hospital payment mechanism and sev- eral case studies of particular interest to the Medicare program. These are published as separate volumes. This report focuses on the policy mechanisms to limit or reduce Medicare costs related to medical technology but draws from the study of DRGs and the case studies. This assessment explores the dual relationship between medical technology and the Medicare program: Medicare policies affect the adoption and use of medical technol- ogies, and the patterns and levels of use of medical technologies significantly affect Medi- care costs. It reviews specific Medicare policies—eligibility, benefits, payment, and ben- eficiary cost-sharing—that have had an influence on the adoption and use of medical technology. It also examines the contribution of medical technologies to increases in Medicare costs. The report identifies several possible changes in Medicare policies that could be used to influence medical technology adoption and use and to restrain Medicare pro- gram costs. These mechanisms fall into four general categories: Medicare's coverage policy for individual medical technologies; methods of Medicare payment to hospitals; methods of Medicare payment to physicians; incentives for the adoption and use of technology that do not directly involve, but may be related to, the Medicare payment mechanism (e.g., encouraging the development of alternative cost-effective health care delivery systems). This study was ably guided by an advisory panel, chaired by Stuart H. Altman. In addition, a large number of persons in the Federal and State Governments and in the health services research field were consulted. We are grateful for their many contri- butions. As with all OTA reports, however, the content is the responsibility of OTA and does not constitute consensus or endorsement by the advisory panel or by the Tech- nology Assessment Board. Key OTA staff involved in the assessment were Anne K. Burns, Cynthia P. King, Lawrence H. Miike, Gloria Ruby, and Judith L. Wagner. JOHN H. GIBBONS Director iii Advisory Panel on Medical Technology and Costs of the Medicare Program Stuart Altman, Panel Chair Dean, Florence Heller School, Brandeis University Frank Baker Vice President Washington State Hospital Association Robert Blendon Senior Vice President The Robert Wood Johnson Foundation Jerry Cromwell President Health Economics Research Chestnut Hill Mass. Karen Davis Johns Hopkins University School of Hygiene and Public Health Robert Derzon Vice President Lewin & Associates Washington, D.C. Howard Frazier Center for the Analysis of Health Practices Harvard School of Public Health Clifton Gaus Center for Health Policy Studies Washington, D.C. Jack Hadley Urban Institute Washington, D.C. Kate Ireland Chairman, Board of Governors Frontier Nursing Service Wendover, Ky. Judith Lave Graduate School of Public Health University of Pittsburgh iv Mary Marshall Delegate Virginia House of Delegates Walter McNerney Northwestern University Morton Miller President National Health Council New York, N.Y. James Mongan Executive Director Truman Medical Center Kansas City, Mo. Seymour Perry Institute for Health Policy Analysis Georgetown University Medical Center Robert Sigmond Blue Cross of Greater Philadelphia Anne Somers Department of Environment and Community and Family Medicine Rutgers University Paul Torrens UCLA School of Public Health Keith Weikel Group Vice President AMI McLean, Va. OTA Project Staff—Medical Technology and Costs of the Medicare Program H. David Banta* and Roger C. Herdman,** Assistant Director, OTA Health and Life Sciences Division Clyde J. Behney, Health Program Manager Anne Kesselman Burns, Project Director Cynthia P. King, Analyst Lawrence H. Miike, Senior Associate Gloria Ruby, Analyst Pamela J. Simerly, Research Assistant Judith L. Wagner, Senior Analyst Kerry Britten Kemp, Health and Life Sciences Division Editor Virginia Cwalina, Administrative Assistant Beckie I. Erickson, Secretary™** Brenda Miller, Word Processor/PC Specialist Jennifer Nelson, Secretary **** Mary Walls, Secretary™ Special Consultant Robert A. Berenson Principal Contractors Benjamin Barnes, Harvard University Montague Brown, Strategic Management Services Mark Chassin, Rand Corp. Gordon DeFriese, University of North Carolina Linda Demlo, University of Iowa Joanne Finley, Maryland State Department of Health Irwin Litt Matthew Menkin Leonard Saxe, Boston University William Stason, Harvard University Paul Wortman, University of Michigan *Until August 1983. **Since December 1983. ***Since January 1984. ****Until January 1984. Contents Chapter Page Glossary of ACTONYMNS . ........viviitertr rrr ares ausaraass ara aenn. Glossary of TeImMS . . . «oo out iii 1. Introduction and SUMMATY ...........uuuunnunnnnniiiiiiieaaaaaannnns Part One: 2. Medicare Policies Affecting Medical Technology .......................... 3. The Impact of Medical Technology on Medicare Costs .................... Part Two: 4. A Framework for Change... ............uuunnnnnnnnnnenaaaannns 5. Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs ......................ooiinnn 6. Medicare Hospital Payment and Medical Technology ..................... 7. Medicare Physician Payment and Medical Technology .................... 8. Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use. ................................. Part Three: 9. Findings and Policy Options ................oiiiiiiiiii Appendixes A. Method of the Study and Case Studies .............................. 0. B. Acknowledgments and Health Program Advisory. ........................ C. Selected Activities in Medical Technology Assessment .................... D. Selected Alternatives to Traditional Health Care Delivery ................. E. Decisionmaking by Medicare Contractors for Coverage of Medical Technologies ..................ccivivvvvervrrrrrrrrarssrrrres RCCrEIICES . «+ «vs vs 4s avmn san nes ann taney aE TE AEE ER may ns vii Glossary of Acronyms AAMC AAPCC ACC ACP ACR ACS ADAMHA AHA AMA BC/BS CABG CAPD CBA CBO CDC CEA CEA/CBA CEAP — Association of American Medical Colleges —average adjusted per capita cost —American College of Cardiology —American College of Physicians —American College of Radiology —American College of Surgeons —Alcohol, Drug Abuse and Mental Health Administration (Public Health Service) — American Hospital Association — American Medical Association —Blue Cross and Blue Shield Association —coronary artery bypass graft surgery —continuous ambulatory peritoneal dialysis —cost-benefit analysis —Congressional Budget Office (U.S. Congress) —C Centers for Disease Control (Public Health Service) —cost-effectiveness analysis —cost-effectiveness analysis/cost- benefit analysis —Clinical Efficacy Assessment Project (ACP) CHAMPUS —Civilian Health and Medical CFA CMSS CON CPI CT DATTA DHHS DRG ECRI ESP ESRD FDA GAO GPPP HCFA viii Program of the Uniformed Services —capital facilities allowance (U.S. Department of Defense) —Council of Medical Specialty Societies —certificate of need —Consumer Price Index —computed tomography scanner —Diagnostic and Therapeutic Technology Assessment (AMA) —Department of Health and Human Services —Diagnosis Related Group —formerly the Emergency Care Research Institute —economic stabilization program —end-stage renal disease —Food and Drug Administration (U.S. Department of Health and Human Services) —General Accounting Office (U.S. Congress) —group practice prepayment plans —Health Care Financing Administration (U.S. Department of Health and Human Services) HCPCS HI HIAA HIMA HMO ICF ICU IOM IPA JCAH KPMCP LOS MCR MDC NAFEC NCHCT NCHS NCHSR NIH NMCES NMR OASH OCP ODR OHTA OTA PHS PMC PPO PRO ProPAC —HCFA Common Procedure Coding System —Hospital Insurance (Part A) program —Health Insurance Association of America —Health Industry Manufacturers Association —health maintenance organization —intermediate care facility —intensive care unit —Institute of Medicine (National Academy of Sciences) —independent practice association —Joint Commission for the Accreditation of Hospitals —Kaiser-Permanente Medical Care Program —]length of stay —Medicare cost report —Major Diagnostic Category —National Association of Freestanding Emergency Centers —National Center for Health Care Technology (Public Health Service) —National Center for Health Statistics (Public Health Service) —National Center for Health Services Research (Public Health Service) —National Institutes of Health (Public Health Service) —National Medical Care Expenditures Survey —nuclear magnetic reasonance —Office of the Assistant Secretary for Health (Public Health Service) —Office of Coverage Policy (Health Care Financing Administration) —Office of Direct Reimbursement (Health Care Financing Administration) —Office of Health Technology Assessment (National Center for Health Services Research) —Office of Technology Assessment (U.S. Congress) —Public Health Service (U.S. Department of Health and Human Services) — patient management category —preferred provider organization —utilization and quality control peer review organization —Prospective Payment Assessment Commission PSRO — Professional Standards Review Organization PTCA — percutaneous transluminal coronary angioplasty QALY —quality-adjusted life-year RAHC —Rochester Area Hospitals’ Corporation SLE —systemic lupus erythymatosis SMI —Supplementary Medical Insurance (Part B) program Glossary of Terms Allowable costs: Hospital costs that are reimbursable under the Medicare program. Ancillary technology: Medical technology used directly to support clinical services, including diag- nostic radiology, radiation therapy, clinical labora- tory, and other special services. Assignment: An agreement by a physician to bill the Medicare program directly and to accept Medicare's reasonable charge as full payment for his or her services. If the physician does not accept assign- ment, the patient is billed by the physician and is responsible for the difference between what Medi- care will pay and what the doctor charges for a par- ticular service. Beneficiary cost-sharing: The general set of financing arrangements whereby the consumer must pay some out-of-pocket cost to receive care. (Also see coinsurance, copayment, deductible, and premium.) Budget neutrality: Specified by the Social Security Amendments of 1983 (Public Law 98-21) to mean that the aggregate payments for the operating costs of inpatient hospital services in fiscal years 1984 and 1985 will be neither more nor less than would have been paid under the Tax Equity and Fiscal Respon- sibility Act (Public Law 97-248) for the costs of the same services. Capital costs: Expenditures for capital plant and equip- ment used in providing a service. Under Medicare's prospective hospital payment system established by the Social Security Amendments of 1983 (Public Law 98-21), hospitals’ capital costs (depreciation, interest, and return on equity to for-profit institu- tions) are treated as pass-throughs (i.e., are not sub- ject to the new system's controls). Capitation: A method of paying for medical care on a fixed, periodic prepayment basis per individual. Payment by “capitation” implies that the amount paid by the individual is independent of the num- ber of services that individual has received. Case mix: The relative frequency of admissions of various types of patients, reflecting different needs SNF —skilled nursing facility SSOP —second surgical opinion program TEAM —Technology Evaluation and Acquisition Methods for Hospitals (AHA) TEFRA —Tax Equity and Fiscal Responsibility Act of 1982 (Public Law 97-248) TPN —total parenteral nutrition for hospital resources. There are many ways of measuring case mix, some based on patients’ diag- noses or the severity of their illnesses, some on the utilization of services, and some on the character- istics of the hospital or area in which it is located. Certificate of need (CON): A regulatory planning mechanism required by the National Health Plan- ning Resources Development Act of 1974 to con- trol large health care capital expenditures. Each State is required to enact a CON law. CON applica- tions by institutions are reviewed by local health systems agencies, which recommend approval or disapproval; they are denied or approved by State health planning and development agencies. Coinsurance: A form of beneficiary cost-sharing whereby the insured pays a percentage of the total cost of health services. Conditions of participation: Requirements that a pro- vider must meet in order to be allowed to receive payments for Medicare patients. An example is the requirement that hospitals conduct utilization review. Copayment: A form of beneficiary cost-sharing whereby the insured pays a specific amount at the point of consumption of health services, e.g., $10 per visit. Cost-benefit analysis (CBA): An analytical technique that compares the costs of a project or technologi- cal application to the resultant benefits, with both costs and benefits expressed by the same measure. This measure is nearly always monetary. Cost-effectiveness analysis (CEA): An analytical tech- nique that compares the costs of a project or of alternative projects to the resultant benefits, with costs and benefits/effectiveness expressed by dif- ferent measures. Costs are usually expressed in dollars, but benefits/effectiveness are ordinarily ex- pressed in terms such as “lives saved,” “disability avoided,” “quality-adjusted life years saved,” or any other relevant objectives. Also, when benefits/ef- fectiveness are difficult to express in a common ix metric, they may be presented as an “array.” CEA/CBA: A composite term referring to a family of analytical techniques that are employed to compare costs and benefits of programs or technologies. Lit- erally, the term as used in this assessment means “cost-effectiveness analysis/cost-benefit analysis.” Coverage: In the Medicare program, coverage refers to the benefits available to eligible beneficiaries, distinguished from payment which refers to the amount and methods of payment for covered services. Deductible: A form of beneficiary cost-sharing in which the insured incurs an initial expense of a spec- ified amount within a given time period (e.g., $250 per year) before the insurer assumes liability for any additional costs of covered services. Depreciation: An estimate of the value of consump- tion of a fixed asset during a specific period of time. Diagnosis Related Groups (DRGs): Groupings of diag- nostic categories drawn from the International Classification of Diseases and modified by the pres- ence of a surgical procedure, patient age, presence or absence of significant comorbidities or complica- tions, and other relevant criteria. DRGs are the case-mix measure mandated for Medicare's prospec- tive hospital payment system by the Social Secu- rity Amendments of 1983 (Public Law 98-21). DRG payment: The system of prospective payment for inpatient services by Medicare which was mandated by the Social Security Amendments of 1983. Effectiveness: Same as efficacy (see below) except that it refers to “. . . average or actual conditions of use. Efficacy: The probability of benefit to individuals in a defined population from a medical technology ap- plied for a given medical problem under ideal con- ditions of use. End-stage renal disease: Chronic renal failure that oc- curs when an individual irreversibly loses a suffi- cient amount of kidney function so that life cannot be sustained without treatment intervention. Hemo- dialysis and kidney transplant surgery are two forms of therapy. Fee-for-service: A method of paying for medical care on a retrospective basis by which each service ac- tually received by an individual bears a related charge. Fee schedules: Set amounts of payment to physicians for particular services, generally established by a regulatory body. Global budgeting: A method of hospital cost contain- ment in which participating hospitals must share a prospectively set budget. Methods for allocating funds among hospitals may vary. Health maintenance organization (HMO): A health care organization that acts as both insurer and pro- vider of comprehensive but specified medical serv- ices by a defined set of physicians to a voluntarily enrolled population paying a prospective per cap- ita fee (i.e., paying by “capitation”). Historical cost depreciation: An estimate of deprecia- tion (see definition) based on the original cost of the fixed asset. Inpatient care: Care that includes an overnight stay in a medical facility. Length of stay (LOS): The number of days a patient remains in the hospital from admission to discharge. Medical technology: The drugs, devices, and medical and surgical procedures used in medical care, and the organizational and supportive systems within which such care is provided. Medicare: A nationwide, federally administered health insurance program authorized in 1965 to cover the cost of hospitalization, medical care, and some related services for eligible persons over age 65, per- sons receiving Social Security Disability Insurance payments for 2 years, and persons with end-stage renal disease. Medicare consists of two separate but coordinated programs—Hospital Insurance (Part A) program and the Supplementary Medical Insurance (Part B) program. Health insurance protection is available to insured persons without regard to income. Medicare carriers: Medicare contractors that compute reasonable charges and make Medicare Part B pay- ments, determine whether claims are for covered services, deny claims for noncovered services, and deny claims for unnecessary use of services. Medicare contractors: Blue Cross/Blue Shield plans or commercial insurers that perform the Medicare pro- gram’s claims processing and payment functions at the local level under the policy and operational guidance of the Health Care Financing Administra- tion. (Also see Medicare carriers, Medicare in- termediaries.) Medicare Economic Index: The index that the Medi- care program uses to determine physicians’ prevail- ing charges, as specified by the Social Security Amendments of 1972 (Public Law 92-603). Specifi- cally, the prevailing charges are calculated by multiplying the 1973 prevailing charges by the cur- rent index, which is promulgated annually for the 12-month period beginning July 1. Medicare intermediaries: Medicare contractors that de- termine reasonable costs for covered items and serv- ices, make payment and guard against unnecessary use of covered services for Medicare Part A payments. Intermediaries also make payments for home health and outpatient hospital services cov- ered under Part B. Medicare vouchers: A proposed administrative change in the Medicare program in which each eligible per- son would be allowed a set amount of money to purchase medical care and/or health insurance. Medigap insurance: Private supplementary medical in- surance covering Medicare deductibles and co- insurance. Outliers: Cases with unusually high or low resource use. DRG outliers are defined by the Social Secu- rity Amendments of 1983 (Public Law 98-21) as atypical cases that have either an extremely long length of stay or extraordinarily high costs when compared to most discharges classified in the same DRG. Outpatient care: Care that does not include an over- night stay in the facility in which care is provided. Part A (Medicare): Medicare's Hospital Insurance pro- gram which covers specified inpatient services in hospitals, post-hospital extended care, and home health care services. Part A, which is an entitlement program for those who are eligible, is available without payment of a premium, although the ben- eficiary is responsible for an initial deductible and/or copayment for some services. Those not automatically eligible for Part A may enroll in the program by paying a monthly premium. Part B (Medicare): Medicare's Supplementary Medi- cal Insurance program which covers medically necessary physician services, hospital outpatient services, outpatient physical therapy and speech pa- thology services, and various other limited ambula- tory services and supplies such as prosthetic devices and durable medical equipment. Part B also cov- ers home health services for those Medicare bene- ficiaries who have Part B coverage only. Part B is optional and requires payment of a monthly pre- mium. The beneficiary is also responsible for a de- ductible and a coinsurance payment for most cov- ered services. Pass-throughs: In a prospective per case payment sys- tem, pass-throughs are elements of hospital cost that are paid on the basis of cost-based reimbursement. For example, under Medicare's new DRG payment system, capital costs, direct teaching, and outpatient services expenses are pass-throughs. Per case payment: A type of prospective hospital pay- ment system in which the hospital is paid a specific amount for each patient treated, regardless of the number and types of services or number of days of care provided. Medicare's DRG payment system for inpatient services is a per case payment system. Preferred provider organization (PPO):A contract agreement between providers (physicians or hospi- tals or both), patients, and insurers that medical care will be delivered at a discounted price as long as the patients use the “preferred providers,” i.e., those who are among the contractors. Premium: A form of beneficiary cost-sharing in which the insured pays a specified amount within a spe- cific time period (e.g., $14.60 per month) as the con- sideration paid for a contract of insurance. Prevalence: In epidemiology, the number of cases of disease, infected persons, or persons with disabilities or some other condition, present at a particular time and in relation to the size of the population. It is a measure of morbidity at a point in time. Price level depreciation: An estimate of depreciation (see definition) based on the current replacement value of the fixed asset. Procedure (medical or surgical): A medical technol- ogy involving any combination of drugs, devices, and provider skills and abilities. Appendectomy, for example, may involve at least drugs (for anesthe- sia), monitoring devices, surgical devices, and the skilled actions of physicians, nurses, and support staffs. Professional Standards Review Organizations (PSROs): Community-based, physician-directed, nonprofit agencies established under the Social Security Amendments of 1972 (Public Law 92-603) to monitor the quality and appropriateness of in- stitutional health care provided to Medicare and Medicaid beneficiaries. Prospective hospital payment: A hospital payment method in which the amount that a hospital is paid for services is set prior to the delivery of those serv- ices and the hospital is at least partially at risk for losses or stands to gain from surpluses that accrue in the payment period. Prospective payment rates may be per service, per capita, per diem, or per case rates. Medicare's DRG payment system for inpa- tient hospital services is a particular form of pro- spective payment. “Reasonable and necessary”: Criteria used by the Health Care Financing Administration or Medicare contractors to determine what services are eligible for Medicare coverage. Reasonable charge: The amount (subject to a patient deductible and coinsurance) Medicare will pay for a physician's service. The reasonable charge is the lowest of: 1) the physician's actual charge; 2) the physician's customary charge (the median of charges filed by a physician during the previous year for the service): and 3) the prevailing charge (calculated by multiplying the Medicare Economic Index by the 1973 prevailing charge which is the 75th percentile of the distribution of customary charges of all area physicians in 1972, weighted by the number of times each physician billed for the service). Recalibration: Periodic changes in relative DRG prices, including assignment of prices to new DRGs. Retrospective cost-based reimbursement: A payment method in which hospitals are paid their incurred costs of treating patients after the treatment has occurred. Technology assessment: A comprehensive form of pol- icy research that examines the technical, economic, and social consequences of technological applica- tions. It is especially concerned with unintended, indirect, or delayed social impacts. In health pol- icy, the term has also come to mean any form of policy analysis concerned with medical technology, Xi especially the evaluation of efficacy and safety. The comprehensive form of technology assessment is then termed “comprehensive technology assess- ment.” Utilization and quality control peer review organiza- Xi tions (PROs): Physician organizations established by the Tax Equity and Fiscal Responsibility Act of 1982 (Public Law 97-248) to replace Professional Standards Review Organizations. Hospitals are mandated to contract with PROs to review quality of care and appropriateness of admissions and readmissions. 1. Introduction and Summary A rational man acting in the real world may be defined as one who decides where he will strike a balance between what he desires and what can be done. It is only in imaginary worlds that we can do whatever we wish. — Walter Lippmann Contents Page Scopeoftie Report... ..... . 0. i a tea hs ea 6 Organization of the Rest of the Report... ................... 0 0 ii ide 6 SUIMIMAIY JL hc ih stam ssn iuns eae 7 Medicare Policies Affecting Medical Technology .......................... 7 The Impact of Medical Technology on Medicare Costs .................... 9 Overview of Areas for Change in Medicare ............. uc ivi did Dida 11 Specific Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs ............... ... 00. 00 iu. 11 Medicare Hospital Payment and Medical Technology ...................... 13 Medicare Physician Payment and Medical Technology ..................... 15 Alternative Approaches to Changing Incentives for Medical Technology Adoption'and Use .................. ........ 4 16 PONY ODHONS, «- oi vrs rss oh tiv eienisisins ns «sed sass isha ay SR 17 1. Introduction and Summary U.S. health care costs have escalated rapidly over the past 15 years, and medical technology® is a primary cause of the increase. Furthermore, now that controlling health care expenditures has become an issue of national prominence in the public and private sectors, increasing attention is being paid to the financial impact of the use of new and existing medical technologies. A major focus of Federal policymakers’ con- cerns about rising health care expenditures is the Medicare program, which provides payment for hospital and other acute care health services for ~ over 30 million elderly and disabled Americans. Since 1974, Medicare expenditures have been in- creasing at an average annual rate of 19 percent (135). Largely because of the Medicare program, the Federal share of national health expenditures has risen continuously since the program's incep- tion in 1966. Medicare expenditures, which repre- sented 48.9 percent of total Federal expenditures for personal health care? in 1970, represented 60.8 percent in 1982 (135). In 1982, Federal expendi- tures under Medicare totaled $52.2 billion. Of that 10TA defines medical technology as the drugs, devices, and sur- gical and medical procedures used in medical care, and the organiza- tional and supportive systems within which they are provided. Personal health care expenditures are national health care ex- penditures minus administrative costs. Photo credit: E. I. du Pont de Nemours & Co. In 1982, Medicare had over 30 million elderly and disabled beneficiaries amount, $36.3 billion went for hospital care, and $11.4 billion went for physicians’ services (135). Program expenditures in fiscal year 1984 are ex- pected to reach $66.5 billion (340). Medicare's beneficiaries, elderly and disabled Americans, are on average sicker than the general population. Furthermore, they are disproportion- ately high users of health care services in general and medical technology in particular. Every class of medical technology—with the exception of ob- stetrical, pediatric, and possibly preventive inter- ventions—is on average applied more often to Medicare beneficiaries than to the population as a whole. In 1980, those over the age of 65 ac- counted for 11.2 percent of the population but 31.4 percent of health care costs (265). Both per- centages are expected to rise significantly in the future because of the aging of the U.S. population. To aid in congressional efforts to contain Medi- care costs, the House Committee on Energy and Commerce and its Subcommittee on Health and the Environment and the Senate Committee on Finance, Subcommittee on Health asked OTA to assess a broad range of mechanisms to limit or reduce Medicare costs related to medical technol- ogy. In addition, they requested a study of the proposed use of Diagnosis Related Groups (DRGs) as Medicare's inpatient hospital payment mech- anism and several case studies of particular in- terest to the Medicare program. These are pub- lished as separate volumes. This report focuses on the policy mechanisms to limit or reduce Medi- care costs related to technology but draws from the study of DRGs and the case studies. The present assessment explores the dual rela- tionship between medical technology and the Medicare program: Medicare policies affect the adoption and use of medical technologies, and the patterns and levels of use of medical technologies significantly affect Medicare costs. It reviews spe- cific Medicare policies—eligibility, benefits, pay- ment, and beneficiary cost-sharing policies—that have had an influence on the adoption and use of medical technology. It also examines the con- 3 4 e Medical Technology and Costs of the Medicare Program tribution of medical technologies to increases in Medicare costs. OTA identified several possible changes in Medicare policies that could be used to influence medical technology adoption and use and to re- strain Medicare program costs. These mechanisms generally fall into the following categories: ® changes in Medicare's coverage policy for specific technologies; ¢ changes in the methods of Medicare payment to hospitals; ® changes in the methods of Medicare payment to physicians; and ® approaches to changing the incentives for the adoption and use of technology that do not directly involve, but may be related to, the Medicare payment mechanism (e.g., encour- aging the development of alternative cost- effective health care delivery systems). Because of the vast number of medical technol- ogies being developed or used and the decentral- ized administration of the Medicare program, technology-specific approaches are likely to be of limited value in containing Medicare costs.? For that reason, broader approaches, many of which involve the use of hospital or physician payment mechanisms to change providers’ or consumers’ financial incentives to use medical technology, are generally considered the major means by which the cost-containment objectives of the Medicare program might be achieved. The change in Medi- care’s inpatient hospital payment system—from retrospective, cost-based payment to prospective, per case payment based on DRGs, as mandated by the Social Security Amendments of 1983 (Pub- lic Law 98-21)—provides a striking example of such an approach. Other broad approaches do not involve the payment mechanism directly but are usually con- sidered in conjunction with payment mechanism changes to alter the incentives for technology use. These include stimulating competition among pro- viders of health care by encouraging the devel- opment of alternative sites or organizations (e.g., *Technology-specific approaches may be particularly valuable, however, in enhancing the adoption of socially valued technologies that may be cost-raising. health maintenance organizations) of health care delivery. They also include administrative changes in the Medicare program (e.g., merging Parts A and B of Medicare) for the purpose of changing incentives for technology adoption and use. There are two additional broad approaches to containing Medicare costs, but they are not dis- cussed extensively in this report. The first ap- proach is simply to limit the amount of money available for Medicare. Applying such a finan- cial squeeze would give providers and patients strong incentives to adopt and use technologies efficiently. However, applying such a limitation to the Medicare program alone, while saving Fed- eral dollars, would likely either shift costs to the private sector or result in Medicare beneficiaries’ reduced access to certain technologies. The sec- ond approach is to use the conditions of partici- pation for Medicare providers (i.e., requirements providers must meet in order to be eligible to re- ceive payment from Medicare) to change the in- centives for technology use. Several points should be kept in mind while reading this report. These points are not presented in order of importance, primarily because the issues involved are intertwined. First, the impact of medical technologies on Medicare costs, or health care costs in general, should not be assessed in isolation from the ef- fect that such technologies have on quality of care. The impact of cost-containing measures on both quality and access is one of the most difficult pol- icy issues to be faced, because the Medicare pro- gram was instituted to increase elderly persons’ access to acute care services. In order to control Medicare costs in the long run, some restrictions on quality or access are likely to be necessary. Unfortunately, the rapid rate of growth in health or Medicare expenditures cannot be stemmed simply by eliminating technologies that do not provide any benefit, because most technologies do provide some benefit, however small or costly the benefit may be (25). Nevertheless, there is substantial evidence to suggest that inappropriate use of medical technol- ogy is common and raises Medicare and health system costs without improving quality of care. Many surgical procedures seem to be overused in Ch. 1—Introduction and Summary © 5 the United States compared to other countries (26). Laboratory examinations and other diag- nostic tests are used at high rates and at times when not indicated by the suspected conditions (90,120,296). Lengths of stay in the hospital are higher in many cases than can be justified by med- ical evidence of benefit (350). Thus, one way to reduce Medicare costs is to encourage the appro- priate use of new and existing medical tech- nologies. Second, there are interactions between Medi- care and the rest of the U.S. health care system. Because of its size and scope, and because other third-party payers often follow Medicare's exam- ple, Medicare's policies and procedures affect all aspects of health care delivery, including financ- ing, administration, organization, and personnel. Furthermore, the program affects the content and costs of health care by its influence on the devel- opment, adoption, and use of medical technology. Nevertheless, it is important to keep in mind that the Medicare program is only one of many public and private institutions that influence the development and diffusion of medical technology. Other important influences are the Food and Drug Administration, the National Institutes of Health, manufacturers of drugs and medical devices, hos- pitals, private health insurers, and professional medical societies. The long-term costs of the Medi- care program are linked with those of the overall health care system, and the leverage of using Med- icare-specific policies to achieve Medicare cost- containment objectives may be limited. Third, reimbursement policy by Medicare and other third-party payers has contributed to the rapid adoption and often excessive use of medi- cal technology. Therefore, policymakers have looked to changes in reimbursement policies to alter the financial incentives for providers and consumers to use medical technology. However, reimbursement is only one of several factors that contribute to the tendency to adopt and use med- ical technology. Other factors include public de- mand for sophisticated technologies, the desire of physicians to do as much as possible for their pa- tients, competition among hospitals to achieve quality and prestige so as to attract patients and physicians, the fear of malpractice suits, and un- certainties about what constitutes appropriate use. 25-337 0 - 84 - 2 Fourth, because of spillover effects from one part of Medicare to another, policy mechanisms involving only one part of the Medicare program may have serious limitations in terms of contain- ing Medicare costs or affecting technology adop- tion and use. Medicare's DRG hospital payment system, for example, excludes physicians’ serv- ices and outpatient care. These exclusions not only provide incentives for the shifting of costs out of inpatient hospital settings but leave physicians’ incentives to use medical technology unaffected. Any cost-containment effort must take into ac- count the fact that physicians play a central role in determining what services are provided to pa- tients in both hospital and other settings. Fifth, what constitutes rational and appropri- ate adoption and use of medical technology de- pends on whether the question is being viewed from a societal perspective, from the perspective of the Medicare program, or from the perspec- tive of individual providers or patients. A rational decision to adopt or use a medical technology is a decision based on the consideration of costs and benefits. A decision by hospitals or physicians to adopt a medical technology that improves the quality of care provided to patients may also raise the costs of the Medicare program. From the per- spective of the providers or patients, such a deci- sion may be entirely rational. However, the costs and benefits to providers and patients are different from those to the Medicare program. Thus, unless the marginal increase in the benefit of improved patient care justifies the marginal increase in costs to the Medicare program, the decision may not be rational for Medicare. Furthermore, what is rational and appropriate from the standpoint of Medicare is not necessarily rational and appro- priate from the standpoint of society as a whole. Sixth, the social and political climate today is quite different from that in 1965, and now that Medicare's goal of improving access to health care for the Nation's elderly has been largely achieved, the primary focus of policymakers is on contain- ing Medicare costs. The principal intent of the 1965 legislation establishing the Medicare program under Title XVIII of the Social Security Act (Pub- lic Law 89-97) was to increase elderly Americans’ access to acute care medical services by remov- ing financial barriers, particularly to hospitaliza- 6 ¢ Medical Technology and Costs of the Medicare Program tion (317). There was far less concern about the cost of services than about the problem of access. The concern about access was also prominent in 1972, when eligibility was extended by Congress to disabled persons and people with end-stage re- nal disease (ESRD). More recently, however, the SCOPE OF THE REPORT This report examines medical technology's im- pact on Medicare costs and Medicare's past and future impact on the development and diffusion of medical technology. Medical technology is the major component of medical care. The incentives that govern the provision of medical care serv- ices work in the same direction (and are of the same magnitude) as those that govern the adop- tion and use of medical technology. In this report, therefore, the term “medical technology” is some- times used synonymously with “medical care” or “medical services.” The bulk of the increases in health and Medi- care costs in the past 20 years is attributable to factors other than changes in the patterns and levels of use of medical technology, such as gen- eral wage and price inflation and growth in the size of the U.S. population age 65 and over. A detailed examination of these factors, however, is beyond the scope of this report. The policy options presented in this report em- phasize controlling costs by changing the incen- tives for technology adoption and use, primarily through Medicare's hospital and physician pay- ment mechanisms. The report does not discuss changes in Medicare eligibility or in the broad Medicare benefit package. A serious problem needing attention that this report does not address costs of the program have soared, and the pres- sures for cost containment have increased. Thus, the challenge that remains for Federal policy- makers today is to solve the problem of control- ling Medicare costs without diminishing past success. is the widening gap between the Medicare bene- ficiary population's needs and the benefit package actually provided. Medicare's benefit package was modeled after insurance plans of the early part of the century, when acute illness was the primary concern and when most patients either got well or died rather quickly (317). Some services criti- cal to chronic disease—preventive measures, custodial or long-term care, drugs in outpatient settings, and many rehabilitative services—were excluded from covered benefits. Ironically, as Medicare has achieved its objective of improved access to acute care services and mortality rates among the Nation's elderly have decreased, mor- bidity from chronic diseases has increased because of the aging of the population. Thus, elderly in- dividuals who cannot afford uncovered services remain an underserved segment of the population. This report does not consider how much money is appropriate to expend on Medicare beneficiar- ies. That decision is essentially a political one. As noted earlier, one way to cut Medicare costs, and change the incentives for technology use and adoption, is simply to cut money out of the sys- tem. The options presented in this report could be implemented regardless of the political deci- sions about how much money is appropriate to spend. ORGANIZATION OF THE REST OF THE REPORT The rest of this chapter presents a summary of the report and briefly lists the issue areas for which there are policy options. The body of the report is organized in three parts. Part One presents information on the interac- tions between medical technology and Medicare. Chapter 2 discusses Medicare's impact on the adoption and use of medical technology. After Ch. 1—Introduction and Summary ® 7 presenting a brief overview of general policies, the chapter describes the specific Medicare policies that have been important to the development and diffusion of medical technology. In chapter 3, the emphasis turns to medical technology's impact on Medicare costs. The chapter examines the patterns of medical technology use experienced by the Medicare population compared to the general population. Then, it reviews the evidence on the contributions of medical technology to general health care costs and to Medicare costs. Those contributions are discussed in the aggregate and with respect to particular technologies. Part Two contains chapters providing informa- tion about policies that have been and could be taken to restrain the costs of medical technologies in the Medicare program. Chapter 4 provides an overview of the issues underlying the remaining chapters. Chapter 5 reviews Medicare's coverage policy and process for individual medical technol- ogies and discusses possible linkages between tech- nology assessment and coverage for the purpose of containing costs. Chapters 6 through 8 exam- ine broader policies that have impacts on medi- cal technology adoption and use. Chapter 6 an- alyzes the implications for medical technology of current and potential methods of hospital pay- ment. Chapter 7 presents a similar analysis of phy- sician payment methods. It also includes a dis- cussion on how physicians influence technology use and how physician cost consciousness may be enhanced. Chapter 8 presents information on broad approaches (those other than direct Medi- SUMMARY Medicare Policies Affecting Medical Technology The very existence of Medicare and other third- party payers expands the market for medical tech- nologies and influences the quantity and kinds of medical technologies that are used and the settings in which they are used. Since the enactment of the Medicare program in 1965, a great deal of leg- islation has been passed with the purpose of curb- ing the escalation in Medicare's costs and control- ling the diffusion of medical technology. To date, care payment changes) to change the incentives for medical technology adoption and use, primar- ily by stimulating competitive behavior among providers. Part Three (ch. 9) presents the main findings and policy options of the study, organized by issue area. Appendix A describes the method used by OTA to conduct the assessment and lists the back- ground papers (including case studies) prepared in conjunction with the project. Appendix B con- tains the acknowledgments and the membership of the Health Program Advisory Committee. Ap- pendix C provides information on public and pri- vate technology assessment activities. Appendix D is a descriptive overview of selected alterna- tives to traditional health care delivery. Appen- dix E presents the results of a survey of Medicare contractors. A series of case studies was used to provide spe- cific examples of issues and problems (see app. A). The report makes reference to the case studies throughout. The full cases themselves are printed as separate volumes of OTA’s Health Technol- ogy Case Study Series. In addition, a technical memorandum entitled Diagnosis Related Groups (DRGs) and the Medicare Program: Implications for Medical Technology was prepared as part of this study and published in July of 1983. Finally, a summary booklet is available. It con- tains information similar to the following sum- mary section and the chapter on policy options. such efforts have been largely ineffective. Indeed, Medicare's policies concerning eligibility, benefits, and payment have acted to promote technology adoption and use. Medicare's eligibility policy has made more medical technology available to millions of the Nation's aged and disabled people. When the pro- gram began in 1966, 19.1 million people aged 65 and over were eligible to enroll. By 1982, the num- ber of Medicare enrollees had increased to 29.5 million. The increase in the size of the Medicare 8 * Medical Technology and Costs of the Medicare Program population is due largely to the growth in the size of the elderly population, but some of it reflects the extension of Medicare eligibility to people with disabilities and ESRD on July 1, 1973. Medicare's benefit policy has had a profound effect on the types and location of modern medi- cal technologies. The Medicare law specifies broad categories of benefits for which the program will pay under two parts: Part A, the Hospital Insur- ance program, and Part B, the Supplementary Medical Insurance program. The law excludes most preventive services and certain other serv- ices, such as custodial and long-term care. Medi- care’s benefit package has undergone few major changes since the program's beginnings. Although the Medicare program covers a vari- ety of services in a variety of settings, its benefit package is concentrated primarily on acute care technologies provided in institutional settings, particularly those provided as inpatient hospital services. In 1982, 66.3 percent ($34.5 billion) of Medicare's $52.2 billion in payments was for in- patient hospital services. There are numerous in- centives inherent in Medicare's benefit policy to provide too many of some kinds of medical tech- nologies and too few of others. Coverage of some technologies (e.g., medical devices, drugs), for ex- ample, varies according to the characteristics of the technology, of the user, and of the setting in which the technology is used. In some cases, as in treatment for alcoholism, Medicare's benefit policy has encouraged the development of a tech- nology in an inpatient setting, despite the fact that treatment in other settings may be as effective and is certainly less costly (348). Medicare's exclusion of benefits for some technologies, including assis- tive communications devices (351), has had an un- favorable influence on their development and dif- fusion. A dramatic specific example of how Medicare's eligibility and benefit policies have affected the ‘Part A benefits include inpatient hospital care, post-hospital ex- tended care services, home health services, and as of Apr. 1, 1982, inpatient alcohol detoxification services. Part B benefits include med- ically necessary physician services, outpatient hospital services, out- patient physical therapy and speech pathology services, home health services for those not eligible for Part A, and various other limited ambulatory services and supplies (e.g., prosthetic devices and durable medical equipment). development and diffusion of costly medical tech- nologies is provided in the case of technologies used to treat ESRD. People with ESRD require some form of dialysis or kidney transplantation to prolong their lives. In 1972, before Medicare eligibility was extended to persons with ESRD, about 10,000 persons were receiving hemodialy- sis. By 1980, following the extension of eligibility, 50,000 were being dialyzed (359). There was also a significant increase in kidney transplantation fol- lowing implementation of the ESRD program (359). Currently, an estimated 93 percent of the U.S. population with ESRD are Medicare benefici- aries (195). Thus, Medicare policies can be clearly identified as a major influence on the diffusion of the technologies used in the treatment of ESRD. Photo credit: National Kidney Foundation, Washington, D.C. Medicare policies are a major influence on the development and diffusion of hemodialysis and other technologies used in the treatment of ESRD Ch. 1—Introduction and Summary ® 9 Medicare's payment policies have had the most profound effect on medical technology adoption and use of any of the program's policies. For many years, Medicare has paid hospitals and other in- stitutional providers on the basis of reasonable cost and paid physicians and other noninstitu- tional providers reasonable charges on a fee-for- service basis. Under both payment methods, pro- viders receive more reimbursement when they use more medical technology. Thus, these payment methods offer providers little deterrent to the in- creased use of technology and little incentive to choose less costly technology. Although Medicare's hospital payment system is now in the process of change,® 17 years of cost- based hospital payment have shaped the health care system today. The original Medicare legis- lation left the specific method of determining rea- sonable cost to administrative decisions. The method adopted was very liberal in allowing hos- pitals considerable discretion in calculating the costs attributable to Medicare (104). Thus, be- cause hospitals have been assured of reimburse- ment by Medicare and by other third-party pay- ers, they have had no financial reason not to spend money on medical technology, especially on socially valued technology. Medicare's method of paying physicians has changed little since Medicare was enacted, al- though minor restraints have been imposed on the rate of increase of physicians’ payment levels. Most physicians’ incomes are determined by the number and intensity of services delivered and the fee received for each service. The use of technol- ogy by fee-for-service physicians is sensitive to the additional revenue they receive (229). In addition, although not intrinsic to the fee- for-service payment method, physician payment levels that Medicare has established for complex and expensive medical technologies are usually disproportionately high (235). In most instances, the reimbursable charge for a technology was es- sMedicare’s hospital payment method was changed by the Social Security Amendments of 1983 (Public Law 98-21), which mandated the phasing in beginning in October 1983 of a prospective per case inpatient hospital payment system based on Diagnosis Related Groups (DRGs). The implications of DRG hospital payment are dis- cussed in a following section. tablished at an early point in the technology's history. Although subsequent technological ad- vances and higher rates of utilization may have substantially reduced the time, judgment, skill, and cost required to use the technology, this change is not reflected in the physician's fee or Medicare's reimbursement level. Furthermore, the existing payment system provides incentives for the use of “technology-intensive” medical care. Under current fees, what are sometimes referred to as “technology-oriented” services, such as diag- nostic tests and surgical procedures, are valued higher than “cognitive” services, such as taking medical histories and counseling. Medicare's beneficiary cost-sharing provisions were the only measures specifically included in the original legislation to help moderate the un- necessary utilization of services. Although there is little empirical evidence concerning the effect of deductibles, copayments, and coinsurance on the use of medical technology specifically in the Medicare program, it is generally believed that such cost-sharing has had little impact on tech- nology use. Supplementary health insurance (“Medigap”) is used extensively by Medicare ben- eficiaries, and it often substantially diminishes or eliminates the burden of these cost-sharing re- quirements. Premium payments, another form of cost-sharing, are clearly not an obstacle to the use of services (28,253). The Impact of Medical Technology on Medicare Costs Changes in the kinds of medical technologies available and changes in the patterns of use of technologies already available continually influ- ence health care costs—at times moderating cost increases and at times exacerbating them. How medical technology contributes to health and Medicare costs is a question that can be addressed either in the aggregate or with respect to particu- lar technologies or classes of technologies. The question from the aggregate perspective is whether changes in medical technology use as a whole have raised or reduced health care or Medi- care costs and, if so, by how much. The aggregate perspective is useful, because it puts technology's relationship to costs into a policy perspective. 10 » Medical Technology and Costs of the Medicare Program The most widely used approach to estimating technology's aggregate contribution to health care costs is to separate the change in total expendi- tures for health care into its component parts: population or enrollment changes, overall wage and price inflation, wage and price inflation in medical care in excess of general inflation, and changes in service intensity .* Changes in technol- ogy use are included in the latter two measures, although these measures also reflect other factors. Using this general approach, OTA estimates that increases in service intensity (labor and nonlabor inputs) per capita accounted for 24 per- cent of the 93-percent increase in per capita hos>~ pital costs from 1977 to 1982. The increase in serv- ice intensity is due in part to an increase in the hospital admission rate (a 5-year increase of 2.1 percent), but the overwhelming part of the in- crease is due to the provision of a greater quan- tity of services per hospital admission. Moreover, OTA'’s empirical and literature analysis supports the general conclusion that changes in service in- tensity have contributed substantially to the growth in hospital costs over the past 20 years. Increasing intensity of care appears to be a less important source of expenditure inflation in total personal health care expenditures in the United States than it is for the hospital sector alone. The combined effect of increasing intensity of care and increasing health care prices in excess of the Con- sumer Price Index accounts for only about 16 per- cent of the growth in per capita personal health care expenditures between 1977 and 1982. Dur- ing that 5-year period, however, these two tech- nology-related components of cost together increased real per capita personal health expendi- tures at an average annual rate of 2.8 percent. It is possible to account for the components of Medicare cost increases, but the interpretation of the estimates is even more clouded than is the in- terpretation of increases in general health care costs. Changes in program eligibility or in cov- ered benefits can lead to dramatic changes in “Service intensity” refers to the quantity of inputs that go into producing a given unit of health care. These inputs include labor, supplies, materials, and equipment used in the provision of care. Service intensity is associated with, but not identical to, medical technology use. measured service intensity that have little to do with changes in medical technology but instead represent a shift in the burden to payment for services already available and used. Changes in per capita service intensity do indicate how much more or less Medicare is paying for now than at some earlier date. Between 1977 and 1982, Medi- care expenditures per enrollee increased 107 per- cent. OTA's analysis indicates that nearly 30 per- cent of the increase in Medicare costs per enrollee from 1977 to 1982 can be attributed to increased use of covered services (25 percent) and increased medical prices in excess of general inflation (3 percent).” Although none of the approaches to measur- ing technology's aggregate contribution to health care or Medicare costs is entirely satisfactory, taken as a whole, the available evidence leads to the conclusion that health care costs have in- creased in part because more is being done for pa- tients today than ever before. More and better trained personnel, more procedures, more drugs, and more and higher priced equipment, materials, and supplies are being used in the delivery of health care to Medicare patients and to the popu- lation as a whole. So far, the trend toward “more” does not appear to be abating. It is not just at the margin, however, that there is an opportunity to reduce Medicare costs. There are many opportu- nities to save health and Medicare costs by alter- ing longstanding patterns of use of medical tech- nology. Furthermore, the aggregate approach to esti- mating medical technology's contribution to health care or Medicare costs is limited, because it ignores the patient benefits associated with cost increases or decreases, it does not take into ac- count the underlying reasons for changes in med- ical decisions or practices, and it does not show that cost-saving and cost-raising changes in tech- nology are not scattered evenly across illnesses. Summary statements about technology’s net in- fluence on health care or Medicare costs mask the "The percent due to medical price inflation may be overstated, and the service intensity percentage correspondingly understated, because the amount Medicare actually pays for services (i.e., the effective price) probably lies somewhat below stated prices. Ch. 1—Introduction and Summary 11 rich assortment of ways in which changes in med- ical technology shape the health care system, the population's health status, and its costs. Thus, in order to provide insight into the underlying reasons for change in medical decisions or practices and to highlight the extent to which the costs of the Medicare program are altered by new technologies, OTA examined seven specific technologies first introduced in the 1960's or 1970's: coronary artery bypass surgery, the drug cimetidine, therapeutic apheresis, pneumococcal vaccine, intensive care units, total parenteral nu- trition therapy, and kidney dialysis. All seven of the technologies have clear patient benefits—in some cases, they are life saving— but for all of the technologies, there is controversy about the most appropriate indications for use. Two of these technologies have been or may be cost saving to Medicare, but five of them have raised or have the potential to raise Medicare costs, in some cases significantly. Above all else, these technologies illustrate how exposed the Med- icare program is to changes in medical technol- ogy that are largely beyond its control. The chal- lenge to Medicare in the face of new technologies that offer both patient benefits and higher costs is how to encourage the most cost-effective use of the most cost-effective technologies. The over- all remaining issue is how Medicare policy can be structured to bring about more cost-effective use of both existing and new medical technologies. Overview of Areas for Change in Medicare OTA's discussion of potential areas for change under Medicare is organized in two parts, cor- responding to the two types of policy mechanisms discussed previously. The first part—policies directed at individual technologies—explores link- ing Medicare's coverage policy and technology assessment to contain costs. The second part— policies providing broad incentives to encourage appropriate adoption and use of technologies— is divided into three sections: hospital payment, physician payment, and alternative or systemwide approaches to changing incentives. Specific Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs A potential method of containing Medicare costs is by influencing the diffusion (i.e., adop- tion and use) of medical technologies. It is gener- ally agreed that Medicare's coverage policy — policy that governs the eligibility of services (tech- nologies) for payment—has influenced decisions about the purchase of some expensive, visible medical technologies. The precise relationship be- tween coverage policy and adoption of other kinds of technologies or use of any technologies remains speculative. Although Medicare and other insurance plans designate broad categories of services, such as in- patient services, as being covered, or eligible, for payment, specific technologies, with few excep- tions, require individual coverage decisions. Medi- care coverage policy for particular technologies not mandated by law has been decided on a case- by-case basis according to Section 1862 of Title XVIII of the Social Security Act. Section 1862 pro- hibits Medicare payment for items and services that are “not reasonable and necessary for the di- agnosis or treatment of illness or injury or to im- prove the functioning of a malformed body mem- ber.” Medicare has refrained from establishing a definitive interpretation of “reasonable and nec- essary” and has relied on a loosely structured and decentralized mechanism to determine whether a technology is covered. The criteria used to deter- mine if a technology meets the broad statutory language of “reasonable and necessary’ are: 1) general acceptance as safe and effective, 2) not experimental, 3) medically necessary, and 4) pro- vided according to accepted standards of medi- cal practice in an appropriate setting. Tradi- tionally, coverage policy has been made in light of Medicare's principles of not interfering with the practice of medicine and assuring beneficiaries a free choice of providers. Some coverage decisions are made at the na- tional level by the central Health Care Financing Administration (HCFA) office, while others are made by regional HCFA offices. Most of the deci- 12 Medical Technology and Costs of the Medicare Program sions, however, are made by Medicare contrac- tors, called intermediaries and carriers, who per- form the Medicare program's claims processing and payment function at the local level under the policy and operational guidance of HCFA. Al- though the details vary, the coverage process is the same at the national level or at the contrac- tor level. First, new technologies and new uses of covered technologies are identified. Second, a decision is made about covering the identified technology for Medicare payment. The decision- makers (contractors or HCFA) may receive ad- vice, which usually involves an evaluation of the safety and effectiveness of the technology. The final step is implementing the coverage decision. Because of the general language of Section 1862 and the absence of regulations or guidelines that implement the section, Government officials and Medicare contractors have had considerable lati- tude in determining which technologies are to be covered for reimbursement. Contractors vary widely in their identification of uncovered tech- nologies, their decisions about the coverage of spe- cific technologies, and their implementation of na- tional coverage decisions (54,86,143,353). As a result, some technologies may be covered and paid for in one area and not in another. There is no national or local listing of procedures that are not covered (163). Problems pertaining to the administration of the coverage process that need attention include: 1) the inadequate identification of emerging and outmoded technologies for coverage decisions; 2) the lack of uniformity in implementing national coverage decisions; 3) the timelag involved in the coverage process, including technology assess- ment; 4) the complex coding system and prolifera- tion of codes; and 5) the incomplete dissemina- tion of information. These problems all poten- tially raise Medicare's costs, although some of them (e.g., numbers 2, 3, and 5) may actually de- crease Medicare expenditures. Of particular interest to cost-containment ef- forts is Medicare's policy of not explicitly consid- ering cost or cost-effectiveness information in making coverage decisions. Also of interest is the fact that Medicare has refrained from a policy of limiting coverage of particular technologies to re- stricted circumstances (e.g., institutions offering specific services or having specialized equipment) and to physicians with specific skills. Although the notion of limiting coverage to selected sites and providers has gained importance with the in- creasing development of sophisticated technol- ogies that require particular expertise, Medicare's principles of refraining from interfering with med- ical practice and assuring beneficiaries a free choice of providers appear to have limited its ap- plication. On the other hand, Medicare does limit coverage of some technologies to appropriate medical conditions. For example, therapeutic apheresis is currently covered for six disease in- dications. In theory, one way to use coverage policy to assist in containing Medicare costs would be to include cost criteria in technology assessments. Cost-benefit analysis (CBA) and cost-effectiveness analysis (CEA) are formal analytical techniques for comparing the positive and negative conse- quences of alternative ways of allocating resources (353). The methodological strengths and weak- nesses of CBA/CEA and the potential for expand- ing their use in coverage decisions was discussed in OTA's 1980 report The Implications of Cost- Effectiveness Analysis of Medical Technology. A methodological issue of particular importance to beneficiaries of the Medicare program is whether to include discounted future medical care costs (due to longer lifespans for patients resulting from the use of medical technology) as a direct cost of a technology. Incorporating cost criteria into an assessment, however, would not necessarily lead to the iden- tification of cost-saving technology. Achieving the objective of identifying technologies that save or raise costs to Medicare before they become estab- lished in medical practice is problematic. The tech- nical complexity of determining the cost effects of emerging and new technologies is compounded by the problem of defining a cost-saving or cost- raising technology. Differences in perspective im- pede arrival at a universal definition of a cost- saving or a cost-raising technology. A new issue for Medicare is how to coordinate coverage policy with the DRG hospital payment system. Although the coverage process and the Ch. 1—Introduction and Summary ® 13 process of adjusting DRG rates share a similar “approval for payment” function, they differ in that a coverage determination focuses on specific technologies, while adjusting DRG payment rates focuses on the larger entity of a diagnostic group, which includes particular technologies. Moreover, the DRG rate adjustment process must include issues of cost as an integral issue, while the cov- erage process at present does not consider cost issues. Nonetheless, the technology assessments performed for the coverage and DRG rate adjust- ment processes no doubt will have similarities and their coordination should be encouraged. Medicare Hospital Payment and Medical Technology The retrospective, cost-based hospital reim- bursement system under which Medicare operated from 1966 until fiscal year 1983 was significantly altered first by the Tax Equity and Fiscal Respon- sibility Act of 1982 (Public Law 97-248) and then by the Social Security Amendments of 1983 (Pub- lic Law 98-21). The latter mandated the phasing in over a 3-year period of a prospective, per case hospital payment system based on DRGs. The new prospective payment system for inpatient operating costs places hospitals at financial risk but also enables them to keep whatever surpluses can be generated. Although capital, outpatient, and direct teach- ing expenses remain pass-throughs,® Medicare's DRG hospital payment system has radically changed the financial incentives for the adoption and use of specific medical technologies in hos- pitals.’ Hospitals now have a financial incentive to increase hospital admissions and decrease lengths of stay. Some patients may be admitted unnecessarily, others may be discharged too early, and some may not get all the elective care in one hospital stay. Thus, hospital admissions and re- admissions will need to be monitored. The DRG payment system also provides hos- pitals with incentives to reduce the number and sPass-throughs are those elements of cost that are not part of the prospective payment system. 9OTA's technical memorandum entitled Diagnosis Related Groups (DRGs) and the Medicare program: Implications for Medical Tech- nology (343) provides a detailed analysis of these incentives. cost of ancillary services. Prior to the implemen- tation of DRG payment, hospital administrators had financial reasons to encourage physicians to use available technologies. Now, hospital admin- istrators will need to gain the support and coop- eration of their physicians in order to keep their inpatient care within the price range of DRG pay- ments. Under the new system, hospital adminis- trators are likely to discourage physicians from using many high-cost technologies. In some cases, the substitution of low-cost technologies for high- cost technologies may result in a decline in quality of care. Thus, quality of care remains an impor- tant issue under DRG payment. Congress has pro- vided some control over quality of care by man- dating the utilization and quality control peer review organizations (PROs). Hospitals must have signed agreements with these organizations in or- der to receive Medicare payments. One of the responsibilities of the PROs will be to monitor the potential admission/discharge/readmission problem. Despite the recent establishment of the DRG hospital payment system, it is quite possible that changes in hospital payment by Medicare will be actively considered in the future. Part of the rea- son is that pressure for cost containment at the Federal level may continue, and part is that in- dividual States may enact hospital cost control systems in which Medicare will agree to partici- pate. Alternative approaches that have been sug- gested or applied by public or private payers and that might be considered for implementation by Medicare generally fall into four major categories: eo alternative hospital prospective payment methods or modifications of Medicare's DRG hospital payment system, e capital payment methods, e limited provider contracts, and e increased patient cost-sharing for hospital services. Congress has adopted DRGs for the Medicare hospital payment system, but improvements of DRGs and of the payment system should be pur- sued. Case-mix classification systems with more desirable properties than DRGs may become available in the future. Innovations in medical devices, drugs, and medical techniques that raise the quality of care for the Medicare population 14 e Medical Technology and Costs of the Medicare Program Photo credit: University of Utah Medical Center Updating the DRG hospital payment system will be necessary to encourage the adoption of technologies that raise the quality of care provided but also raise hospital per case costs but also increase hospital per case costs may not be readily adopted unless DRG payment rates are updated. Refinements of Medicare's DRG-based hospital payment system are anticipated in light of the series of congressionally mandated studies and the charge to the Prospective Payment Assess- ment Commission to recommend changes in DRG relative weights and categories. Other approaches to prospective payment of hospitals are certainly possible, and the current Medicare law encourages States to experiment with these as part of all-payer systems. Innovative prospective payment meth- ods such as per capita hospital payment and area- wide global budgeting may hold promise in some areas. How Medicare will pay for hospitals’ invest- ments in capital plant and equipment under DRG payment is an issue that has yet to be resolved. Traditionally, Medicare has reimbursed hospitals for interest and historical cost depreciation ex- penses associated with all capital equipment. This payment method has increased hospitals’ demand for capital but has also made it difficult and costly for some hospitals to obtain additional debt fi- nancing. Currently, under DRG payment, capi- tal costs are treated as pass-throughs (i.e., reim- bursed, as they always have been, as they are incurred with no limit on the amount that a hos- pital can be paid). Of particular concern with a capital cost pass-through under DRG payment is the incentive for hospitals to adopt expensive cap- ital equipment that reduces operating costs but raises total costs per case. Congress has recognized that capital costs are still a problem for Medicare, and Public Law 98-21 requires the Department of Health and Human Services to study how capi- tal costs should be paid in connection with the DRG hospital payment system. Two possible alternatives to the pass-through are to incorporate a flat rate for hospital capital into the DRG rates and to build hospital-specific capital allowances into the DRG system. Al- though the flat rate approach is generally more efficient than pass-through capital payment, it does raise questions of fairness among hospitals and equity of access to medical technologies among patients. In a flat rate payment system, hospitals that in the past had lower ratios of cap- ital to operating costs would receive more pay- ment than they had in the past. The hospital- specific approach would tend to reward those hos- pitals that were most highly capitalized in the past, leaving those with less capital forever to receive lower payments. Two additional approaches to affecting the use of medical technologies through hospital payment are limited provider contracting and increased beneficiary cost-sharing for hospital services. Both methods have significant limitations. Limited pro- vider contracts for hospital care'® would involve selecting certain hospitals for the provision of in- patient care to Medicare beneficiaries. Overall, although contracting may save program dollars, it represents an abandonment of the principle of assuring beneficiaries freedom of choice of pro- 1°Currently, State Medicaid agencies may apply for waivers from the freedom-of-choice provision of the Social Security Act. Most waivers to date have been for case management systems that re- strict the providers from whom a Medicaid beneficiary can obtain primary care, although California has adopted an approach of con- tracting with hospitals for inpatient care for Medicaid beneficiaries. Ch. 1—Introduction and Summary © 15 viders on which Medicare was built and forces subsidies of hospital care from other payers. Medicare Physician Payment and Medical Technology Any cost-containment effort must take into ac- count the fact that physicians are key decision- makers with respect to the use of medical tech- nologies. Physicians determine the amount of medical services to be provided, when patients need hospitalization, and when they need other types of care. There is substantial evidence to suggest that in- appropriate use of medical technologies is com- mon and raises costs without improving quality of care. Such excessive use exists within the norms of medical practice and across the spectrum of technologies available to physicians. Physicians’ habitual behavior can cause excessive use of med- ical services. Until recently, medical education trained physicians to do all they could for their patients’ well-being without concern for the cost. In response to restraints in their payment, physi- cians have changed the number and mix of serv- ices they provide. The practice of defensive medi- cine in response to fears of lawsuits may also increase unnecessary use, and thus cost, of med- ical technologies. Physician behavior with regard to the use of medical technologies may be modified by finan- cial incentives, educational programs, utilization review programs, and other programs such as sec- ond surgical opinion programs. Studies show that the results of different programs and interventions vary both in effectiveness and longevity. Changes in physician payment methods can also influence physicians’ incentives for the use of medical technologies. Physicians who are paid on a fee-for-service basis have financial incentives to see more patients more often and provide more technologies. Physicians (or practice plans in which they participate) paid on a per capita basis have financial incentives to increase the number of their patients but to keep the number of pa- tient visits low (or nonexistent) and to use par- ticularly low-cost technologies. The financial in- centives under a fee schedule system depend on the particular type of schedule adopted. Under fee schedules based on patient visits, physicians have incentives to schedule more visits but disincen- tives to use a large number of technologies (par- ticularly those whose costs are high in relation to the fee per visit received). Under fee schedules based on episodes of illness, physicians have in- centives to treat for more episodes but to keep patient visits for each episode and the use of costly technologies at a minimum. Most changes in Medicare physician payment methods would necessitate a reformulation of the diagnostic and procedural codes for physician services that are currently used by the program. The present coding system makes it fairly easy for physicians to adopt and use medical technol- ogies. Furthermore, the large number of proce- dural codes makes it fairly easy for physicians to bill for expensive services and to make expensive coding errors. Changes in Medicare's physician payment methods that could help contain costs for the Medicare program by influencing the adoption and use of medical technologies are of two general types. One is requiring patients to assume more responsibility for their health care costs, either through increases in beneficiary cost-sharing or a reduction in the types of benefits Medicare cov- ers. It should be noted, however, that elderly ben- eficiaries already have greater out-of-pocket expenses than the younger population, and in- creased cost-sharing may reduce their access to health care. The second type of change involves restraining the amount or changing the methods of Medicare payment to physicians (e.g., by fee schedules or freezes on current fee levels). Either approach could result in cost savings for the Medi- care program, but each would have different ef- fects on the adoption and use of medical technol- ogies and on access to medical care by Medicare beneficiaries. Changing Medicare's claim-by-claim voluntary physician assignment policy! would 11Medicare permits physicians the option of being paid directly by Medicare, called “accepting assignment,” or being paid directly by the patient. If a physician does not accept assignment, the Medi- care reasonable charge, which is paid directly to the patient, may be less than the physician's actual charge, and the patient is respon- sible for the difference between the two. 16 * Medical Technology and Costs of the Medicare Program strengthen the implementation of the other changes, although it might discourage some phy- sicians from treating Medicare patients. Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use Alternative approaches that could be used by Medicare to foster the appropriate adoption and use of medical technologies, and ultimately save costs, include two general policy mechanisms: 1) methods to foster competitive behavior by pro- viders, and 2) administrative changes in Medicare. These mechanisms include changes involving the general health care system that Medicare could embrace and changes in the structure of the Medi- care program itself. It is generally believed that increases in the costs of the Medicare program and of the overall health care system can be contained by the rational adop- tion and use of medical technologies, which in- cludes using technologies in appropriate settings. An important method of stimulating such adop- tion and use is to foster competitive behavior by providers. In most cases, it is through policies en- couraging the use of alternative sites and organi- zations for health care delivery that competitive behavior is expected to occur. Alternatives to fee- Photo credit: National Institutes of Health Home health care as a substitute for an extended hospital stay may be underutilized for-service, solo physician office practices and traditional inpatient hospital settings include site alternatives, such as freestanding ambulatory sur- gery centers, emergency care centers, hospices, hospital outpatient departments, home health care, and nursing homes; and organizational al- ternatives, such as health maintenance organiza- tions, the use of primary care gatekeepers, and preferred provider organizations. Long-range cost containment in the Medicare program is constrained by the kinds of health care delivery systems available and the limited influ- ence that Medicare financing can have on the set- tings of care and kinds of technologies provided. In recent years, the Medicare program has granted exceptions to specific alternative sites of care (e.g., freestanding ambulatory surgery centers) and en- couraged the demonstration and evaluation of alternative organizations for health care delivery (e.g., preferred provider organizations). Thus, Medicare's efforts in developing competition with the types of care predominantly available have been to identify and encourage other types of pro- vider practices and modes of delivery. In the long run, it is hoped, alternatives of these types will lead to cost-effective health care. A complementary approach to increasing com- petition among providers involves moving from the current Medicare program structure to mak- ing available other types of health insurance cov- erage to beneficiaries. The most discussed possi- bility is the use of vouchers, wherein Medicare beneficiaries would receive a specified amount of money to purchase health insurance from the mar- ketplace instead of participating in the traditional Medicare program. Important decisions regarding competition for policymakers in the Medicare pro- gram are: 1) the relative emphases to be placed on the insurance versus the alternative delivery systems approach, and 2) the pace of adopting the various competitive approaches into Medi- care. To increase the capability of Medicare to embrace competitive approaches, the program could undergo an administrative change—merg- ing Parts A and B. Merging the two parts could alleviate the financial problems of the Medicare program and improve the quality of care for pa- tients. Ch. 1—Introduction and Summary ® 17 POLICY OPTIONS The final chapter of this report presents policy generating effective and efficient adoption options for congressional consideration. Rather and use of technology? than to recommend specific actions, OTA'’s pol- e How can Medicare's physician payment icy is to provide Congress with a series of alter- method be used to improve incentives for native actions and discussions of the possible con- appropriate technology adoption and use? sequences of implementing them. The policy e What broad approaches, other than those options in this report are organized by the fol- directly involving Medicare's payment mech- lowing issue areas: anism, could be used by Medicare to encour- age the appropriate adoption and use of med- e How can the Medicare coverage process for teal tachuolosy? specific technologies be improved? e How can Medicare's hospital payment sys- Findings and options related to each issue are pre- tem incorporate appropriate incentives for sented in chapter 9. Part One 2. Medicare Policies Affecting Medical Technology Life can only be understood backwards, but it must be lived forwards. Soren Kierkegaard Contents Page Introduction... .... vod Bo, nosh Rem ns he 23 Development and Diffusion of Medical Technologies ............. ........ 23 Overview of Medicare Policies That Affect Medical Technology ............. 25 Eligibility 1... saison is ale Ta TT 26 Benefits ....c... c,h oa ae 27 Payment... .o.:. vs eroishiby cde, a EE G 30 Discussion... cei, de ali ints LC DLL Ua a Ye a 38 LIST OF TABLES Table No. Page 1. Limitations Placed on Medicare Payment Levels ..................... . 26 2. Number of Elderly and Disabled Beneficiaries Enrolled in Medicare by Type of Coverage, Selected Years From 1966 to 1982... .... 27 3. Medicare Benefits and Limitations, as of January 1984 ........ .... .. . 28 4, ESRD Patient Population, 1972101982 ........ 0. 0 0 200 oi 35 FIGURE Figure No. Page 1. A Model for the Development and Diffusion of Medical Technologies . ... .. 24 2. Medicare Policies Affecting Medical Technology INTRODUCTION Because of its size and scope and because other third-party payers often follow its example, the Medicare program is a major force in the struc- ture and performance of the U.S. health care sys- tem. Medicare currently accounts for more than 35 percent of national health spending for hospi- tal care and more than 18 percent of national spending for physicians’ services (135). Further- more, in fiscal year 1984, program spending is ex- pected to increase by 16.3 percent to $66.5 bil- lion (340). Medicare's policies and procedures affect all aspects of health care delivery in the United States, including financing, administration, organization, and personnel. Medicare also affects the content of U.S. health care by its influence on the development and diffusion (i.e., adoption and use) of medical technology. Despite its importance, however, the Medicare program is only one of many institutions that af- fects the development and diffusion of medical technology. Other institutions that affect the direction and pace of technological change include Federal agencies such as the National Institutes of Health and the Food and Drug Administration (FDA), as well as private organizations ranging from manufacturers of drugs and devices, hospi- tals, private insurers, and professional medical societies.! Some programs, such as the certificate- of-need program, have been enacted to influence medical technology but have not had the intended impact. This chapter analyzes the extent and limits of Medicare's contribution to the development and diffusion of medical technology. It begins by describing a model of the process of technologi- cal change and then considers the effects on this process of Medicare's eligibility, benefits, and pay- ment (including beneficiary cost-sharing) policies. The effect of medical technology on Medicare costs, another important interaction between Medicare and medical technology, will be con- sidered in chapter 3. 1For an analysis of public and private sector roles in the develop- ment and diffusion of medical technology, see earlier OTA reports, including Strategies for Medical Technology Assessment (359) and The Impact of Randomized Clinical Trials on Health Policy and Med- ical Practice (352), as well as Toward Rational Technology in Medi- cine by H. D. Banta, C. J. Behney, and J. S. Willems (24). DEVELOPMENT AND DIFFUSION OF MEDICAL TECHNOLOGIES? Broadly defined by OTA, medical technologies are the drugs, devices, medical and surgical pro- cedures used in medical care, and the organiza- tional and supportive systems in which such care is provided (341). In the past few decades, numer- ous and impressive changes have been made in the types of medical technology available to the The development and diffusion of medical technology are de- scribed in detail in previous OTA reports, including Assessing the Efficacy and Safety of Medical Technologies (341) and The Impact of Randomized Clinical Trials on Health Policy and Medical Prac- tice (352). health care system. In the case of drugs, devices, and procedures, many new technologies, new uses of established technologies, and improvements on technologies have been developed and marketed. Furthermore, sophisticated managerial technol- ogies, such as computer-based hospital informa- tion systems, are being purchased and used with increasing rapidity. Alternatives to the traditional modes and sites of health care delivery—the hos- pital and the physician's office—that have pro- liferated recently include a variety of alternative organizational arrangements, including hospital 23 24 Medical Technology and Costs of the Medicare Program chains, hospital management corporations, free- standing emergency care centers, and ambulatory care centers (see ch. 8). The process of technological change occurs in two stages: the development of a technology and the subsequent diffusion of the technology into medical practice (24). This process may be divided into sequential steps within the developmental and diffusion stages for the purpose of analysis, as shown in figure 1. But the process of change is often less linear and systematic than the gener- alized version depicted in the figure. The development of medical technology occurs in various sites and with a variety of funding sources, depending on the type of technology. While most of the basic biomedical research and some applied research and technology develop- ment in this country is funded by the Federal Government, the greatest portion of applied re- search and technology development is funded by private industry. Much of modern medical tech- nology is a combination of drugs, medical devices, and human skills, and its development is very complex and not well understood. The role of Medicare and other payment sys- tems in the development of drugs, devices, and procedures has received little study. In theory, however, payment systems can influence the de- velopment of some types of technology that are produced by the private sector. In order for a private firm to justify committing funds for the The forthcoming OTA report, Federal Policies and the Medical Devices Industry (345), will provide information on the role of pay- ment systems in the development of medical devices. Figure 1.—A Model for the Development and Diffusion of Medical Technologies Extent of human use =) > ) 4 > — >» > ~ =~ ~- Innovation ~ wn Basic Applied Targeted First Clinical Early Late Acceptea research research development human trials adoption adoption practice Wa Nt Clinical investigation Basic research Applied R&D and testing SOURCE: Office of Technology Assessment. Diffusion Ch. 2—Medicare Policies Affecting Medical Technology ® 25 research and/or development of a potential new medical device, for example, the firm must per- ceive the existence of a market for the innovation (345). A payment system that favors the adop- tion of new devices would help a firm make a deci- sion to proceed with development. The develop- ment of technologies such as hospital information systems may be subject to market influences as well. Theoretically, Medicare's new inpatient hos- pital payment system mandated by the Social Security Amendments of 1983 (Public Law 98-21) should stimulate the development of new hospi- tal information technologies. It is clear that pay- ment systems have played an important part in the development of some new organizational pat- terns of medical care such as preferred provider organizations or multihospital systems (325). The diffusion of a medical technology into the health care system has two phases: the initial phase in which decisions are made to adopt the technology, and a subsequent phase in which deci- sions are made to use the technology (24). Adop- tion has been studied far more by researchers than use has and has also been the subject of much greater direct involvement by the Government (24). Decisions to adopt medical technology within organizations such as hospitals are made by physicians, hospital administrators, and pur- chasing departments. Decisions about using tech- nology are primarily made by physicians, al- though patients’ decisions are also important. The use of a technology is obviously dependent on its adoption. The exact relationship between the two phases, however, has not been established. Before adoption (or rejection) can occur, knowledge about a technology must be commu- nicated to potential adopters (24). One focus of studies on the adoption of medical technology, therefore, is how knowledge about technology is communicated. Research on communication about drugs has led to the description of a two-step model. The first step is the flow of information from industry to those physicians who are opin- ion leaders. The second step is the transfer of in- formation from the physician opinion leaders to their followers through informal channels (324). Recent research on the adoption of hospital in- formation systems found a similar two-step model (16). Another focus of studies on the adoption of medical technology is on the factors influencing adoption. Such factors include the characteristics of the technology (e.g., the complexity of under- standing and using it), the characteristics of the adopter (e.g., level of training in the case of phy- sicians, organizational structure in the case of hos- pitals), and characteristics of the environment (24). Third-party payment, including Medicare payment, is one of the environmental factors affecting medical technology adoption. Third- party payers pay more than 90 percent of all hos- pital expenditures, thus facilitating the adoption of costly hospital technologies (24). The factors affecting the use of technology in- clude physician training, increasing physician specialization, concerns about malpractice suits, industry promotion of its products, the organiza- tion of medical care, and payment for medical services (24). Unfortunately, however, the degree of their influence is not known. How physicians behave in their use of medical technologies and how payment methods and other characteristics of the health care system influence their behavior are questions addressed in chapter 7. OVERVIEW OF MEDICARE POLICIES THAT AFFECT MEDICAL TECHNOLOGY The principal intent of the 1965 legislation that established Medicare under Title XVIII of the Social Security Act (Public Law 89-97) was to in- crease elderly persons’ access to medical services by removing financial barriers to such services, particularly to needed hospitalization (317). In 1965, there was less concern about the cost of the services than there was about the problems of ac- cess, primarily because there was little reason for concern. After the inception of the program, how- ever, the costs of Medicare escalated dramatically, as did all health care costs. Thus, most of the 26 ¢ Medical Technology and Costs of the Medicare Program Medicare legislation enacted after 1965, with the exception of laws increasing the numbers and kinds of populations eligible for Medicare (see sec- tion on eligibility), has been passed with the in- tention of holding down Medicare's cost increases (see table 1). Efforts to control costs to date have met with little success. One of the factors that contributes to Medicare costs and to health care costs as a whole is the adoption and use of medical technology. How the adoption and use of medical technology influences the cost of the Medicare program will be discussed in chapter 3. This chapter provides an overview of Medicare's eligibility, benefits, and payment (including beneficiary cost-sharing) policies and briefly describes the influence of each of the pol- icies on the adoption and use of medical tech- nology .* Eligibility To increase elderly persons’ access to main- stream health care services, the Medicare law mandated eligibility for insurance benefits, in- cluding specific technologies, for most Americans 65 years and over. Eligibility was extended to dis- abled persons and most persons with end-stage renal disease (ESRD) on July 1, 1973, by the Social “As noted in ch. 1, “medical services” is often used interchangeably with “medical technology” in this report. Table 1.—Limitations Placed on Medicare Payment Levels Public Law 92-603 (Social Security Amendments of 1972): * Limits on hospital routine operating costs * Use of a Medicare Economic Index in limiting rises in physician fees Authority for prospective reimbursement experiments A Professional Standards Review Organization to review medical necessity of services Public Law 97-35 (Omnibus Reconciliation Act of 1981): * Limitation on Part B premium increases suspended for 1 year so that premiums could be increased to comprise 25 percent of Part B costs Public Law 97-248 (Tax Equity and Fiscal Responsibility Act of 1982): * Per case limit on operating costs in hospitals, with potential to keep some of the savings as an incentive * Extension of limits on hospital routine operating costs to ancillary services Public Law 98-21 (Social Security Amendments of 1983): * Diagnosis Related Groups (DRGs) for hospital payment SOURCE: Office of Technology Assessment. Security Amendments of 1972 (Public Law 92- 603). Largely because of increases in the number of Americans who are 65 and over, the size of the Medicare population has increased substantially since the program's inception (see table 2). This trend is expected to continue. As of July 1, 1982, Medicare beneficiaries num- bered close to 30 million people, about 12.7 per- cent of the U.S. population. Because of the size and characteristics of the population eligible for Medicare benefits, there is a substantial market for medical technology. By definition, Medicare enrollees are either aged or disabled and thus are disproportionately high users of health services in general and of medical technology in particu- lar (see ch. 3). Elderly people represent 90 per- cent of Medicare beneficiaries (see table 2). Chronic conditions, most often conditions of mid- dle and old age, require medical services for long periods of time (273). The elderly population visits physicians and uses hospitals and nursing homes (organizational medical technologies) much more often than the younger population (194). In 1982, for example, people aged 65 or over represented only 11 percent of the noninstitutionalized popula- tion but accounted for 29.8 percent of the hospi- tal short-stay days of care (408). The older the elderly individual, the more health care services are provided, particularly hospitalization and skilled nursing care (328). And the proportion of older individuals in Medicare's elderly population is increasing. In 1966, 37 percent of Medicare's elderly enrollees were 75 years or older; in 1981, however, the figure was 41 percent (328). Disabled people represent only 11 percent of Medicare enrollees (see table 2). Nevertheless, their eligibility for Medicare benefits has affected Medicare expenditures for services (328). The Congressional Budget Office estimates that in 1984, Medicare payment for a disabled person will be $2,136, while payment for an elderly enrollee will be $1,773 (328). The patterns of use of health care services by the disabled, however, have not been studied (328). People with ESRD require some form of dialysis or kidney tranplants to prolong their lives. Medi- care’s ESRD population represents 0.26 percent of Medicare enrollees (see table 2). An estimated 93 percent of the U.S. population with ESRD is Ch. 2—Medicare Policies Affecting Medical Technology * 27 Table 2.—Number of Elderly and Disabled Beneficiaries Enrolled in Medicare by Type of Coverage, Selected Years From 1966 to 1982 Number of Total number of Number of Number of elderly and disabled Medicare elderly? disabled® beneficiaries Enrollment year? beneficiaries beneficiaries beneficiaries with ESRD? HOBO: ix; wi & 2orw 3 won «ere 19,108,822 19,108,822 — — 1973 oe 23,545,363 21,814,825 1,730,538 NAS 1978 io vo cant mR E 24,201,042 22,272,920 1,928,122 18,564 1978 viv musmmrrminns 27,858,742 24,947,954 2,910,788 60,608 1982 ois 29,494,219 26,539,994 2,954,225 76,117 agnroliment year begins July 1. ball beneficiaries aged 65 and over, including those with end-stage renal disease. CAll beneficiaries under age 65, including those with end-stage renal disease. dEnd-stage renal disease. eNA—information not available. SOURCE: U.S. Department of Health and Human Services, Health Care Financing Administration, 1966-1979 Data Notes: Per- sons Enrolled for Medicare, 1979, HCFA publication No. 03079 (Baltimore, Md.: HCFA, January 1981); and H. A. Silver- man, Medicare Program Statistics Branch, Health Care Financing Administration, personal communication, August 1983. enrolled in Medicare (195). Thus, Medicare pol- icies can be clearly identified as a major influence on the diffusion of the technologies used in the treatment of this disease. The effects of Medicare benefits and payment policies in this area are dis- cussed further below. Benefits Title XVIII of the Social Security Act specifies the broad categories of benefits for which the Medicare program will pay under the two parts: Part A, Hospital Insurance, and Part B, Sup- plementary Medical Insurance.® Part As primary purpose is to provide insurance against the costs of inpatient hospital care. Other benefits include payment for post-hospital extended care services, home health services, and, as of April 1, 1982, inpatient alcohol detoxification services (see table 3). Part B covers medically necessary physician services, outpatient hospital services, outpatient physical therapy and speech pathology services, and various other limited ambulatory services and supplies, such as prosthetic devices and durable medical equipment (see table 3). Part B also cov- ers home health services for those Medicare ben- eficiaries who have Part B coverage only. Part A is an entitlement program and is available with- sThe Medicare law specifies broad categories of services, such as hospital inpatient services, that the program will pay for or “cover.” It also lists a number of specific services that it will not cover. HCFA or Medicare contractors decide the coverage status of particular tech- nologies not mentioned in the legislation. For a full discussion of coverage policy for such specific technologies, see ch. 5. out payment of a premium to those eligible. Par- ticipation in Part B is voluntary and requires payment of a monthly premium.” Except for in- dividuals who choose not to participate in Part B, premiums are deducted automatically from social security checks. In 1982, 99 percent of the elderly and 92 percent of the disabled people en- rolled in Part A were also enrolled in Part B (328). Although Medicare pays for a wide variety of services in a variety of settings, Medicare's bene- fit package is concentrated primarily on acute care technologies provided in institutional settings, particularly those provided as inpatient hospital services. Of Medicare's $52.2 billion in payments for 1982, $34.6 billion (66.3 percent) was for in- patient hospital services (135,151). In 1978, Medi- care paid for almost 75 percent of the elderly’s hospital bills; other public sources paid for almost 13 percent, and the remaining 12 percent was paid for by private health insurance (7 percent) or directly by the patient (5 percent) (124). Medi- care’s impact on hospital use can be seen from examining hospital discharge rates. From 1965 to 1982, the discharge rate for persons 65 years and over (i.e., Medicare beneficiaries) from acute care hospitals increased 36 percent. The discharge rates for other age groups during the same period, how- ¢Individuals not eligible for Part A include those who work for a nonprofit organization that has chosen not to join Social Secu- rity, those who work for a foreign employer overseas, the Presi- dent of the United States, and others. The Part B premium was $14.60/month as of Jan. 1, 1984, and is due to increase on Jan. 1, 1985. 28 * Medical Technology and Costs of the Medicare Program Table 3.—Medicare Benefits and Limitations, as of January 1984 Kind of care Medicare pays Beneficiary pays Comments Part A Hospitalization Psychiatric... .o.ouvus igen cnn vs Skilled nursing facility®. ....... Home health services Part B Home health services ......... Physician and other medical services Chiropractors’ services Most routine foot care......... Dental care DENIUIeS. . cc suis 2s wnins vin « Hearing and eye exams ........ Eyeglasses and hearing aids .. Routine physical exams Prosthetic devices Durable medical equipment .... Medical supplies.............. Blood Outpatient mental illness Outpatient physical therapy or speech pathology ............ End-stage renal disease treatment Hospice 61-90 days 91-150 days After 150 days—no coverage Same as hospitalization 1-20 days 21-100 days After 100 days—no coverage Unlimited visits Reasonable costs Unlimited visits Reasonable costs 80% of approved charges after deductible is met Pneumococcal vaccine and those required for treatment and ordered by physician Manual manipulation of the spine Nothing Jaw surgery and setting Nothing Nothing .Nothing Nothing Those needed to substitute for an internal body organ; i.e., heart prostheses. Also artificial limbs and eyes, arms, legs, back, and neck braces If rented, help with approved charges; if bought, monthly payments until Medicare's share is paid or equipment no longer necessary. If equipment is for long-term use, payment is made in a lump sum Dressings, splints, and casts For all but first three pints $250/yr As treatment: In doctor's office, 80% of approved charges after deductible From physical therapist, $400/yr maximum From clinic, home health agency, or other agencies, 80% of approved charges after deductible 80% of approved charges: Hospital outpatient dialysis: $131/treatment Independent clinic dialysis: $127/treatment Home dialysis: $127/treatment Prospectively per day, the following: —Routine home care............. $46.25 —Continuous home care: Total cont. care rate. .......... $358.67 For 8 hrs. cont. care ........... 119.56 Hourlyrate..................... 14.94 —Inpatient respite rate ........... 55.33 —General inpatient care rate . . . . .. 271.00 Initial deductible ($356) Daily copayment ($89) Daily copayment ($178) Same as hospitalization Nothing Daily copayment ($45) Nothing SMI basic premium=$14.60/mo. Nothing Initial yearly deductible ($75) Deductible does not apply All other costs All costs All other costs Total costs Total costs Total costs Total costs All other costs All other costs (i.e., common first aid supplies) First three pints or replacement All costs above $250/yr $75 deductible 20% coinsurance All costs above $400/yr $75 deductible 20% coinsurance $75 deductible, 20% coinsurance 5% of cost to program for: —Drugs and biologicals (per drug) (not to exceed $5 per prescription) —Inpatient respite care (per day) (not to exceed inpatient hospital deductible) Deductible and copayments are adjusted annually 150 days of coverage includes a lifetime reserve of 60 days that can be used only once Only 190 days of coverage, usable only once Beneficiary must be eligible for Part A Beneficiary eligible for Part B only Coverage ends 12 months after the month maintenance dialysis treatment stops or 36 months after month of kidney transplant Beneficiary also pays Medicare deductibles and coinsurance payments, and the difference between reasonable and actual charges on unassigned claims for covered services other than hospice care. Hospice coverage consists of two 90-day periods and one 30-day period, to be taken in that order aCoverage for skilled nursing facility services was limited to only 9.1 enrollees bCoverage for home health agency services was limited to only 33.6 enrollees CRespite care is defined in the regulations as short-term (not reimbursed after 5 days) inpatient care provided to the individual only when necessary to relieve the family members or other persons caring for the individual during period of hospice election. Data Sources: 1. U.S. Department of Health and Human Services, Health Care Financin (Baltimore, Md.: Social Sec urity Administration, January 1983). 2. Federal Register, 48(243):56008, Dec. 16, 1983. 3. Federal Register, 48:21254-2 4. U.S. Department of Health and Human Services, Health Care Financing Administration, Division of Public Information, personal communication, Jan. 23, 1984. 1291, May 11, 1983. SOURCE: Office of Technology Assessment. per 1,000 in 1980 (Medicare and Medicaid Data Book, 1983). per 1,000 in 1980 (Medicare and Medicaid Data Book, 1983). g Administration, Your Medicare Handbook, January 1983 edition, SSA publication No. 05-10050 Ch. 2—Medicare Policies Affecting Medical Technology ¢ 29 ever, either decreased, remained the same, or in- creased slightly (see ch. 3). Medicare's benefit policy has favored the de- velopment of some technologies in an inpatient setting. For example, Medicare coverage empha- sizes treatment for alcoholism provided in tradi- tional acute care institutional settings, rather than that provided in the freestanding inpatient alco- holism facilities that have developed over the last 15 years or in outpatient alcoholism treatment centers. As a result, a substantial network of in- patient alcoholism treatment facilities has devel- oped, despite evidence that outpatient treatment may be as effective and certainly is less costly (348).° Since Medicare's enactment, practically all of the Nation's elderly have gained access to serv- ices provided in hospitals and to a lesser extent to services provided in ambulatory settings (81). Some analysts contend that the quality of life for Medicare beneficiaries has improved with access to inpatient hospital services such as surgical serv- ices. The frequency of certain surgical procedures that improve the functional status of the elderly, such as cataract operations and arthroplasty, has increased dramatically as a result of Medicare (91,92). Furthermore, U.S. mortality rates, which had been reaching a plateau in the early 1960's, resumed their decline in the 1970's (406). From 1968 to 1977, death rates for elderly men declined at an average annual rate of 1.5 percent and for elderly women at an average annual rate of 2.3 percent (406). The decline in U.S. death rates is almost twice the decline in Canada and European rates over the same period (81). The beginning of the sharp decline in mortality rates among older people in the United States was coincident with Medicare's enactment. Some analysts have attrib- uted the decline to improved medical care treat- ment (284), and more specifically, to the services available under the Medicare program (92). This example illustrates the increase in hospital services for elderly people relative to the general population. It does not, however, demonstrate an increase in hospital services relative to outpatient care. The Effectiveness and Costs of Alcoholism Treatment, OTA Health Technology Case Study #22 (348), analyzes the reimburse- ment issues concerning the treatment of alcoholism in depth. 10 Arthroplasty is plastic surgery of a joint or joints, or the forma- tion of movable joints. At the same time that the mortality rates from many diseases, including heart disease and dia- betes, have been falling among the elderly, how- ever, the prevalence of such diseases has increased. The reason may be that improved medical treat- ment of acute episodes of these conditions de- creases mortality rates and thereby increases the prevalence of these conditions as chronic illnesses. The prevalence of other major illnesses—athero- sclerosis, cancer, emphysema, cirrhosis, osteo- arthritis—has also increased among the elderly (128). Chronic conditions require long-term care more than episodic acute care. Long-term care for the elderly requires social services as well as health services. Indeed, there is a school of thought that considers medical services to have a subsidiary role in long-term care (137). Thus, the most appropriate role of Medicare, a health insurance program, with respect to chronic conditions and long-term care is a matter of debate. Medicare's benefits do not include all health- related services. Section 1862 of Title XVIII of the Social Security Act specifies that, notwithstanding other provisions of the title, “no payment may be made under part A or part B for any expenses incurred for items or services . . . which are not reasonable and necessary for the diagnosis or treatment of illness or injury or to improve the functioning of a malformed body member.” Sec- tion 1862 specifically excludes Medicare coverage of many preventive services (including routine physical checkups; eyeglasses and examinations for the purpose of prescribing or fitting eyeglasses; hearing aids and examinations for hearing aids; and immunizations other than pneumococcal vac- cinations, which were added as a benefit in 1981), custodial care, most cosmetic surgery and dental services (except for special cases that require hospitalization), personal comfort items, and or- thopedic shoes. In part because of funding disapprovals by Medicare and other third-party payers, assistive communication devices, used in the rehabilitation of persons disabled by severe speech impairments, are little used (351). Although Medicare covers such devices under Part A (i.e., for use only while 11Gee the forthcoming OTA report Technology and Aging in America (summer 1984) for a detailed discussion of long-term care, including technology and financing issues. 30 e Medical Technology and Costs of the Medicare Program the person is actually in the hospital) so as to enable nonvocal patients to communicate with hospital and skilled nursing facility staff (351), it does not cover these devices under Part B because of an administrative decision that the devices are not prosthetic devices needed for the functioning of a malformed body member. Not only has sales volume been lower than anticipated, but innova- tions in the field appear to have been held back, in part, by the lack of coverage by Medicare and other insurers (351). Clearly, many medical technologies have been developed and diffused without Medicare cover- age. One example is eyeglasses and vision aids. In 1977, among those 65 years of age and older, 193 per 1,000 population purchased glasses or contact lenses or had them repaired. The major source of payment, 78.8 percent, was the family (394). Whether there would be greater adoption and utilization by the elderly of vision aids such as eyeglasses and contact lenses if the technologies were covered by Medicare is not known. Payment For many years, reimbursement by Medicare has been based on reasonable costs in the case of hospitals and other institutional providers and reasonable charges on a fee-for-service basis in the case of physicians and other noninstitutional sup- pliers of services. Under these payment methods, providers receive greater reimbursement when they use more technology, so they have little fi- nancial incentive to use technologies judiciously, with consideration of their costs and benefits. In particular, cost-based reimbursement policy has been instrumental in facilitating the acquisition by hospitals of sophisticated, capital-intensive technology (24). In an attempt to change the financial incentives to provide hospital care, the Social Security Amendments of 1983 (Public Law 98-21) changed the basis of Medicare payment for inpatient hospi- tal services from retrospective cost-based re- imbursement to prospective payment based on Diagnosis Related Groups (DRGs).?? The DRG 12Gee discussion section for an overview and ch. 6 for a detailed discussion of Medicare's current hospital payment system and its expected effects on medical technology. prospective payment system does not apply to all hospitals or to many other segments of the health care delivery system. The system is still being im- plemented, and it is too early to evaluate its ef- fects. Medicare's method of paying physicians and other noninstitutional suppliers of services on the basis of reasonable charges has changed little since Medicare was enacted. Payment for Hospital Services Box A describes Medicare's traditional method of payment for hospital services. The Medicare law passed in 1965 specified that Medicare pay hospitals the reasonable cost of providing serv- ices to beneficiaries. The method or methods to be used in determining reasonable cost were left to administrative decisions. Since Medicare pur- chased only a portion of each hospital's costs, the costs attributable to Medicare patients (“allowable costs”) had to be calculated. Like Blue Cross, Medicare adopted a method that allowed hospi- tals considerable discretion in calculating at- tributable costs (104). Under Medicare's cost- based payment method applied to inpatient serv- ices, there was no financial reason for a hospital not to spend money on technology, because it was assured of reimbursement. There have been a number of changes in the rules and guidelines in attempts to moderate Medi- care’s hospital expenditures, but such attempts have had only qualified success. Until 1982, the single most important innovation in the Medicare hospital reimbursement system was the 1974 im- plementation of “Section 223" limits (see Box A). Although the objectives of Section 223 were to moderate the rate of increase in Medicare's hos- pital outlays (362), the results were disappoint- ing. The new limits affected only a few high-cost hospitals and were relatively easy to circumvent by reclassifying formerly routine services into ancillary (and therefore chargeable) items. Sec- tion 223 limits may also have encouraged the spread of intensive care unit beds (32). The limits may have encouraged hospitals to increase lengths of stay. Finally, the limits never pertained to cap- ital costs (depreciation, interest, and in the case of for-profit hospitals, return on equity). Hospi- tals were paid depreciation based on actual his- torical expenditures and interest payments as in- curred. Ch. 2—Medicare Policies Affecting Medical Technology © 31 -_— Cost-based treatment of capital costs, which re- mains to this day, provides no disincentive for hospitals to adopt the capital-intensive technol- ogies that they wish to adopt. The inclusion of depreciation cost as an allowable cost means that the cost of the equipment and buildings is passed through the hospital to the Medicare program. The treatment of capital costs by Medicare has also facilitated the ability of hospitals to borrow capital with little risk. Low-cost borrowing has made it easier for hospitals to purchase buildings and equipment than it would be if the hospitals found it necessary to generate capital for such ex- penditures internally. Medicare's traditional ap- proach to capital investment is seen in another capital-related provision. Profitmaking hospitals can include a reasonable return on equity (capi- tal invested and used in providing patient care) as an allowable cost. Thus, with limited finan- cial constraints, it has been to some hospitals’ advantage to increase their technological capabil- ity in response to demand rather than through a process of assessment of need, however defined. Until very recently, the organization of the hos- pital industry has also provided incentives for technology adoption. 32 e Medical Technology and Costs of the Medicare Program In August 1982, Congress made some major revisions in Medicare's traditional cost-based reimbursement system for hospitals by passing the Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) (Public Law 97-248). Among other things, TEFRA imposed a hospital-specific limit on the amount of inpatient operating costs per case that Medicare would reimburse (see Box B). Medicare's inpatient hospital payment method was changed more dramatically when the Social Security Amend- ments of 1983 (Public Law 98-21) mandated the phasing in over a 3-year period beginning in Oc- tober 1983 of a new inpatient hospital payment method based on a national set of per case prices for patients in 470 separate DRGs (see Box B).? The DRG prices will apply to virtually all short- term acute-care general hospitals in the United States. Capital costs will continue to be paid as under the old system until the end of the transi- tion period. At the time Congress passed the 1983 amendments, it contemplated, but did not specify the method for, the incorporation of payment for capital into the DRG pricing system. Payment for Physicians’ Services! Except for the imposition of minor restraints on the rate of increase of physicians’ upper limit payment levels, Medicare's method of paying phy- sicians and other noninstitutional suppliers of services for charges on a fee-for-service basis has changed little since Medicare was enacted. Box C provides a description of this payment method. Most charge-based payment by Medicare—over 70 percent in 1982—is made to physicians for care provided in ambulatory and institutional settings (338). Under Medicare's current method of payment for physician services, most of the charge for a service is passed through to the beneficiary, who, in turn, is protected from some part of the charge by Medicare. This method of physician payment 13See OTA's technical memorandum entitled Diagnosis Related Groups (DRGs) and the Medicare Program: Implications for Medi- cal Technology (343) for further discussion of Medicare's new hos- pital payment system. 14See ch. 7 for a detailed discussion of physician payment and medical technology. 15The effect of Medicare's beneficiary cost-sharing policy on tech- nology is discussed in a following section. encourages the use of medical technology by pro- viders, particularly if the charge-to-cost ratio is high. When payment is based on a fee paid for each service, physicians’ revenues are determined to a large extent by the number and intensity of services delivered and the fee received for each service. The use of technology by fee-for-service physicians is sensitive to the additional revenue they receive (229). Medicare's method of paying physicians also gives them an opportunity to acquire and use ex- pensive, technically sophisticated technology in their offices. When a technology is provided as an office service, the physician's capital invest- ment in the technology may be incorporated into the charge to the patient for the service. In 1980, 18.9 percent of the 1,471 operational computed tomography (CT) scanners, were in private of- fices and clinics (347). At the time, scanners were expensive. Even the so-called lower priced models of head scanners listed from $95,000 to $200,000 (347). Although no payment method automatically provides incentives for one technology over another, any payment method can be structured to do so. The payment levels that Medicare and other third-party payers have established under the charge-based method of paying physicians provide incentives for the use of new and often expensive technologies. As noted in Box C, Med- icare carriers refer to relative value systems when establishing charges for new technologies.’® In most instances, the reimbursable charge was es- tablished at an early point in the history of the technology (52). Although later technological ad- vances and higher rates of utilization may have substantially reduced the time, judgnient, skill, and cost required to use the technology, this change is not reflected in the physicians’ fee level or Medicare's reimbursement level (235). Further- more, the payment level that Medicare has estab- lished for complex and expensive technology is 1¢Relative value systems establish a quantitative but nonmonetary scale on the worth of one procedure as compared to all other pro- cedures (315). For example, if administration of a measles vaccine had a relative value of 2.2 and the conversion factor is 10, then the payer would pay the physician a maximum of $22 for the immuni- zation. Ch. 2—Medicare Policies Affecting Medical Technology * 33 Box B.—Prospective Per Case Payment to Hospitals Under Medicare The Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) (Public Law 97-248) made major revisions in Medicare's retrospective cost-based hospital reimbursement system. In addition to extend- ing the existing “Section 223 limits” (see Box A) to include ancillary and special care unit operating costs, TEFRA imposed a hospital-specific maximum limit on the amount of inpatient operating costs per case that would be reimbursed. The new approach, which became effective in October 1982, has two key elements: 1) the limit is determined either by the hospital's own per case cost in a previous year or the average per case cost of similar hospitals; and 2) the hospital stands to gain a small portion of per case savings it can generate. TEFRA put no limit on capital costs (depreciation and interest), direct teaching expenses, or outpatient services. These remained “pass-through” items, i.e., items not subject to the new system's controls. A more sweeping revision of Medicare's hospital payment system was signed into law in April 1983. The Social Security Amendments of 1983 (Public Law 98-21) mandated the phasing in over a 3-year period of a prospective payment system for inpatient hospital services. Payment is to be based on a national set of per case prices for patients in 470 Diagnosis Related Groups (DRGs). DRGs are a set of patient classes developed to reflect differences in resource needs among different kinds of patients. Several types of hospitals (psychiatric, long-term, children’s, and rehabilitation hospitals) and hospital units (psychiatric and rehabilitation units operating as distinct parts of acute care hospitals) are excluded from the prospective payment system and will continue to be reimbursed on the basis of reasonable costs. Capital costs and the costs of direct medical education remain pass-through items. Under the DRG system, Medicare payment is made at a predetermined, specific rate for each discharge. During the 3-year transition period, which began in October 1983, a declining portion of the total prospective rate is to be based on hospital's historical costs in a given base year, and a gradually increasing portion is to be based on a blend of federally determined regional and national rates. Begin- ning in the fourth year, Medicare payment for inpatient care will be based on a set of national DRG rates. The price for a DRG will be adjusted for the hospital's urban or rural location and area wage rate. For 1984 and 1985, the prospective payment system must be “budget neutral,” i.e., payments may not be greater than nor less than the payments that would have been paid under TEFRA. The DRG prospective payment system applies to-all inpatient discharges from short-term acute care general hospitals in the United States except for a small number of cases (set by statute as 6 percent of the total) with unusually long lengths of stay or high charges. The rate of payment for these “outlier” cases will be increased by the estimated cost of care during the extended stay. The initial set of DRG prices is based on the 1981 average inpatient operating cost per case for each DRG in a 20-percent sample of Medicare claims. The law requires that the DRG prices be updated regularly in two ways. First, an overall annual rate of increase is applied to all DRGs to keep pace with the general level of inflation and rate of technological change in the economy. Second, the relative weights (i.e., the ratio of one weight to another) must be assessed and recalibrated at least once every 4 years, with the first recalibration scheduled for October 1985. The recalibration must reflect changes in treatment patterns, technology, and other factors that alter the relative use of hospital resources among DRGs. A Prospective Payment Assessment Commission established by the law is responsible for making rec- ommendations regarding the annual payment increase and recalibration and for evaluating any such adjustments made by the Secretary of the Department of Health and Human Services. Public Law 98-21 requires the Medicare program to participate in any State-legislated alternative prospective payment program that: covers at least 75 percent of the State's population; makes provi- sions for competitive health plans; assures the Federal Government that access to hospital care for Medi- care and Medicaid beneficiaries will not decline; and assures the Federal Government that hospital costs will not be higher under the State program. Thus, it encourages States to experiment with hospital pay- ment systems that cover third-party payers in addition to Medicare and differ from DRG payment. 34 e Medical Technology and Costs of the Medicare Program usually disproportionately high. Relative value scales place higher values on “technology-ori- ented” procedures and devices than on other serv- ices, such as cognitive procedures and office visits (235). Payment for Treatment of End-Stage Renal Disease The Social Security Amendments of 1972 man- dated payment under Medicare's ESRD program for both hemodialysis and kidney transplantation. Before the ESRD program was established in 1973, there were few freestanding dialysis centers, and most hemodialysis was performed in hospitals or in patients’ homes. The original Medicare regu- lations pertaining to ESRD included financial dis- incentives for home dialysis as compared to fa- cility dialysis.'” By 1977, there were 895 approved For example, out-of-pocket costs were required for home dialysis supplies and equipment, and reimbursement was not provided for the services of a home dialysis assistant nor for renting equipment, ordering supplies, and other bookkeeping requirements. Home dial- ysis patients also incurred out-of-pocket costs for home modifica- tion and higher electric and water bills. Ch. 2—Medicare Policies Affecting Medical Technology ® 35 dialysis centers in the ESRD program (262), and the percentage of patients on home dialysis had E30 decreased significantly (see table 4). Some of the Ti PN - decrease in home dialysis may have been due to ' : the stresses on family life, which led patients to use facility dialysis when Medicare coverage be- came available. Other factors contributing to the increased use of facility dialysis included the per- sonal philosophy of the physician or hospital treating the patient, increased age and morbidity of dialysis patients that reduced their suitability for home dialysis, and the for-profit status of a significant percentage of dialysis facilities. The number of patients receiving kidney trans- plants increased strikingly in 1973 (see table 4). After 1973, the number grew at a slower pace and then plateaued between 1977 and 1978 because of the lack of improvement in graft success rates, a decreased donor pool, and financial disincen- tives for undergoing transplantation that were in the Medicare regulations. When the financial dis- incentives, including termination of benefits the 12th month after transplant surgery, were re- moved in 1978, the number of transplants started to increase (359). Photo credit: National Kidney Foundation, Washington, D.C. Payment and coverage policies for end-stage renal disease Escalating costs of Medicar e's ESRD program technologies have fostered their use were addressed in two revisions to the original Table 4.—ESRD Patient Population, 1972 to 1982 Number of Percentage of Number of hemodialysis patients on Year kidney transplants? patients home dialysis? 1972 1,993 (2,852) 10,000 40% 1973. .....cosaminnisnismisg 3,017 11,000 35.9 TOT. inne smn immerse www 3,190 18,875 327 1975 3,730 22,000 28 1978 .civsvimisvninbe amass 3,504 30,131 23.7 (13) VOTT wus nip + 5 mi # 00m 2 0 4 sw ww 3,973 32,435 11.6 (20) 1978 3,949 36,463 12.4 VO79 ....onsomas wns ns ans vw 4,271 45,565 13.0 (10) 1980 oii 4,697 50,000 NAP 1981 4,885 NA 1982 ovo 5,358 NA aNumbers in parenthesis reflect conflicting reports in the literature. bNA—information not available. SOURCE: Office of Technology Assessment, Strategies for Medical Technology Assessment, OTA-H-181 (Washington, D.C. U.S. Government Printing Office, September 1982); and National Kidney Foundation, Washington, D.C., personal com- munication, Feb. 9, 1984. 36 e Medical Technology and Costs of the Medicare Program law. The End-Stage Renal Disease Program Amend- ments of 1978 (Public Law 95-292) established a prospective reimbursement method to encourage efficiency and cost effectiveness. To encourage home dialysis by eliminating the 20-percent co- insurance requirement and to avoid high equip- ment rental payments, one of the provisions pro- vided for reimbursement by Medicare of the full costs (100 percent) of home dialysis equipment, installation, maintenance, and repair. The Omnibus Budget Reconciliation Act of 1981 (Public Law 97-35) discontinued 100-percent reimbursement for home dialysis equipment but called for further changes to promote home dial- ysis. Under regulations implementing the law, each dialysis facility receives a certain payment rate per treatment, adjusted for geographic wage differences, regardless of whether the treatment is furnished in the facility or supervised in the pa- tient’s home. Dialysis facilities have to accept the prospective payment rate as payment in full. Phy- sicians receive a monthly capitation payment that is equal for home dialysis and facility dialysis (111). Financial incentives favoring one dialysis loca- tion over another are related to the difference be- tween reimbursement rates and unit costs. The dif- ference between reimbursement rates and unit costs creates strong disincentives for performing hemodialysis in hospital dialysis centers, moderate incentives to perform hemodialysis in independ- ent centers, and very strong incentives for home dialysis supervised by either hospital or independ- ent centers. The strong incentive for home hemo- dialysis could be moderated somewhat if unit costs rise as a result of the need for more home health aides for sicker patients or those without much family support. Average physician capitation fees under the 1983 composite rate formula will increase from $1,848 per year to $2,208 per year (+ 19 percent) for home dialysis, and decrease from $2,640 per year to $2,208 per year (— 16 percent) for center dialysis. Thus, changes in capitation rates for phy- sician supervision of dialysis also heavily favor home dialysis over center dialysis (344). Beneficiary Cost-Sharing “In one sense . . . Medicare can be said to have been designed to increase utilization” (318). Yet cost-sharing provisions were included in the original Medicare legislation as a possible mod- erating influence on the unnecessary utilization of services (322). Box D describes the beneficiary cost-sharing provisions of the Medicare program. The premise behind deductibles, copayments, and coinsurance is that price deters patients from seeking care and thereby lessens the use of unnec- essary services. Furthermore, once beneficiaries decide to seek care, price is considered to influ- ence patients and providers to choose less expen- sive technologies. Premium payments, another form of cost- sharing, are not considered an obstacle to the use of services (28,253). Premium cost is too far removed from the use of a technology to affect patients’ or physicians’ behavior at the time of its use. A number of studies of populations not in the Medicare program suggest that cost-sharing re- strains the use of medical services (30,138,244, 245,246,255,259,260,261,299).'®* When benefici- aries must immediately pay for part of the cost of additional services, they choose to use fewer services than when fully insured. Low-income groups, in particular, are deterred from using serv- ices as a result of cost-sharing (30,299). Early results of a large, well-designed and ex- ecuted study—the Rand National Health Insur- ance Study (247)—substantiate the above find- ings. Individuals enrolled in health insurance plans with high coinsurance rates (50 or 95 percent, sim- ilar to income-related catastrophic coverage) were less likely than individuals enrolled in plans with no coinsurance or a low coinsurance rate (25 per- cent) to visit a physician and to be admitted to a hospital. Individuals in plans with higher co- For a general discussion of cost-sharing and the adoption and use of medical technology, see the OTA assessment Medical Tech- nology Under Proposals To Increase Competition in Health Care (355). Ch. 2—Medicare Policies Affecting Medical Technology © 37 Box D.—Beneficiary Cost-Sharing Under Medicare Beneficiaries’ expenses in the Medicare program consist of deductibles and copayments under Part A (Hospital Insurance) and premium payments, deductibles, and coinsurance under Part B (Supplemen- tary Medical Insurance). Those over 65 who are no t automatically entitled to Medicare (e.g., those who work for a nonprofit organization that has chosen not to join Social Security) can participate by month- ly payments of the actuarial cost of coverage. Part A deductibles and copayments are calculated on the basis of a benefit period (a benefit period begins when a beneficiary enters a hospital and ends when the beneficiary has been out of a hospital or skilled nursing facility for 60 days in a row). During each benefit period, Part A will pay for 90 days of inpatient hospital care of which the beneficiary has to pay the first $356. After 60 days of inpatient hospital care, the beneficiary is required to pay a daily copayment of $89 until the 90th day of care. If more than 90 days of care are required in any o lifetime reserve of 60 days that requires a copaymen ne benefit period, the beneficiary can draw upon a t of $178 per day. Part A also requires a beneficiary copayment of $45 per day for the 21st through 100th day in a skilled nursing facility. Under Part B, the beneficiary is responsible for the first $75 of approved charges in a calendar year and coinsurance of 20 percent for the remainder of approved charges. If a physician does not accept assignment (agree to accept the level of reimbursement calculated by Medicare in exchange for direct payment of Medicare's 80-percent share), the beneficiary is financially responsible for the difference be- tween the charge billed by the physician and the allowable charge determined by Medicare. In 1966, premiums contributed half of Part B revenues, while general revenues subsidized the other half. Subsequent legislation limited increases in the premiums to no more than the percentage increase in Social Security cash benefits. By 1978, the percent contribution of premiums to meet Part B program costs had fallen to below 25 percent (134). The Omnibus Budget Reconciliation Act of 1981 (Public Law 97-35) and the Social Security Amend- ments of 1983 (Public Law 98-21) suspended the limitation on Part B basic premium increases for the period between July 1, 1983 and January 1, 1984. During this period, premiums increased so that they represent 25 percent of program costs. Premiums rose from $13.50/month on July 1, 1983, to $14.60/ month on January 1, 1984. insurance rates also had a lower number of phy- sician visits. There was no significant difference in hospital spending per hospital admission. Applying results of available studies of cost- sharing on different age and sex groups to the Medicare beneficiaries may not be appropriate. There is evidence that the influence of cost-sharing on hospital use is sensitive to the age and sex of the patient (243). There are crucial differences in health status and health practices between the Medicare population and others. Not surprisingly, even before Medicare was enacted, the elderly used hospitals more than others. For example, from July 1962 to June 1964, those 65 years and older represented 9 percent of the population, but used over 25 percent of hospital days (318). 25-337 0 - 84 - 4 Little empirical evidence is available on the ef- fects of deductibles, copayments, and coinsurance specifically in the Medicare program on the adop- tion and use of technology. However, a study of the use of supplementary health insurance by Medicare beneficiaries provides some insight into how Medicare's cost-sharing policy has affected the adoption and use of technology (199). Public and private supplementary (Medigap) health in- surance is used extensively by Medicare benefi- ciaries. In 1976, 63 percent of aged Medicare bene- ficiaries had some form of private supplementary health insurance and 14 percent had public sup- plementary coverage, primarily from Medicaid (6 percent had both public and private supplementa- tion). Only 29 percent had no supplementary insurance. 38 e Medical Technology and Costs of the Medicare Program The study found that supplementary insurance increased the use of both hospital and physician services by elderly Medicare beneficiaries (197). Supplementary health insurance greatly increased the use of inpatient hospital services by elderly Medicare beneficiaries with or without chronic health problems. Most of the gains in utilization of hospital services came from the admission of more people into hospitals rather than from in- creases in length of stay. The investigators sug- gest that the “Part A deductible (approximately equal to the average charge for an inpatient hos- pital day) represents a significant barrier to the utilization of hospital services by the elderly” (199). The effect of cost-sharing under Part B of Medi- care on the use of physicians’ services depended on whether the elderly Medicare beneficiary did or did not have a chronic health problem (approx- imately 78 percent of the elderly Medicare popula- tion have a chronic health problem). Part B cost- sharing provisions did not deter individuals with chronic health problems from seeking health care from physicians (199). On the other hand, the Part DISCUSSION Medicare policies—payment policies, in partic- ular—have influenced the adoption and use of some medical technologies. Cost-based hospital reimbursement, with pass-through for capital ex- penditures, has not discouraged hospitals from purchasing new technologies. Payment for phy- sician services and technologies at hospitals and other health care delivery sites provided financial incentives for their use without careful consider- ation of their impacts on costs. Medicare payment policies generally have assured hospitals that they would be paid for the cost of new technologies. This assurance has had a direct effect on hospitals’ decisions to adopt new technologies. Russell found that adoption of co- balt therapy, for example, was influenced by Med- icare (289). In addition, since hospitals have fewer nonpaying patients since the inception of Medi- care and Medicaid, they are not losing as much B deductible and coinsurance provisions had a decided effect on the use of physicians’ services by elderly Medicare beneficiaries without chronic conditions. Thus, it appears that cost-sharing under Medi- care “leads to significantly lower levels of hospi- tal and physician utilization than would have pre- vailed in the absence of the program's deductibles and coinsurance” by some members of the elderly Medicare population (199). The more fundamen- tal question of whether cost-sharing affects the use of necessary services by the elderly requires health status data. A recent Rand study of nearly 4,000 people found that there were only small dif- ferences at the end of the study between the health status of those people with “free care” (no cost- sharing) and people who were required to pay a portion of their medical bill (various levels of cost- sharing were aggregated for the analysis) (49). However, the study population was limited to in- dividuals between the ages of 14 and 61 without any disability, making the applicability of its find- ings to Medicare beneficiaries questionable (270). money to bad debt, so they are better risks for loans. Thus, the presence of Medicare patients also has an indirect effect on hospitals’ decision- making regarding adoption of medical tech- nologies. The use of medical technologies is largely the responsibility of physicians. Under Medicare poli- cies of retrospective, charge-based reimbursement, physicians have had no financial constraints on the number of such technologies provided. In- stead, they have known that the hospital gained revenue for each test or procedure. Medicare's coverage policy excludes payment for items that are “not reasonable and necessary” for diagno- sis, treatment, or improved functioning of a mal- formed body member. This has allowed physi- cians much flexibility in their medical technology use (see chs. 5 and 7). Under the new Medicare DRG prospective hospital payment system, with Ch. 2—Medicare Policies Affecting Medical Technology © 39 its incentives to reduce ancillary services, hospi- tal administrators will have to work more closely with physicians regarding use of technologies. It is important to remember that physicians have always been important actors in both the adoption and use of medical technologies. In ad- dition to purchasing new technologies for their office practices, physicians are often the ones who suggest the purchase of new technologies to hos- pital administrators or boards of trustees. In their decisions, the administrators or boards may con- sider the importance of the individual physicians in admitting patients and the various specialties competing for the technologies, as well as the cost of the new equipment and its benefits to patients. They also may consider the extent to which the physicians use the technologies already available. The DRG hospital payment system may change the impact of Medicare on technology. Use of hos- pitalization and tests and procedures during hos- pital stays are constrained under the new system. Efforts to control costs during hospitalizations may extend to the adoption of technologies that will lower hospital costs per case. More technol- ogies may be moved out of inpatient settings to ambulatory settings, where Medicare payment has not yet changed. Such movement depends on the development of specific technologies, also (e.g., those used in freestanding ambulatory surgery centers—see ch. 8). Thus, the Medicare program has influenced technology adoption and use. Yet, the strength of this influence has been limited by the fact that it is only one payer among several. Where Medi- care beneficiaries make up a large portion of the patient population, such as in hospitals, Medicare policies have more impact. Medicare's influence with physicians—because physicians are the strongest factor in technology adoption and use decisions—needs to be strengthened in order to contain program costs and to rationalize technol- ogy decisions. Trt CS yim—y ump rms mewn c= a sg ae Fe 3. The Impact of Medical Technology on Medicare Costs There is no gathering the rose without being pricked by the thorn. Bidpai Contents Page IDtrodUCHION ... «coitus ih ss eer aR Res te 43 Medicare Beneficiaries’ Use of Medical Technology .......................... 43 Medical Technology's Aggregate Impact on Medicare Costs .................. 45 The “Intensity of Care” Approach to Measuring Technology's Contribution to Health Costs’... ........ 0... coo ll mes al 47 Other Estimates of Technology's Contribution to Health Care Costs... ...... 51 Conclusions From the Aggregate Studies ............... 6... 00 lds. 52 Selected Medical Technologies and Medicare Costs. ......................... 53 Coronary Artery Bypass Graft Surgery ...... ..... 0b a iilad bala 53 Clmebtidine:, von doa sided Bias s iids « Sn di aE aa 54 Therapeutic ApReresiS . ... ... sini is snr ad oe ase gis aie 56 Pneumococcal Vaceine. i... {ova id oat bd al a 56 Intensive Care Units... ...h.. iil in tvs vm sil To as iis nt a a, 57 Total Parenteral Nutrition. .... .os88 .. ives ido She io i a 58 Hemodialysis for Chronic Renal Failure ........... ovis iia iis 59 CONCIUSIONS ccoiv vis 1s cis + Titers siionttir Sas sv 2 as is vs see ea 60 LIST OF TABLES Table No. Page 5. Discharges From and Days of Care in Short-Stay Non-Federal Inpatient Hospitals by Patient Age, Selected Years From 1973 to 1982. ..... 44 6. Operative Procedures in Short-Stay Hospitals by Patient Age, 1973-80 . . . .. 44 7. Use of Ambulatory Physician Visits With Specified Diagnostic Services by Age, 1077 .«.... 0. i vir od aa me 46 8. Decomposition of Hospital Costs, (1077-82 . cot ire casi 49 9. Summary of Studies of Hospital Cost Inflation .......................... 49 10. Increase in Personal Health Care Expenditures, 1977-82 .................. 50 11. Real Medicare Expenditures per Enrollee, 1977 and 1982 .................. 51 12. Days of Care for Ulcer Disease for Patients Discharged From Short-Stay Hospitals, 1077-79 oi. ee. oor ioe at oh Fe i 55 13. Operations for Ulcer Disease for Patients Discharged From Short-Stay Hospitals, 1977-79... /. ivi vin ss sede svn Bs 55 3. The Impact of Medical Technology on Medicare Costs INTRODUCTION Changes in the kinds of medical technologies available and changes in the patterns of use of technologies already available continually influ- ence health care and Medicare costs—at times moderating cost increases and at times exacer- bating them. As noted in the previous chapter, various factors affect the adoption and use of medical technology. This chapter examines the patterns of medical technology use experienced by the Medicare population compared to the gen- eral population. Its primary purpose, however, is to explore the nature and size of medical tech- nology’s contribution to health care and Medicare costs. How medical technology contributes to health care and Medicare costs is a question that can be addressed either from an aggregate perspective or from the standpoint of particular technologies or classes of technology. The question from the aggregate perspective is whether changes in medical technology use as a whole have raised or reduced health care or Medi- care costs and, if so, by how much. This perspec- tive is useful, because it puts technology's re- 1As noted in ch. 1, OTA defines medical technology as “the drugs, devices, and surgical and medical procedures used in medical care, and the organizational and supportive systems within which such care is provided.” The term “medical services” is often used inter- changeably with “medical technology” in this report. lationship to costs into a policy perspective. Changes in the use of medical technology reflect changes in the behavior of medical decisionmak- ers. Quantitative estimates of technology's aggre- gate contribution to health care costs, therefore, reflect the importance of changes in medical deci- sions, which can be presumed to be influenced by policy, relative to changes in other, less control- lable factors such as population growth or gen- eral wage and price inflation. The aggregate approach is limited, however, because it ignores the patient benefits associated with cost increases or decreases, it does not take into account the underlying reasons for changes in medical decisions or practices, and it does not show that cost-saving or cost-raising changes in technology are not scattered evenly across ill- nesses. In short, it offers no way of knowing whether any particular technology-related rate of change in health care or Medicare costs is too high or too low. Analyzing how specific technologies or classes of technology affect health care or Medicare costs can be more enlightening, particularly when the information that results is combined with data on efficacy and patterns of adoption and use. Anal- yses of the cost implications for Medicare of seven specific medical technologies are provided in this chapter. MEDICARE BENEFICIARIES’ USE OF MEDICAL TECHNOLOGY By definition, Medicare enrollees are either aged or disabled. Furthermore, they are disproportion- ately high users of health care services in general and of medical technology in particular (126). Although a high proportion of health care expend- itures for the elderly is for nursing homes and other long-term care services, it is important to recognize that the elderly are high users of serv- ices provided in hospitals, where medical technol- ogy is concentrated. Age-specific hospital dis- charge rates for selected years from 1973 to 1982 are shown in table 5. Not only has the number 43 44 ¢ Medical Technology and Costs of the Medicare Program Table 5.—Discharges From and Days of Care in Short-Stay Non-Federal Inpatient Hospitals by Patient Age, Selected Years From 1973 to 1982 Discharges per 1,000 population Average annual Percent change Age 1973 1975 1978 1979 1980 1981 1982 percent change 1973 to 1982 <15 ............ 70.8 71.5 68.8 70.8 71.8 72.9 71.2 0.1% 0.6% 1544 ........... 154.4 155.4 155.1 151.8 151.3 148.7 145.0 -0.6 —6.0 4564 ........... 182.3 194.7 193.1 192.4 196.0 195.3 195.5 0.7 72 =65............ 341.8 359.3 381.9 361.5 405.2 396.5 398.8 1.5 16.7 All ages ....... 154.0 158.8 159.8 156.9 161.9 160.2 167.9 0.9 9.0 Days of care per 1,000 population Percent change Age 1973 1978 1979 1980 1973 to 1980 <15 321.9 304.8 314.7 316.4 1.7% 16-44 LLL 878.5 824.7 817.8 793.1 -9.7 BBA ,....ounminnsennrrnss 1,661.0 1,638.7 1,604.3 1,607.0 -1.0 265... 4,136.4 4,183.8 4,182.5 4,327.5 4.6 Allages................. 1,192.3 1,165.4 1,158.2 1,163.0 -25 Data sources: 1.1973 and 1978 data, U.S. Department of Health and Human Services, National Center for Health Statistics, Health—United States, 1980, DHHS publication No. (PHS) 81-1232 (Hyattsville, Md.: DHHS, December 1980). 2.1975 and 1980 data, U.S. Department of Health and Human Services, National Center for Health Statistics, Health—United States, 1962, DHHS publication No. (PHS) 83-1232 (Hyattsville, Md.: DHHS, December 1982). 3.1979 and 1981 discharge data, U.S. Department of Health and Human Services, National Center for Health Statistics, Health—United States, 1983, DHHS publica- tion No. (PHS) 84-1232 (Hyattsville, Md.: DHHS, December 1983). 4.1979 days of care data, U.S. Department of Health and Human Services, National Center for Health Statistics, Health—United States, 1981, DHHS publication No. (PHS) 82-1232 (Hyattsville, Md.: DHHS, December 1981). 5. 1982 discharge data, U.S. Department of Health and Human Services, National Center for Health Statistics, “1982 Summary: National Hospital Discharge Survey,” No. 95, DHHS publication No. (PHS) 84-1250 (Hyattsville, Md.: DHHS, Dec. 27, 1983). SOURCE: Office of Technology Assessment. of discharges per 1,000 population increased dra- matically for the elderly population, but in the period from 1977 to 1982, the number increased more rapidly for the elderly than for other seg- ments of the population. Furthermore, once in the hospital, elderly people experience longer lengths of stay than do younger people.? It is interesting to note, however, that the covered days of care for the Medicare population increased only 4.6 percent between 1973 and 1980. “The longer lengths of stay experienced by the elderly are due in part to the use of the hospital as a substitute for nursing homes. “Backed-up” hospital patients awaiting nursing home placement have been estimated to account for 10 percent of the annual occupancy of some hospitals (371) and in 1980 cost Medicare $1.5 billion in hospital expenditures (371). In the hospital, elderly patients use many med- ical technologies more frequently than the rest of the population. Table 6 presents age-specific data on surgical operations in short-stay hospitals be- tween 1973 and 1980. In 1980, the rate of surgery among the elderly was 61 percent higher than the rate in the population as a whole. Furthermore, from 1973 to 1980, it increased by 37 percent, while the rate for the population as a whole in- creased only 22 percent (5). Intensive care units (ICUs) typically represent the confluence of medical technology and inten- sive nursing care in a complex system of care for the critically ill. In 1980, about 18 percent of Medi- care hospital stays involved intensive or coronary Table 6.—Operative Procedures in Short-Stay Hospitals by Patient Age, 1973-80 (rate per 1,000 population) Age 1973 1974 1975 1976 1977 1978 1979 1980 <15............ 41.9 42.2 40.5 41.0 41.0 37.6 37.2 36.0 1544 ........... 96.4 99.8 101.5 99.5 104.7 100.6 125.7 121.3 4584 ......0iu04 113.2 117.3 123.3 122.5 124.6 119.0 121.3 122.8 265 ............ 140.7 145.9 154.8 154.9 165.9 172.2 183.4 193.2 All ages? ...... 89.5 92.9 95.4 95.4 99.7 97.0 110.5 109.9 2Because of rounding, the sum of procedures for all ages may not total. SOURCE: I. N. Haug and R. Seeger (eds.), Socio-Economic Factbook for Surgery, 1982, Surgical Practice Department, American College of Surgery, 1982. Ch. 3—The Impact of Medical Technology on Medicare Costs © 45 care units (161). The available evidence seems to indicate that the representation of the elderly is the same or only slightly greater in ICUs than it is in the hospital as a whole (354). Thus, while the elderly are likely to require more intensive care than other segments of the general population, once in the hospital they appear to be placed in ICUs no more often than the nonelderly popula- tion (354). Once in an ICU, however, elderly pa- tients generally receive more interventions than other patients (57). According to Knaus, the key factor influencing the use of resources once a pa- tient is in an ICU is acute and chronic health status, not age in and of itself (190). Elderly pa- tients in ICUs are simply sicker than other ICU patients. The more frequent and intensive use of specific medical technologies by elderly patients translates into a greater representation of the elderly among high-cost patients within the hospital. Thus, for example, a 1976 study of almost 27,000 patients in three short-term hospitals found that 23.8 per- cent of the patients were over 65 years of age, but 41 percent of the high-cost patients® were over 65 (437). Furthermore, the National Medical Care Ex- penditures Survey conducted in 1977 found that the mean charge per hospital admission was $2,198 for patients 65 or older compared to $1,251 for the nonelderly population, a difference of $947 (395). The difference reflects not only the greater use of specific medical technologies by the elderly but also the longer inpatient stays generally ex- perienced by the elderly (10.3 days per admission compared to 7.1 days in the general population in 1977) (395). In 1977, the average daily rate for a semiprivate room was approximately $91 (180). *High-cost patients were defined in the study as patients whose total charges fell in the upper 20 percent of charges for patients in the hospital. Thus, of the $947 extra charge per stay, about $503 can be attributed to the extra use of ancil- lary technologies by the elderly and the rest to the longer length of stay. The general pattern of high use of medical tech- nology by the elderly extends beyond the hospi- tal to ambulatory care settings as well. As shown in table 7, the rate of ambulatory visits to physi- cians for diagnostic services is higher among elderly persons than among other segments of the population. Interestingly, however, the rate of X- ray testing in patients who do visit a physician for diagnostic reasons is not higher in the elderly. The elderly are also relatively high users of pre- scription drugs, despite the fact that Medicare does not pay for outpatient prescription drugs. In 1977, about 75 percent of the population 65 or older had at least one prescription compared to 58 percent of the general population (398). Fur- thermore, during the decade preceding that year, the intensity of use of prescriptions by Medicare Part B beneficiaries had increased (147). Finally, the use of medical equipment and supplies out- side of hospitals and nursing homes was more than twice as frequent in the elderly as in the gen- eral population (397). It is hardly startling that elderly people use more health care services and medical technologies in the aggregate and use them more intensively than the rest of the population. The importance of this fact lies in its implications for the Medi- care program. Changes in types of medical tech- nologies available or the patterns and conditions of the use of such technologies are likely in the aggregate to have strong effects on the costs of the Medicare program precisely because of the in- tensity with which Medicare enrollees use tech- nology. The next section attempts to explore the extent of that impact. MEDICAL TECHNOLOGY'S AGGREGATE IMPACT ON MEDICARE COSTS In order to investigate the aggregate contri- bution of changes in medical technology (i.e., changes in the kinds of technologies available and in the ways in which technologies are used in the practice of medicine) to Medicare costs, one must first examine the impact of technology on over- Table 7.—Use of Ambulatory Physician Visits With Specified Diagnostic Services by Age, 1977 Annual number of visits per 1,000 population Percent of persons with at least one visit Total population Visits with any Visits with Visits with Visits with any Visits with Visits with Age (in thousands) diagnostic services X-ray laboratory tests diagnostic service X-rays laboratory tests <6years.................... 18,216 746 119 528 39.1 7.9 30.9 cl 50,647 652 204 355 34.1 12.7 21.7 19to24vyears................ 22,299 1,211 231 654 46.9 13.4 26.5 25tob4years................ 78,472 1,327 262 565 51.5 16.0 24.4 55to64years................ 20,180 1,614 307 673 55.4 18.1 27.0 =>65years ................... 28,284 1,881 254 971 56.0 15.9 34.0 Allages ................... 212,098 1,189 236 574 46.6 14.4 25.8 SOURCE: L. F. Rossiter and C. M. Horgan, “Unequal Financial Incentives for Diagnostic and Preventive Health Care,” paper prepared for 109th Annual Meeting of the American Public Health Association, Medical Care Section, Los Angeles, Calif., November 1981. weiboid 81edIpe ay} Jo S}S0D pue ABOJOUYIS. [BIIPON o 9b Ch. 3—The Impact of Medical Technology on Medicare Costs ¢ 47 all health care expenditures. There are several methods for measuring technology's contribution to health care costs. The most common method is the “intensity of care” approach. The “Intensity of Care” Approach to Measuring Technology’s Contribution to Health Costs The intensity of care approach involves divid- ing a change in total expenditures for health care‘ into its constituent parts: * population or enrollment changes, e overall wage and price inflation, e wage and price inflation in medical care in excess of general inflation, and e changes in “service intensity.” Changes in technology use are included in the lat- ter two measures, although these measures also reflect other factors. Service intensity refers to the quantity of in- puts that go into producing a given unit of health care. Such inputs include labor, supplies, mate- rials, and equipment. Labor intensity refers to the quantity of personnel used to produce a unit of health care. Nonlabor intensity refers to the quan- tity of materials and supplies as well as the capi- tal plant and equipment used in producing the unit of health care. Although changes in service intensity have been labeled the “technology factor” (132), service in- tensity is not synonymous with medical technol- ogy use. To understand both the usefulness and limitations of estimates of changes in service in- tensity, it is helpful to consider how measures of intensity are related to the changes in medical technology whose effects are desirable to identify. One way to relate service intensity to the use of technology is to examine how hypothetical changes in medical technology would be likely to alter the operations, and thus costs, of health care institutions. The introduction of a new device in a hospital, for example, often involves both cap- ital (nonlabor) and some operating (both labor and nonlabor) costs for its application and main- This approach may also be used for one or more of the major components of health care costs. tenance. If the device is more sophisticated than the average technology in the hospital, it may re- quire more highly trained technicians, thus driv- ing up the average wages of hospital personnel. But the services provided by the device might sub- stitute for other services, thereby reducing labor and nonlabor costs in other areas. Or the new de- vice may have negligible effects on hospital oper- ations and simply be a product improvement, with a concomitant increase in product price.® Finally, a new device may draw into the hospital patients who would otherwise not be hospitalized, thereby increasing admissions and the routine (la- bor and nonlabor) costs associated with a hospi- tal stay as well as the costs of the service itself. Of course, these admissions might reduce the costs of other sectors of health care, such as ambulatory care or drugs. These observations suggest that the effects of changes in medical technology on health care costs must be traced through the changes’ specific ef- fects on hospital costs and other components of health care costs. Changes in hospital costs due to technological change are reflected in two measures: e service intensity, or the quantity of inputs per admission and the frequency of hospital ad- missions; and e the technical sophistication of inputs as re- flected in changes in the input prices (or wages) relative to general price level changes. Thus, changes in technology affect service inten- sity, but they also affect another component of hospital cost. Each of these components of hospital cost is also affected by forces unrelated to technology. For example, both the quantity of labor used in the hospital and the average wage paid to hospital personnel may be driven up because hospitals have inadequate incentives to be efficient or be- cause the hospital work force has been recently unionized (131). The price and quantity of medi- cal equipment, materials, and supplies might also sA new fully programmable cardiac pacemaker, for example, would be more expensive than more traditional pacemakers, but it would have little effect on the hospital's costs of pacemaker in- sertion; the full effect on hospital cost would be the increased price of the pacemaker reflecting its enhanced capabilities. 48 » Medical Technology and Costs of the Medicare Program increase relative to general inflation because of inadequate incentives for efficiency in the hospi- tal sector: Finally, hospital admissions are altered by changes in the incidence of illness and the gen- eral aging of the population, among other things. Thus, the components of hospital cost likely to be affected by changes in medical technology are also likely to be influenced by other factors. It appears, then, that the separation of health care cost increases into their components provides at best an oblique view of the contribution of changes in medical technology use to costs. The aforementioned caveats having been noted, the evidence on the components of hospital and health care cost (or expenditure) inflation is presented below. Several analysts have divided changes in hos- pital costs into their constituent parts, including service intensity (3,117,126,419,430). Waldman (419) estimated that increases in service intensity (i.e., labor, supplies, and equipment) accounted for about one-half of the annual change in the daily cost of hospital care between 1951 and 1970. Studies of increases in hospital costs per day through the mid-1970’s found similar results (3, 116). Feldstein and Taylor (117), for example, found that slightly less than one-half of the rise in average daily hospital costs between 1955 and 1975 was due to an increase in the intensity of services delivered per day. Altman and Wallack (3) found that roughly one-third to one-half of the annual increase in daily hospital costs between 1971 and 1976 was the combined result of an in- crease in the intensity of services and an increase in the price of hospital inputs relative to general wage and price inflation. Freeland and Schendler’s recent analysis of the 283-percent increase in national expenditures for hospital care over the period 1971 to 1981 found that 59 percent of the increase could be explained by overall inflation in the economy and growth in the U.S. population (126). The remaining 41 percent of the increase in national expenditures for hospital care was due to three technology- related factors: ® increased hospital admissions per capita (8.6 percent); ® increased intensity, or input use, per admis- sion (20.8 percent); and ® increased hospital input prices in excess of general inflation (11.7 percent). From 1971 to 1981, these three factors raised na- tional expenditures for hospital care about 157 percent. } Table 8 presents data on hospital cost increases for the period 1977 to 1982. OTA estimates that increases in service intensity (labor and nonlabor imputs) per capita accounted for 24 percent of the 93-percent increase in per capita hospital costs during the most recent 5-year period. A small part of this effect is due to the higher admission rate (a 5-year increase of 2.1 percent), but the over- whelming part of the intensity increase is due to higher intensity per hospital admission. The results of the aforementioned analyses are summarized and compared in table 9. The esti- mated growth in the intensity of hospital inputs clearly depends on the time period studied and the denominator unit. However, all five analy- ses support the conclusion that the intensity of hospitals’ services has contributed substantially to the growth in hospital costs over the past 20 years. When the components of growth of total per- sonal health care expenditures in the United States are considered, increasing intensity of care ap- pears to be a less important source of expendi- ture inflation than it is for the hospital sector alone. Table 10 shows estimated growth in real per capita personal health care expenditures be- tween 1977 and 1982 (when population growth and general price inflation, as measured by the Consumer Price Index, are taken into account). The combined effect of increasing intensity and increasing health care prices in excess of the Con- sumer Price Index is a relatively small proportion (about 16 percent) of the increase in per capita personal health care expenditures during the 5- year period. Nevertheless, these two technology- related components of cost together increased real per capita health care expenditures at an average annual rate of 2.8 percent during the period. It is possible to account for the components of Medicare cost increases over an appropriate time Ch. 3—The Impact of Medical Technology on Medicare Costs e 49 Table 8.—Decomposition of Hospital Costs, 1977-82 Percent Average Difference change annual 1977 1982 1982-1977 1977-1982 percent change 1. Total hospital costs (millions) ................ $51,647 $104,876 $53,229 103.1% 15.2% 2. Total U.S. population (thousands) ............. 219,760 231,534 11,774 5 11 3. Total adjusted hospital admissions? (thousands) 39,012 41,947 2,935 8 15 4. Total full time equivalent employees Qhousands) spasms mrevmrpnsswrsnisuesnes 2,573 3,306 733 29 5.1 5. Consumer Price Index (1977=100) ............ 100 159.3 NAP NA NA Service intensity per adjusted admission: 6. Hospital costs per adjusted admission ........ $1,324 $2,500 $1,176 89 13.6 7. Nonlabor costs per adjusted admission ....... $563 $941 $378 92 13.9 8. Nonlabor inputs per adjusted admission (7+ 5). $563 $591 $28 21 3.8 9. Labor costs per adjusted admission .......... $761 $1,421 $660 87 13.3 10. Index of hospital labor costs per full time equivalent employee (1977=100).............. 100 156 NA NA NA 11. Labor inputs per adjusted admission (9+10) ... $761 $910 $149 19 3.6 12. Change in labor and nonlabor inputs per adjusted admission 8+11)................... NA NA 177 NA NA Service intensity per capita: 13. Hospital costs per 1,000 population ........... $235 $453 $218 93 14.0 14. Nonlabor costs per 1,000 population .......... $100 $196 $96 95 14.3 15. Nonlabor inputs per 1,000 population (145) ... $100 $123 $23 23 4.2 16. Labor costs per 1,000 population ............. $135 $257 $122 92 14.0 17. Labor inputs per 1,000 population (16+10) ..... $135 $164 $29 21 4.1 18. Change in labor and nonlabor inputs per capita 31 Eb Inne NA NA 52 NA NA apdjusted admissions include outpatient visits, which are weighted in equivalent units. bNA—Not applicable. Data sources: 1. Most data derived from American Hospital Association, Hospital Statistics, table 6, 1978 and 1983 editions. 2. Population data from U.S. Department of Commerce, U.S. Statistical Abstract. ’ 3. Consumer Price Index found in U.S. Department of Health and Human Services, National Center for Health Statistics, Health—United States, 1982, DHHS publica- tion No. (PHS) 83-1232 (Hyattsville, Md.: DHHS, December 1982). SOURCE: Office of Technology Assessment. Table 9.—Summary of Studies of Hospital Cost Inflation Source and study period covered Freeland & Waldman Feldstein Altman & Wallack Schendler OTA 1951-70 1955-75 1971-76 1971-81 1977-82 Hospital cost per patient day: Annual average percent change. ..... 8.6% per year 12% per year 17.1% per year _— a Proportion due to service intensity ......... oi 50% 48% 30.6 to 50.5% — _— Hospital cost per admission: Annual average percent change ...... _— —- er 13% per year 13.2% per year Proportion due to service IRBNSHY ov vnrumesmervmevuivusams — — — 20.8% 15% Hospital cost per capita: Annual average percent change...... — — — _ 14.0% per year Proportion due to service INEENASIY os smsnvrsn ss mrs me sme swims — —_— es - 24% Data sources: 1.S. Waldman, “The Effect of Changing Technology on Hospital Costs,” Research and Statistics Note, Social Security Administration, Office of Research and Statistics, DHEW publication No. (SSA) 72-11701, Feb. 28, 1972. 2. M. S. Feldstein, Hospital Costs and Health Insurance (Cambridge, Mass.: Harvard University Press, 1981). 3.S. H. Altman and S. S. Wallack, “Technology on Trial—Is It the Culprit Behind Rising Health Costs? The Case For and Against,” in Medical Technology: The Culprit Behind Health Care Costs, proceedings of the 1977 Sun Valley Forum on National Health, DHEW publication No. (PHS) 79-3216, 1979. 4. M. S. Freeland and C. E. Schendler, “National Health Expenditure Growth in the 1980s: An Aging Population, New Technologies, and Increasing Competition,” Health Care Financing Review 4(3):1-58, March 1983. SOURCE: Office of Technology Assessment. Table 10.—Increase in Personal Health Care Expenditures,® 1977-82 Average annual Percent Percent Percent Percent Percent percent change 1977 1978 change 1979 change 1980 change 1981 change 1982 change 1977-82 1. Total personal health care expenditures (billions) ......... $148.7 $166.7 121% $188.9 13.3% $219.4 16.1% $255.0 16.2% $286.9 12.5% 14.0% 2. U.S. population (millions) ...... 220 222 0.9 225 1.4 227 0.9 229 0.9 22 1.3 1.1 3. Personal health care expenditures per capita........ $657.9 $750.9 11.1 $839.6 11.8 $996.5 15.1 $1,113.5 15.2 $1,236.6 11.0 12.8 4. Consumer Price Index ......... 100.0 107.6 7.6 119.9 11.4 136.1 13.5 150.0 10.2 159.3 6.2 9.8 5. Medical care index (1977=100) (12-month period ending September) .................. 100.0 108.5 8.5 118.3 9.1 131.0 10.7 146.1 11.5 162.2 11.0 10.2 6. Real personal health care expenditures per capita @B +4) $675.9 $697.9 3.3 $700.3 0.3 $710.1 1.4 $7423 45 $776.3 4.6 2.8 7. Real health inputs per capita (service intensity (3 + 5)....... $675.9 $692.1 2.4 $709.7 25 $737.8 3.9 $762.1 3.2 $762.4 0.1 2.4 8personal health care expenditures are national health care expenditures minus administrative costs. Data sources: : 1. Total health care expenditure data for 1977 to 1981, U.S. Department of Health and Human Services, National Center for Health Statistics, Health— United States, 1982, DHHS publication No. (PHS) 83-1232 (Hyattsville, Md.: DHHS, December 1982). 2.1982 health care expenditure data, U.S. Department of Health and Human Services, Health Care Financing Administration, Statistical Information Services, personal communication, Sept. 1, 1983. 3. U.S. population, U.S. Department of Commerce, U.S. Statistical Abstract. 4. Medical care price index, U.S. Department of Health and Human Services, Health Care Financing Administration, Health Care Financing Trends, DHHS publication No. 03091, winter 1981. 5.1981 and 1982 medical care index data, U.S. Department of Commerce, Bureau of Labor Statistics, personal communication, Sept. 1, 1983. SOURCE: Office of Technology Assessment. weibolg 81eoipe ey} JO S1S0D puke ABojouyoa] [BOIPON eo 0G Ch. 3—The Impact of Medical Technology on Medicare Costs ® 51 interval, but the interpretation of the estimates is more clouded than it is for general health care costs. Changes in program eligibility, such as the inclusion of disabled people in 1972, or in cov- ered benefits, such as the expansion of home health care benefits in 1980, can lead to dramatic changes in measured service intensity that have little to do with changes in medical technology but instead represent a shift in the burden of pay- ment for services already available and used. Changes in per capita service intensity do indicate how much more or less of health care services Medicare is paying for now than at some earlier date. Table 11 provides per capita estimates for 1977 and 1982. The data presented in table 11 indicate that most of the 107-percent increase in Medicare ex- penditures per enrollee between 1977 and 1982 is due to general price inflation. But 25 percent of the increase in Medicare expenditures per enrollee from 1977 to 1982 is due to Medicare's payment for more services per enrollee, and another 3 per- cent is due to the increased prices of medical serv- ices in excess of general price inflation.’ Thus, nearly 30 percent of the increase in Medicare costs per enrollee from 1977 to 1982 can be attributed to two technology-related components of costs. The percent due to medical price inflation may be overstated, and the service intensity percentage correspondingly understated, because the amount Medicare actually pays for services (i.e., the effective price) probably lies somewhat below stated prices. Table 11.— Real Medicare Expenditures per Enrollee, 1977 and 1982 1977 1982 1. Medicare expenditures per enrollee ............ cian. $927.54 $1,925.40 2. Consumer Price Index (1977=100) ........ccviuinrnnnn.. 100.00 159.30 3. Medical care price index N977= 100) 5» wuss mis gnsmmsnmss mes 100.00 162.40 4. Real Medicare expenditures per enrollee (1 + 2) ....ovvinienn. $927.54 $1,208.66 5. Real Medicare inputs per enrollee (1 += 3) .................. $927.54 $1,185.59 Data sources: U.S. Department of Health and Human Services Health Care Financing Ad- ministration, The Medicare and Medicaid Databook, 1981, HCFA publication No. 03156 (Baltimore, Md.: HCFA, April 1982). U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Statistical Information, personal communication, Sept. 1, 1983; U.S. Department of Com- merce, Bureau of Labor Statistics, personal communication, Sept. 1, 1983. SOURCE: Office of Technology Assessment. Other Estimates of Technology’s Contribution to Health Care Costs The service intensity approach has its limita- tions as a way of estimating technology's contri- bution to health care costs. A few analysts have used different approaches and data bases to look at the question. Redisch (267), for example, analyzed cost and operating data for a sample of about 1,500 hos- pitals and found that approximately 40 percent of the rise in operating costs per admission re- sulted from the increased use of eight types of ancillary services, all of which must be ordered by the physician. (The services were pathology, nuclear medicine, anesthesiology, pharmacy, lab- oratory, diagnostic X-ray, therapeutic X-ray, and blood bank.) Whether the increased use of ancil- lary technologies in the hospital has corresponded to reductions in the cost of other kinds of health care, however, is unknown. Several analysts have used a “residual ap- proach” to measure the impact of technological change on hospital or health care expenditures. In this approach, expenditures over time are re- gressed on a number of variables influencing sup- ply or demand for health care services.” The unex- plained residual of changes over time is then assumed to measure the effect of technological change. In a study of hospital costs from 1962 to 1968, Davis (82) found that 38 percent of the total an- nual increase in hospital cost per admission was unexplained by variables reflecting supply and de- mand conditions. This residual translates into a 2-percent annual increase in hospital expenses per admission attributed to technological change. Other analysts have used the residual approach to estimate the impact of technological change on total health care costs (130,231). In one study, which covered the period 1930 to 1975, Mushkin and colleagues (231) estimated that technological change reduced total health care expenditures at ’In the regression process, each variable receives a weight that represents the relative degree to which that variable explains or con- tributes to the change in expenditure. Some percent of the change cannot be explained by the variables. This percent is called the re- sidual. 52 e Medical Technology and Costs of the Medicare Program an annual rate of 0.5 percent. In a similar study, which covered the period 1947 to 1967, however, Fuchs (130) found that technological change raised expenditures at an annual rate of 0.6 percent. The difference in the findings of these two studies could, as Mushkin claimed, be due to dif- ferences in the periods studied or the variables chosen for study. Altman and Wallack (3) have pointed out significant limitations of the approach used in both studies. One limitation is the sensi- tivity of any residual estimate to the variables cho- sen for inclusion. In Altman and Wallack’s words, “even relatively small errors in specification [of the variables chosen] or in the statistics used to estimate the model can lead to the conclusion that technology has had a positive impact on rising health care costs when the true result is negative, or vice versa” (3). An even more important limita- tion of the approach is the narrow interpretation of technological change embodied in the residual. A major portion of the increased use of medical technology may well be attributed to demand- related factors such as the growth of third-party payment or personal income over the periods of study. Since these variables were included as variables in the regressions, the contribution to health care costs of changes in medical technol- ogy is underestimated. In short, the residual ap- proach gives too narrow a view of just how changes in the quantity, quality, kinds, and set- tings of use of medical technology have influenced health care costs. Another useful approach to looking at technol- ogy’s impact on health care costs is to focus on a specific illness and to document the array of medical practices and procedures used to treat the condition at two different times. The costs of treating the illness using the practices current in each time period can be estimated, and the dif- ference in these costs can be considered the effect of technological change on the cost of illness. However, it should be noted that this approach does not account for changes in the rates of use of treatments. Furthermore, only a few conditions can be studied because of the high cost of this kind of analysis. Trends detected in studies of a few illnesses certainly do not represent all illnesses and may not even represent the most important ones. Scitovsky and McCall (298) took this approach to explain the net increase from 1964 to 1971 in the average cost of treatment for eight conditions: otitis media, forearm fracture, appendicitis, ma- ternity care, breast cancer, pneumonia, duode- nal ulcer, and myocardial infarction. In almost every instance, there were both cost-raising and cost-saving changes in treatment. However, the authors noted that the costs of treatment of con- ditions requiring hospitalization rose at a con- siderably faster rate than those of conditions treated on an ambulatory basis. Among the fac- tors leading to higher costs were shifts to more expensive drugs, increases in the number of lab- oratory tests per case, and the use of more miscellaneous inpatient and outpatient services. The most dramatic cost increases occurred in the treatment of myocardial infarction, traceable prin- cipally to the increased use of ICUs during the time period.® The increase in the cost of treating this condition was greater than the decrease in the costs of five other illnesses combined. Of course, the net effect on health care costs would depend on the relative frequency of the various conditions in the population. Conclusions From the Aggregate Studies Although none of the approaches to measur- ing technology's aggregate contributions to health care cost is entirely satisfactory, taken as a whole, the available evidence leads to the conclusion that U.S. health care costs have increased in part be- cause more is being done for patients today than ever before. More and better trained personnel, more procedures, more medicines, and more and higher priced equipment, materials, and supplies are being used in the delivery of health care to Medicare patients and to the Nation as a whole. And, the trend toward “more” is not abating. The intensity of service use continues to increase. Despite the net increase in service intensity, the evidence also demonstrates the variation in tech- *For more information, see the forthcoming case study in OTA’s Health Technology Case Study Series entitled Intensive Care Units (ICUs): Costs, Outcomes, and Decisionmaking (354), prepared by Robert A. Berenson, M.D. Ch. 3—The Impact of Medical Technology on Medicare Costs ¢ 53 nology’s effects on costs. In the past 5 years, the hospital sector appears to have experienced rela- tively greater increases in intensity than has the health care sector as a whole. And, as Scitovsky and McCall's (298) research illustrates, cost-raising and cost-saving changes in technology are not scattered evenly across illnesses. The real cost of treating some illnesses has declined as a result of technological change, while that of others has in- creased dramatically. Thus, summary statements about technology's net influence on health care or Medicare costs mask the rich assortment of ways in which changes in medical technology shape the health care system and its costs. SELECTED MEDICAL TECHNOLOGIES AND MEDICARE COSTS To highlight the extent to which the costs of the Medicare program are altered by new technol- ogies, this section describes seven technologies first introduced in the 1960's or 1970's and exam- ines their actual and potential impact on Medi- care costs. The seven technologies are: coronary artery bypass graft surgery, the drug cimetidine, therapeutic apheresis, pneumococcal vaccine, intensive care units, parenteral nutrition therapy, and kidney dialysis. All seven of the technologies have clear patient benefits—in some cases, they are even life sav- ing—but for all of them, there are uncertainties about the most appropriate indications for use. Five of the technologies have raised or could raise Medicare's costs, in some cases significantly. Two have saved or could save Medicare costs. Above all else, these seven technologies illustrate how ex- posed the Medicare program is to changes in med- ical technology that are largely beyond its con- trol. In the face of new technologies that offer both patient benefits and higher costs, the challenge for Medicare may be how to encourage the use of those that are most cost effective. Coronary Artery Bypass Graft Surgery Coronary or arteriosclerotic heart disease, often caused by narrowing and blocking of the arteries that supply blood to the heart, is the number one cause of death in the United States. In 1982, heart disease was responsible for approximately 500,000 deaths (408). Furthermore, in 1968, this disease was the most frequent condition diagnosed for pa- 25-337 0 - 84 - 5 tients at the time of discharge from hospitals in this country (198). Coronary artery bypass graft surgery (CABG), a procedure in which a graft is used to bypass a constricted portion of the coronary artery and thus to improve oxygen supply to the heart mus- cle, has become the primary surgical approach to treatment of coronary artery disease (53). Since coming into practice in the early 1970's, the pro- cedure has diffused quite rapidly: approximately 25,000 operations were performed in the United States in 1973; at least 70,000 in 1977; 86,000 in 1979; 100,000 in 1980 (266); and 170,000 in 1982 (87,341). The rate of CABG in the United States has been estimated to be from 4 to 10 times as high as that of the United Kingdom, although the incidence of coronary artery disease is similar in the two countries (266,297). Data from a 15-institution registry of patients undergoing evaluation for suspected coronary ar- tery disease during the period from 1974 to 1979 reveal that 10 percent of such patients were 65 years of age or older (186). About 15.2 percent of the bypass procedures performed in Maryland in 1980 were on patients 65 or older (68). How- ever, 1982 data from the National Hospital Dis- charge Survey suggests that almost 30 percent of all such procedures were performed on those 65 years of age or older (87). Almost all evaluations of CABG have shown that the surgery is more effective than medical management in relieving angina pectoris (a con- dition characterized by severe chest pain). After surgery, angina is lessened in 80 to 90 percent and totally relieved in 60 to 70 percent of patients. The available data can be interpreted as suggesting that surgery is far more effective than medical manage- 54 e Medical Technology and Costs of the Medicare Program ment in improving that aspect of quality of life (266). Two clinical evaluations have demonstrated the life-extending properties of CABG in patients with coronary artery disease involving three ves- sels or the left main coronary artery, but the life- extending properties of the procedure are more uncertain when only one or two arteries are in- volved and when left ventricular function is se- verely restricted (422). Recently, the results of a clinical trial covering 15 medical centers revealed that CABG has not been shown to extend life in patients with mild or no chest pain and should probably be delayed until chest pain increases. The trial included patients under 65 years of age who did not have narrowed left main coronary arteries and who had mild or moderate chest pain, or those who had had at least one heart attack already but no chest pain. The investigators found no difference in mortality between medical and surgical management. Patients with surgery had greater relief from chest pain and better exercise tolerance, but the surgical group was hospitalized more often. Perhaps most telling, chest pain grad- ually worsened in both groups, and since a sec- ond operation is more hazardous than the first, the investigators concluded that “there is no pen- alty for waiting.” The investigators estimate that about 25,000 of the 170,000 CABG procedures performed in 1982 would be contraindicated by these findings (192). CABG itself is costly, estimated at approx- imately $15,000 to $20,000 in 1981, including hos- pital and surgical fees (422). But, the surgery also saves part of the costs of medical management of coronary artery disease and avoids the cost of treating heart attacks that are prevented by the surgery. When these savings in medical costs are taken into account, the net costs associated with CABG surgery range from $10,000 to $19,000, depending on the presenting condition of the pa- tient (422). If the age distribution of bypass surgery patients in the United States follows that reported by the National Hospital Discharge Survey, then approx- imately 50,000 procedures were performed on Medicare's aged population in 1982. This would imply that the procedure cost the Medicare pro- gram and its beneficiaries approximately $500 mil- lion to $950 million in that year.’ At this cost, Medicare buys for some elderly patients substan- tial benefits in the form of improved quality and extra years of life. For a substantial minority (esti- mated at 15 percent or 7,500 procedures), the pro- cedure may offer little in the way of improvement. Thus, an estimated $75 million to $142 million of the 1982 expenditures by or on behalf of Medi- care patients for CABG surgery may have been unnecessary. Even disregarding these potentially excess costs, CABG has had a substantial impact on annual Medicare costs. Cimetidine Peptic ulcer disease is a relatively common ill- ness with important ramifications for Medicare. In 1976, about 620,000 hospital discharges in the United States were for peptic ulcer, representing a rate of about 175 per 100,000 people (432). Fur- thermore, over 25 percent of the hospital stays for peptic ulcer involved surgery (433). The in- cidence of peptic ulcer disease increases with age (119). In 1978, fully 40 percent of hospital days of care for ulcer disease were for those 65 years or older (see table 12). In addition, the rate of ulcer-related surgery was twice as high for the elderly as for the general population (see table 13). In 1975, the total direct and indirect costs of ulcer disease in the United States were roughly esti- mated to be in the neighborhood of $2 billion (121). In August 1977, a new drug was approved for use in the United States for the short-term treat- ment of duodenal ulcers.'® This drug, known as cimetidine, acts by blocking stimulation of gastric acid secretion. Clinical evidence has demonstrated that cimetidine promotes healing of ulcers com- pared to placebo (121). Several analysts have investigated cimetidine’s impact on the use of health services. Studies in the United States and abroad have documented °Although not all of the costs of an individual surgery occur dur- ing the year in which the procedure is performed, the estimate is reasonably accurate for the Medicare population as a whole. 1°In 1976, duodenal ulcers accounted for approximately one-half of all hospitalized peptic ulcer cases in the United States. Gastric and unspecified ulcers accounted for the other half. Ch. 3—The Impact of Medical Technology on Medicare Costs © 55 Table 12.—Days of Care for Ulcer Disease for Patients Discharged From Short-Stay Hospitals, 1977-79 (per 100,000 population) 1977 1978 1979 All >65 All >65 All =>65 ICDA code ages years ages years ages years 531—Stomach ulcer........... 524.6 2,119.5 484.1 2,186.0 412.3 1,773.3 532—Duodenal ulcer .......... 751.3 2,554.1 636.9 2,220.2 559.7 1,955.5 533—Peptic ulcer ............. 294.4 940.4 263.9 871.4 248.4 708.5 534—Gastrojejunal ulcer. ...... 30.5 140.5 37.3 90.3 30.7 105.3 Data sources: U.S. Department of Health and Human Services, National Center for Health Statistics, “Detailed Diagnosis and Surgical Procedures for Patients Discharged From Short-Stay Hospitals: United States,” for years 1977, 1978, and 1979 (Hyattsville, Md.: DHHS). SOURCE: Office of Technology Assessment. Table 13.—Operations for Ulcer Disease for Patients Discharged From Short-Stay Hospitals, 1977-79 (per 100,000 population) 1977 1978 1979 All >65 All >65 All >65 ICDA procedure ages years ages years ages years 46.2—Partial gastrectomy .......... 24.1 77.7 18.5 37.7 4.0 8.1 46.8—Vagotomy................... 21.5 46.9 13.6 26.2 1.8 8.1 Data sources: U.S. Department of Health and Human Services, National Center for Health Statistics, “Detailed Diagnosis and Surgical Procedures for Patients Discharged From Short-Stay Hospitals: United States," for years 1977, 1978, and 1979 (Hyattsville, Md.: DHHS). SOURCE: Office of Technology Assessment. reductions in duodenal ulcer surgery rates imme- diately following the introduction of cimetidine in 1977. Fineberg and Pearlman (122) estimated that in 1978 the number of surgeries in the United States was 21,000 to 31,000 less than would have been predicted from the trend prior to 1977. In 1979 and 1980, the number of procedures was below the expected rate (but there was no sta- tistical significance) (122,312). Thus, cimetidine may delay surgery to a greater extent than it replaces it." The introduction of cimetidine coincided with a dramatic decrease in the rate of hospitalization for peptic ulcer disease in young adults (15 to 44 years old). There was only a modest decrease in the rate for all patients in the United States be- tween 1977 and 1978 (432). This fact suggests that the elderly population may not have experienced a substantial reduction in hospitalization as a re- sult of the drug's availability. 1t is important to note that the expected surgery rate in the absence of the cimetidine was calculated on the basis of a declining linear time trend. It is questionable whether such a trend would nor- mally continue as rates of surgery decline to low levels. One might expect surgery rates to level off at some point in time. A recent analysis reported on the impact of cimetidine on the costs of ulcer disease in Rhode Island (272). Although this study was limited by the available data, the researchers had access to hospital charges for patients undergoing ulcer sur- gery. Ulcer surgery rates declined in Rhode Island after the introduction of cimetidine, and a pro- portion of this decline was ascribed to cimetidine’s availability. The authors estimated that this re- duction in surgery meant statewide savings in medical care of between $185,000 and $450,000, depending on the extent to which it can be assumed that a reduction in surgery keeps ulcer patients out of the hospital. The evidence on the economic evaluation of cimetidine reviewed above highlights the impact that a single drug can have on the patterns of hos- pital and medical care. It also demonstrates the difficulty of determining whether these changes in patterns of use save health care or Medicare costs without the passage of enough time to mon- itor such changes. Today, physicians prescribe cimetidine for a variety of indications that are not among those approved by the Food and Drug Ad- ministration, including prevention of gastrointes- 56 e Medical Technology and Costs of the Medicare Program tinal bleeding in hospitalized patients (63,290). The economic impact of cimetidine in these areas has not been investigated, yet it could surpass the effects of cimetidine in treating ulcer disease. Therapeutic Apheresis'? Therapeutic apheresis is not a new procedure, but the extent of its use has grown rapidly dur- ing the past 5 years. It is a procedure in which a patient's plasma or blood cellular parts or both are separated and then removed from the blood and most often replaced by substitute plasma or a related physiological solution. It is believed that abnormal or harmful substances or cells are there- by removed, leading to a cure or arrest of disease. At present, apheresis is primarily accepted as an acute therapy in a small group of relatively ob- scure diseases, and the number of patients under- going treatment is approximately 20,000 (183). Results reported in the scientific literature have been dramatic, and apheresis is being used to treat an increasing number of medical conditions. Skep- ticism over the validity of such claims along with the high costs of apheresis, however, have touched off recent controversies over this proce- dure’s use. From 1977 through 1980, procedure volume in- creased more than 700 percent, from around 5,000 to over 40,000 procedures per year. In the late 1970's, the rate of growth far outpaced the esti- mates. For example, the now defunct National Center for Health Care Technology originally esti- mated its use in 1979 at “hundreds of procedures. It turned out to be around 16,000. A lot of peo- ple were doing it but not reporting it” (95). The costs of apheresis have become a particu- larly volatile issue. Each treatment costs between $400 and $1,200. Furthermore, each patient re- quires a number of treatments, usually varying between 5 and 15. (Sometimes as many as 30 treatments are needed initially, but the number tapers off with time.) Estimates of current national expenditures on apheresis therapy range from $3.2 million to $240 million. If apheresis therapy is ex- tended in the future to the wider array of diseases 12This section is based on a case study prepared for this project by OTA, entitled The Safety, Efficacy, and Cost Effectiveness of Therapeutic Apheresis (349). to which it has been only experimentally applied thus far, total national treatment costs could range from $650 million to over $7 billion per year (349). In 1981, the Health Care Financing Adminis- tration (HCFA) issued the first national instruc- tions on the coverage of apheresis under Medi- care. Only a small group of relatively rare diseases were listed as acceptable indications for the pro- cedure. These included: myasthenia gravis; leu- kemia; and macroglobulinemia and hyperglobu- linemias, including multiple myeloma (382). In 1983, a few additional uses were added to the list, including thrombotic thrombocytopenic purpura as a last resort treatment; life-threatening rheu- matoid vasculitis; life-threatening forms of Good- pasture’s syndrome, when the patient has not re- sponded to more conventional forms of therapy; and glomerulonephritis, when the patient has not responded to more conventional forms of ther- apy (74). Moreover, Medicare coverage of apher- esis has been limited only to procedures performed in the hospital inpatient or outpatient setting. The ultimate cost of therapeutic apheresis to the Medicare program will depend on whether cov- erage is extended to new indications and on the distribution of the affected diseases in the elderly population. By far the largest potential cost will arise if therapeutic apheresis is used for those who suffer from rheumatoid arthritis, which afflicts an estimated 5 million to 7 million people in the United States. Most observers believe that apher- esis would be used on patients who have failed to respond to traditional forms of therapy. At present, the consensus of professional opinion is that apheresis for treatment of rheumatoid ar- thritis is an experimental therapy (349). Clinical evaluation of the use of this procedure in rheuma- toid arthritis has been limited, but one controlled study found no statistically significant differences between the short-term response of patients re- ceiving apheresis together with drug therapy and the response of patients receiving drug therapy alone (287). Pneumococcal Vaccine An estimated 10 to 35 percent of all cases of pneumonia are bacterial infections caused by pneumococci (358). There are over 80 known Ch. 3—The Impact of Medical Technology on Medicare Costs ¢ 57 serotypes of pneumococcal bacteria (293), but a much smaller number is responsible for the ma- jority of pneumococcal pneumonias in the world. Vaccines for various combinations of pneumo- coccal serotypes have been produced at different points in time since the turn of the century, but in 1978, a vaccine offering protection against 14 serotypes of pneumococcal bacteria responsible for about 70 to 85 percent of pneumococcal infec- tions was introduced into commercial production (18). At the time of the vaccine’s introduction, im- munizations were specifically excluded as Medi- care benefits by Section 1862 of the Social Secu- rity Act. The ultimate potential for the pneumococcal vaccine is uncertain because of a lack of knowl- edge about the incidence of pneumococcal pneu- monia in various population groups, the dis- tribution of pneumococcal serotypes in these pneumonias, and the effectiveness of the vaccine in various patient groups, particularly high-risk groups (18,168). It is unknown, for example, whether a reduction in pneumococcal infections of the 14 types contained in the vaccine will be met with a concomitant increase in the incidence of other types of pneumococcal pneumonia, espe- cially in high-risk patients (31). Since estimates of economic costs must rely on estimates of these rates, they are themselves subject to a great deal of uncertainty. In 1979, noting these data and methodological problems, OTA performed a cost-effectiveness analysis of a pneumococcal vaccination program (358). OTA's analysis compared the net societal medical care costs and health effects (measured in terms of quality-adjusted life-years) that would result from vaccination. Under the base case set of assumptions, vaccination would increase net medical care costs'? for vaccinees in all age groups, but would also yield health benefits that could not be obtained through treatment. Furthermore, vac- cination of the elderly (those 65 years of age and older) was relatively cost effective in comparison to many existing health programs. BOTA's analysis included the discounted value of future medi- cal care costs arising from increased life expectancy among vaccinees. Working largely with data provided by OTA, the Congressional Budget Office analyzed the likely impact of covering the pneumococcal vac- cine as a Medicare benefit on Medicare expendi- tures (331). That study found that a vaccine ben- efit would be cost saving to Medicare after 3 to 5 years, depending on the assumptions made about vaccination rates, levels of reimbursement for vaccination, and the inclusion of medical costs arising from increased life spans. Partly as a result of these analyses, Congress amended the Social Security Act to allow Medi- care coverage of pneumococcal vaccination. At present, it appears that vaccination rates in the Medicare population are low, and estimates of Medicare cost impacts are not available (305). Intensive Care Units The ICU is an example of a technology which has proliferated widely despite the absence of studies of efficacy or cost effectiveness. Because of the difficulty of separating the intensity of care from the setting in which it is provided, it is dif- ficult to know whether intensive care would be as effective if provided on the general hospital floor as in the physically and administratively sep- arate ICU. For many medical problems, however, treatment in an ICU has become the standard method of treatment. A recent National Institutes of Health spon- sored consensus panel concluded that it is im- possible to generalize about whether ICU care improves outcome for the varied ICU patient pop- ulation. The panel agreed that ICU intervention is unequivocably lifesaving for some conditions, particularly where there is an acute, reversible problem, such as drug overdose or major trauma. It was less certain about the effectiveness of ICU care in other conditions, particularly in the pres- ence of a severe, debilitating chronic illness, such as cancer or cirrhosis of the liver (409). Despite the uncertainty about the indications for ICU care, almost 80 percent of short-term gen- 14For more information, see OT A's forthcoming technical memo- randum Update on Federal Activities Regarding the Use of Pneu- mococcal Vaccine, OTA-TM-H-23 (Washington, D.C.: U.S. Gov- ernment Printing Office, May 1984). 58 e Medical Technology and Costs of the Medicare Program eral hospitals have at least one ICU (354). Large hospitals are likely to have two or more ICUs, organized along specialty lines. Overall, 5.9 per- cent of total hospital beds in non-Federal, short- term community hospitals in 1982 were beds in adult intensive and coronary care units (9). In 1980, 7 percent of hospital Medicare charges were for intensive and coronary care units (161). This figure understates the full costs of ICU care be- cause it does not include the ancillary charges for patients. In any case, it is a representation of charges instead of costs, which may be higher. It is estimated that the costs of adult intensive and coronary care unit care represent over 15 percent of total hospital inpatient costs, or $4,742.5 mil- lion in 1982 (354). Inclusion of the other types of specialized ICUs, such as neonatal and burn care units, would bring the percentage up to about 20 percent of total hospital costs, or almost 1 per- cent of the Nation's gross national product. According to 1979 Medicare data, 18 percent of Medicare discharges included a stay in inten- sive or coronary care units (160). From reports from individual hospitals, it appears that the representation of the elderly Medicare population in ICUs is about the same as in the hospital as a whole (354). Age alone does not appear to be a significant factor limiting use of ICUs in the United States. It is noteworthy that in other coun- tries, ICU patients have a significantly lower mean age (354). The literature on the outcomes of ICU care has demonstrated consistently the inverse relationship between the cost of ICU care and the likelihood of survival. The sickest ICU patients, many of whom do not survive their hospital stay, consume a disproportionately high share of ICU costs. Under Medicare's cost-based hospital reimburse- ment system, the high-cost patients were not par- ticularly burdensome financially to the hospital. However, under Medicare's recently initiated pro- spective payment system for inpatient hospital care, many of these long-stay, high-cost ICU patients will become financial losers to the hos- pital (354). Total Parenteral Nutrition Total parenteral nutrition (TPN) refers to the intake of nutrients directly into the bloodstream, circumventing the digestive tract (14). Its primary use is in eliminating malnutrition in patients who cannot adequately digest food or whose nutrition- al needs are elevated because of disease or injury. To receive TPN, a patient must have his or her nutritional needs assessed by a doctor or dietician and must have a catheter implanted in a large vein. Clinically, a patient must be on TPN for a va- riety of reasons, most commonly inflammatory bowel disease (e.g., Crohn's disease), ischemic bowel infarction, and cancer-related problems, in- cluding damage due to radiation therapy (158). Indications for the use of TPN have been the sub- ject of considerable discussion in the medical pro- fession in the last 5 years. Some physicians ad- vocate the use of TPN to bolster patients before surgery and to improve cancer patients’ tolerance to therapy. Others suggest that TPN has little in- fluence on the outcome in these cases and may actually promote tumor growth (139,181). Before the late 1960's, prolonged maintenance of patients with digestive dysfunction was not possible. The development of TPN came about through advances made in four areas: improved knowledge of human nutritional needs, improved surgical procedures, improved catheter composi- tion and design, and improved infusion control devices. The development of volumetric infusion pumps, especially the cassette-type electronic pump introduced in 1974, was the watershed for safe and reliable infusion that made overnight parenteral feeding practical. TPN can be delivered either in the hospital or in the home. Before 1979, all home TPN patients were treated as hospital outpatients. In that year, a private firm, Home Health Care of America, entered the market, offering a package of supplies and services (189). Today there are some 30 to 40 commercial home TPN providers, several of Ch. 3—The Impact of Medical Technology on Medicare Costs ® 59 which are owned by firms that manufacture solu- tions and supplies (37). In either setting, hospital or home, TPN is an expensive long-term therapy. A study at the Cleveland Clinic found that costs to the hospital for home TPN were about one-fourth of TPN costs in the hospital, but even in the home, the average per-patient first-year costs were estimated at $21,465 (in 1978 dollars) (421). The most im- portant factor in cost was the quantity of dis- posable supplies, including nutritional solutions, which accounted for almost 90 percent of the total cost. Other studies have estimated costs of a typi- cal home TPN patient to be about $40,000 to $45,000 per year (158,189,301). Medicare coverage of TPN delivered to hospi- tal inpatients has never been at issue. TPN pro- vided in a hospital setting has been covered as a Part A hospital benefit since the technology was developed. In 1977, HCFA began to cover home TPN on the advice of the Public Health Service. At that time, HCFA did not anticipate home TPN as a major expense; it was expected that only about 10 patients per year would need home cov- erage and that most of these patients would not live long (56). Because intravenous nutrients are classified as drugs and are therefore not individ- ually reimbursable under Medicare's Part B, HCFA declared the whole home TPN system a prosthetic device, and therefore subject to Part B coverage. In 1981, HCFA tightened the require- ments for home TPN, listing seven diagnoses for which it was appropriate. Other indications can be approved on a case-by-case basis (62,184). All persons eligible for Medicare and partici- pating in the Part B program are covered for home TPN supplies. Persons younger than 65 and not otherwise eligible can receive Medicare coverage if their need for home TPN renders them unable to work, but over half of home patients consider themselves fully functional once they receive the needed nutrition (241). TPN covered under Part B is reimbursed on the basis of reasonable charges, but because there have been relatively few home TPN patients, it has been difficult to establish charge screens. As with other Part B services, TPN at home is subject to the deductible and coinsur- ance provisions of Medicare. The use of home TPN has undergone tremen- dous growth in recent years, much of which was stimulated by the increase in coverage by Medi- care and other insurers. For example, a registry maintained by the New York Academy of Medi- cine reported a 103-percent increase in the num- ber of patients on home TPN between 1979 and 1981 (241). It has been estimated that about 200 TPN patients were discharged to the home in 1978 (301), while estimates for 1983 are around 4,000 home patients (37). It is unknown to what extent this increase represents a substitution of home TPN for inpatient nutrition services and to what extent it represents a net increase in the number of patients receiving TPN. The home parenteral nutrition registry esti- mates that about 22 percent of patients receiving home TPN in 1979 were Medicare enrollees (241). If the cost estimate of $40,000 per year for TPN in the home is accepted, and if it is assumed that 22 percent of the 4,000 patients on home TPN are Medicare beneficiaries, then Medicare currently pays in the neighborhood of $28 million per year for home TPN.?® A lack of data precludes esti- mation of the total cost to Medicare of providing TPN in the hospital, but it is likely to be greater than that for home TPN. Although over $28 million in annual Medicare expenditures for a technology that extends life is small in relation to total Medicare expenditures ($52.2 billion in 1982), this case illustrates the ex- tent to which the impact on Medicare cost of a new technology can be grossly underestimated at the time coverage is introduced. Hemodialysis for Chronic Renal Failure Hemodialysis represents the classic case of a life- saving technology whose development dramat- ically affected the costs of Medicare. Although hemodialysis has been available since 1945 for temporary treatment of acute and reversible renal failure, its application to patients with end-stage renal disease (ESRD) was first made possible in 1960, when Quinton and Scribner developed a subcutaneous arteriovenous shunt (a plastic tube connected to an artery and a vein in the arm or 15(0.22 X 4,000 X $40,000 X 0.80). 60 ¢ Medical Technology and Costs of the Medicare Program leg) (271). Without the shunt, filtering the blood as often as necessary was not possible on a long- term basis because the blood vessels would col- lapse. In the early 1960's, hemodialysis became accepted as a life-extending therapy for victims of chronic kidney failure. The cost of hemodialysis for ESRD varies with the setting in which treatment is provided. One study estimated that the costs of dialysis in 1980 were approximately $25,000 per patient year for in-center treatment and $13,000 per patient year for home treatment after the first year (279). Because of the high costs and the obvious life- extending properties of both hemodialysis and its Photo credit: National Kidney Foundation, Washington, D.C. Hemodialysis is a lifesaving medical technology that has affected the costs of the Medicare program CONCLUSIONS The evidence presented in this chapter was in- tended to illuminate how changes in medical tech- nology alter the cost of the Medicare program. Although its aggregate impact on Medicare costs cannot be estimated precisely, medical technol- ogy has clearly added to the costs of the Medi- care program and to health care costs as a whole. Today, Medicare is buying more services for its beneficiaries than ever before, and the pressure to adopt new beneficial but cost-raising technol- ogies continues. Despite this conclusion, the in- crease in the provision of services to Medicare competitor, kidney transplantation, a debate began in the mid-1960’s over who should be re- sponsible for paying for treatments of patients with ESRD. The debate culminated in 1972 with the passage of the Social Security Amendments (Public Law 92-603), which extended Medicare coverage of treatment for ESRD to over 90 per- cent of the ESRD population. Factors that led to the congressional decision to pay for ESRD treat- ment included a recognition that the alternative to life sustainment by dialysis or kidney transplan- tation was death, that ESRD treatment was very expensive, and that there occurred 7,000 to 10,000 uremic deaths a year because of the limited avail- ability of dialysis facilities. In 1972, 40 patients per million population were receiving long-term hemodialysis treatment in the United States, almost entirely under the auspices of nonprofit organizations. The number now ex- ceeds 200 per million population (a fivefold in- crease) and is one of the highest in the world (96,269). The cost of Medicare's ESRD program grew from $250 million in 1974 to an estimated $1.8 billion in 1982 (378), greatly exceeding original congressional estimates of the potential costs (279). In 1979, benefit payments for ESRD ex- ceeded 5 percent of total Medicare expenditures and were fully 10 percent of expenditures from the Supplemental Medical Insurance fund of Med- icare, although renal patients constitute only 0.2 percent of the Medicare population (38,279). beneficiaries and inflation in the price of medical care represent less than a third of the 107-percent increase in Medicare expenditures per capita from 1977 to 1982. Increases in enrollment and general price inflation account for the bulk of Medicare expenditure inflation. The descriptions of the seven new medical tech- nologies provided in this chapter highlight the dif- ficulty of predicting at the outset how technologi- cal change in medicine is likely to affect Medicare costs in the future. New technologies are and will Ch. 3—The Impact of Medical Technology on Medicare Costs * 61 continue to be developed regardless of Medicare's policies. Some of them substantially prolong life or improve its quality for Medicare beneficiaries. Application of new technologies to the Medicare population can have large and unanticipated im- pacts on Medicare expenditures. But the cases demonstrate quite clearly that the extent of im- pact on Medicare program expenditures depends on whether Medicare chooses to cover a technol- ogy, and if it chooses to, to influence the condi- tions under which it is used. It is important to note that changes in service intensity are measures of the incremental effects, not the cumulative effects, of technology on total Medicare costs. It is not just at the margin that there is an opportunity to reduce Medicare costs by altering the patterns of technology adoption and use; there are many opportunities to save costs by altering longstanding patterns of use of medical technology. It might be desirable to have the new cost-raising but life-extending technol- ogies widely adopted and used and the use of many existing ineffective technologies substan- tially reduced. The issue for the remainder of this report is how Medicare policy can be structured to bring about the most cost-effective use of both new and existing medical technologies. Part Two 4. A Framework for Change No great improvements in the lot of mankind are possible, until a great change takes place in the fundamental constitution of their modes of thought. John Stuart Mill 4. A Framework for Change INTRODUCTION The following chapters examine policies that have been or could be used to restrain the costs of medical technologies in the Medicare program. Several points underlie the analyses in these chapters. First, the impact of medical technologies on the costs of medical care should not be assessed in isolation from the effects that such medical tech- nologies have on patient care. The impact of cost- containing measures on quality and access is one of the more difficult policy issues to be faced, be- cause the Medicare program was instituted on a payment basis that had few controls on costs. Now that costs are a primary concern of Federal policymakers, some restrictions on quality and access are likely to occur. Nevertheless, there is substantial evidence to suggest that inappropriate use of medical technology is common and raises costs without improving quality of care. Methods of controlling the costs of medical technologies can vary widely and have varying impacts. Direct methods, for example, are meth- ods that are intended to control the use of spe- cific medical technologies on a technology-by- technology basis. Such methods could be used: 1) to control the actual adoption or use of par- ticular technologies, as in the coverage process that assesses specific technologies before they are approved for payment; or 2) to provide informa- tion on the costs of technologies so that payment for their use could be more reasonably related to their costs. Indirect methods include methods that use the payment mechanism to provide broad in- centives to medical care providers not to over- utilize medical technologies and to make patients more cost conscious in their use of medical serv- ices. Indirect methods are now considered the ma- jor means through which long-term cost-contain- ment objectives might be achieved. Particularly when used in conjunction with indirect methods, however, some direct methods, such as review of capital spending, utilization review, and some other types of technology assessment activities, may also be valuable. Second, there are interactions between Medi- care and the rest of the U.S. health care system. Because of its size and scope, the Medicare pro- gram’s policies and procedures affect all aspects of health care delivery, including financing, ad- ministration, organization, and personnel. Fur- thermore, the program affects the content and costs of health care by its influence on the devel- opment, adoption, and use of medical technol- ogy. Medicare's leverage in the health care sys- tem is partly due to the fact that Medicare alone finances over one-third of the country’s hospital care, the setting where technology use is concen- trated (15). It is also partly due to the fact that other third-party payers often follow Medicare's example. Nevertheless, it is important to keep in mind that the Medicare program is only one of many public and private institutions that have an in- fluence on the development and diffusion of med- ical technology. Other important influences are the Food and Drug Administration, National In- stitutes of Health, manufacturers of drugs and medical devices, hospitals, private health insurers, and professional medical societies. Thus, for ex- ample, the leverage of using Medicare-specific payment policies to influence the development and diffusion of medical technology may be limited. Third, because of spillover effects from one part of Medicare to another, policy mechanisms in- volving only one part of the Medicare program may have serious limitations in terms of contain- ing costs or affecting the adoption and use of tech- nology. Medicare's hospital payment system based on Diagnosis Related Groups (DRGs), for example, excludes physician services and outpa- tient care. These exclusions provide financial in- centives for the shifting of technologies and costs 67 68 ¢ Medical Technology and Costs of the Medicare Program out of inpatient hospital settings, while leaving physicians’ incentives to use medical technology unaffected. Any cost-containment effort must take into account the fact that physicians play a cen- tral role in determining what services are provided to patients, both in hospital settings and in non- hospital settings. Most—if not all—cost-contain- ment strategies depend on the ultimate influence of physicians for their success. Cost-contain- ment strategies can be targeted directly at physi- cians—although such strategies are regarded by most observers as short-term and inadequate approaches—or they can be directed at the prac- tice of medicine through changes in the organiza- tional and financial arrangements under which physicians provide care. For the success of strat- egies that depend on incentives rather than on di- rect regulation, it is essential that physicians believe that the incentives are advantageous. Fourth, the social and political climate today is quite different from that in 1965, and now that Medicare's goal of improving access to health care for the Nation's elderly has been largely achieved, the primary focus of policymakers is on contain- ing Medicare costs. The intent of the original Medicare law (Public Law 89-97) was to increase elderly persons’ access, by removing financial bar- riers, to mainstream medical services, particularly to needed hospitalization (318). The concern about access to medical services was also promi- nent when disabled persons and those with end- stage renal disease were added to the list of eligi- ble beneficiaries. There was far less concern about the cost of the services than there was about the problems of access, primarily because there was little reason to be concerned. Early principles in addition to improving access included assuring beneficiaries freedom of choice of providers and not interfering in the practice of medicine. Un- fortunately, Medicare's adherence to these original principles has contributed to the current cost crisis. Today, in part because the original goals have been largely—though certainly not entire- ly—attained, the overriding goal for policymakers is to solve the problem of controlling Medicare costs. The challenge is to achieve that goal with- out diminishing past success. The aforementioned points are closely inter- twined. The relationship between cost contain- ment and its effects on quality and access to med- ical care is but one example. Equally problematic is the widely held belief that specific policies that could be implemented in the short-term and di- rected at specific segments of the health care sys- tem will provide only temporary relief in medi- cal care cost inflation. On the other hand, long- term success is increasingly dependent on broad but still untested ideas of the kinds of strategies (e.g., “competitive” systems, alternative delivery sites and organizations) that could lead to ade- quate cost containment. One fundamental dilem- ma, therefore, is whether policymakers can be precise about cost-containment processes for which the desired outcomes are quite limited, while still exploring the kinds of processes that would lead to the desired long-term or broader cost-containment outcomes. ORGANIZATION OF THE FOLLOWING CHAPTERS Chapter 5 examines the potential of linking Medicare's technology-specific coverage policy with technology assessment activities as a means of influencing the adoption and use of specific medical technologies for the ultimate purpose of containing Medicare program costs. In the past few years, assessment of the health effects—i.e., safety and efficacy—of some technologies has become part of the process of arriving at cover- age decisions. Two current issues are whether costs should be considered in Medicare coverage decisions and whether coverage of new technol- ogies should be limited to specific sites and pro- viders. This chapter provides information on the Medicare coverage process and technology assess- ment as practiced in the public and private sec- tors, and analyzes the strengths and limitations of the current coverage process, technology assess- ments, and possible linkages between the two processes. It also discusses the role of coverage Ch. 4—A Framework for Change ¢ 69 policy and technology assessment under Medi- care’s DRG hospital payment system. Chapter 6 discusses the implications of the DRG hospital payment system for the adoption and use of medical technologies. These implications are varied and to some extent unknown. Much will depend on the way in which the system is im- plemented and the refinements that will follow. DRG payment levels, especially relative to the speed with which hospitals can reduce costs, will have a major effect on the ability of hospitals to adopt new medical technologies. The way in which capital is paid for will also be an impor- tant influence in determining how much and what kinds of new technologies are adopted. Chapter 6 also includes a discussion of alternative ap- proaches to hospital payment and of the impli- cations of these approaches for medical tech- nology. Chapter 7 describes the Medicare physician payment system and analyzes the impact of pro- posals to limit physician payment or increasing beneficiary cost-sharing under Part B on medical technology adoption and use. Physicians deter- mine the amount of medical services provided and decide when patients need to be hospitalized, dis- charged, or provided other types of institutional and noninstitutional care. This chapter examines evidence of excessive use of technologies and methods of enhancing cost consciousness among physicians. Such methods include programs to help ensure appropriate technology use by phy- sicians that might be incorporated in the Medi- care program. For example, the law authorizing DRGs also puts into place a mechanism for quality assurance and utilization review by requiring hos- pitals to contract with regional peer review or- ganizations. 1The recent changes in Medicare's hospital payment system estab- lished by the Social Security Amendments of 1983 (Public Law 98-21) are more extensively discussed in OTA'’s July 1983 technical memo- randum Diagnosis Related Groups (DRGs) and the Medicare Pro- gram: Implications for Medical Technology (343). 25-337 0 - 84 - 6 Except for the imposition of minor restraints on the rate of increase of payment levels, Medi- care’s charge-based method of payment for phy- sicians’ services under Part B has been little changed since 1966. This method provides finan- cial incentives to physicians for increased tech- nology use by the way that fees are set and by the coding system used. The achievement of the cost-containment objectives of the DRG-based payment system for hospital services could be par- tially impeded through movement of some tech- nologies and services out of the hospital setting. For that reason, the law establishing the DRG sys- tem requires that data necessary to compute the amount of inpatient physician charges based on DRGs be collected and that the Department of Health and Human Services report to Congress on the prospects of including physicians in DRG payment. The possibility that changes in Medi- care payment for physicians’ services may lower beneficiaries’ access to medical care is also exam- ined, principally through an analysis of the im- plications of changing physician assignment pol- icy in the Medicare program. Chapter 8 explores mechanisms other than hos- pital or physician payment that Medicare could use to foster the appropriate adoption and use of medical technology. These include stimulating competition among providers of health care by encouraging the development of alternative sites and organizations of care such as health mainte- nance organizations, home health care, and am- bulatory surgical centers, and the use of vouchers and other methods. Examination of these four areas—current Medi- care coverage policy and related technology as- sessment activities, changes in hospital payment, changes in physician payment, and other meth- ods to encourage the appropriate adoption and use of medical technology —lead to the final chap- ter of this report. Thus, chapter 9 presents OTA’s conclusions and policy options. b. Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs Indeed, what is there that does not appear marvelous when it comes to our knowledge for the first time? Pliny the Elder Contents Page Introduction’... 0 cut. ER a 73 Definitions Ji... oii ip Br Lh Rene 73 Medicate Coverage... ....i i J... pao 74 TheCoverageProcess .........0... i iio oo ed 76 Coverage Policy Under Medicare's DRG Hospital Payment System. ......... 82 Evaluating Technologies for Coverage Decisions ..................... .. 84 Current Activities... ... o.oo... a a oa Sl 84 Analytic Methods of Comparing Costs and Benefits ....................... 85 Cost-Savine Technology... 0 a... 0 ade 86 Discussion LL... se aR es 88 LIST OF TABLES Table No. Page 14. Sources of Information Used by Medical Consultants in Making Coverage Decisions... cue... nivaescisnnen. Sia 0S 78 15. Sources of Coverage Questions Submitted to HCFA’s Central Office . .. . . 79 16. Reported Coverage Decisions by Medicare Contractors . ............... . 81 FIGURE Figure No. Page 2. Model of Medicare's Coverage Process for Individual Medical Technologies... 76 5. Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs INTRODUCTION The introduction of a new technology some- times has large and unanticipated impacts on Medicare expenditures (see ch. 3). The extent of impact depends on whether Medicare chooses to cover (i.e., pay for) the technology, and, if it chooses to cover it, to specify the conditions of its use. Coverage policy, i.e., policy that governs the eligibility of services (technologies) for pay- ment, has been a significant factor in hospitals’ decisions regarding the purchase of expensive, vis- ible medical technology (24,289). The relationship between coverage policy and adoption of other kinds of medical technology or the use of any medical technology remains speculative. Title XVIII of the Social Security Act specifies broad, general categories of medical and health services (e.g., hospital services and physicians’ services) and some specific items (e.g., home di- alysis supplies, pneumococcal vaccine) that the Medicare program will cover (see ch. 2). It also lists a number of specific services and items that the program will not cover. For the most part, however, decisions about which technologies Medicare will pay for are made at the national level by the Health Care Financing Administra- DEFINITIONS Coverage is generally defined as “the guaran- tee against specific losses provided under the terms of an insurance policy.” The term is frequently used interchangeably with benefits or protection. Coverage also means “the extent of insurance of- fered by a policy” (335). Insurance plans (includ- ing Medicare) specify, to varying degrees of preci- sion, the benefits they will pay for. Thus, coverage tion (HCFA) or at the local level by Medicare con- tractors. In the past few years, rapid technological devel- opment has led to an increasing need for decisions by Medicare and other third-party payers about the coverage of specific technologies. Many cov- erage decisions are based on an assessment of the health effects of the particular technology. For the most part, these assessments are not rigorous. In- deed, it is estimated that only 10 to 20 percent of technologies used in medical practice have been shown to be efficacious by controlled trials (341). Evaluation of the nonmedical effects, for exam- ple, economic and social effects, of specific tech- nologies is usually not part of an assessment for coverage purposes. This chapter discusses the possibility of refin- ing Medicare's coverage policy, for example, by using appropriate technology assessments, as a means of influencing the diffusion of medical tech- nology. Changes in Medicare's coverage policy for specific technologies may provide an incre-. mental approach to controlling Medicare costs. refers both to the broad categories of benefits spe- cified in the law or in a plan as well as to the spe- cific services actually provided and paid for. In the Medicare program, coverage is distinguished from payment or reimbursement: coverage refers to benefits available to eligible beneficiaries, and payment refers to the amount and methods of payment for covered services (434). 73 74 ¢ Medical Technology and Costs of the Medicare Program “... [Technology assessment is simply a broader form of policy research than is commonly conducted. The goal of technology assessment, as of all policy research, is to provide decision- makers with information on policy alternatives, such as allocation of research and development funds, formulation of regulations, or development of legislation” (23). A comprehensive assessment MEDICARE COVERAGE The basis for decisions by HCFA or Medicare contractors regarding the coverage status of med- ical technologies not otherwise specifically men- tioned in Title XVIII of the Social Security Act is Section 1862. Section 1862, among other things, prohibits payment by Medicare for any expenses incurred for items and services which are not “rea- sonable and necessary for the diagnosis or treat- ment of illness or injury or to improve the func- tioning of a malformed body member. . . .” The provision applies “notwithstanding any other pro- visions” of the title. Coverage policy for specific technologies is ex- pressed in the development, issuance, and imple- mentation of coverage decisions. Such policy is made in light of Medicare's twin principles of not interfering with the practice of medicine and of assuring beneficiaries a free choice of providers. For the most part, questions regarding Medicare coverage status arise with respect to new technol- ogies or new applications of covered technologies, although occasionally, the coverage status of cov- ered, established technologies is reexamined. The focus of this chapter is on the Medicare coverage process for new technologies and new uses of cov- ered technologies.! There is a basic contradiction between Medi- care’s stated intention of not interfering with the practice of medicine and the delivery of health care and its coverage policy that judges technol- ogies to be used in medical practice. A decentral- ized approach to the coverage process attempts to minimize the contradiction by accepting the "The term “new technologies” henceforth refers to both new tech- nologies and new uses of established technologies. examines the technical, economic, social, and legal consequences of technological applications. A less comprehensive assessment of a medical technol- ogy may focus only on the health effects of the technology. The typical meaning of the term “technology assessment” in health policy today is an evaluation of a technology's efficacy and safety and sometimes costs. premise that medical practice varies from one geo- graphic area to another. Some Medicare coverage decisions for specific technologies are made at the national level by HCFA's central office. Most of the decisions, how- ever, are made by Medicare contractors who per- form the Medicare program's claims processing and payment function under the policy and oper- ational guidance of HCFA. Medicare contractors, called intermediaries and carriers, are either Blue Cross/Blue Shield plans or commercial insurers. On the U.S. mainland, 84 intermediaries admin- ister Part A (institutional services) of the Medi- care program, and 61 carriers administer Part B (physicians services) (see app. E).2 HCFA's 10 re- gional offices assist contractors with coverage decisions and transmit information between HCFA's central office and Medicare contractors. Because of the general language of Section 1862 and the absence of regulations that implement the section, HCFA officials and Medicare contractors alike have had considerable latitude in determin- ing which technologies are to be covered for reim- bursement. Medicare coverage policy is continuously evolving and is developed and implemented in a decentralized manner. National policy has devel- oped largely as a result of questions from in- dividual contractors about whether they should pay for specific technologies (366,435). HCFA in- forms contractors about the coverage status of “There are two governmental bodies that perform similar func- tions: the Office of Direct Reimbursement and the Group Health Plan Operations Staff in the Bureau of Program Operations. HCFA is moving to contract out the intermediary functions of the Office of Direct Reimbursement to regional intermediaries. Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs © 75 some specific technologies through transmittal let- ters and a Medicare manual. However, HCFA's coverage instructions have no standing in law or regulation, so the contractors’ compliance is essen- tially voluntary (366). There is variation among Medicare contractors in a number of areas (54,143,353, and app. E): e their identification of specific medical tech- nologies that are not covered, e their decisions about the coverage of specific technologies, and e their implementation of national coverage decisions made by HCFA. Because of the variation among contractors, some technologies may be covered and paid for in one geographic area and not in another. There is no national or local listing of procedures that are not covered (163). Part of the variation stems from the absence of precise definition of “reasonable and neces- sary.” There are, however, specific criteria that are applied to a technology to determine if the technology meets the broad statutory language of “reasonable and necessary.” These criteria are Photo credit: The University of Utah Medical Center The Jarvik 7 artificial heart is an example of an experimental technology that has not been approved for Medicare coverage found in program instructions prepared by HCFA and sent to the Medicare contractors. The tech- nology must be (435): e generally accepted as safe and effective, ® not experimental, ¢ medically necessary, and e provided according to accepted standards of medical practice in an appropriate setting. Of particular interest to cost-containment ef- forts is that Medicare's policy is to exclude the explicit consideration of cost information in mak- ing coverage decisions. At one point, in the con- text of a proposed regulation to define the mean- ing of “reasonable and necessary” more clearly, HCFA debated establishing criteria and standards for taking nonmedical factors, including economic factors, into account in making coverage deter- minations (268). Active consideration ended with the change in administration in 1981. Another important point is that Medicare has refrained from a policy of limiting coverage of particular technologies to restricted circumstances (e.g., to institutions offering specific services or having specialized equipment, or to physicians with specific skills). Although the notion of limit- ing coverage has gained importance with the in- creasing development of sophisticated technol- ogies that require particular expertise, the dictum of refraining from interfering with medical prac- tice appears to be foremost.* On the other hand, Medicare does limit cover- age of some technologies to appropriate medical conditions. Thus, for example, in August 1981, HCFA announced the coverage of specific types of therapeutic apheresis for three conditions but 3A dramatic exception was heart transplantation. As a result of the controversial nature of the technology, including economic, social, ethical, legal, and moral concerns, the evaluation and subse- quent coverage decision has been delayed for additional research evidence, including cost-effectiveness data (107). “There appears to be a lessening of adherence to the concept of not covering technologies for limited situations. Although no pol- icy change has been announced, the coverage of apheresis, which became effective on Jan. 31, 1983, is limited to the performance of apheresis only in the inpatient or outpatient hospital setting (74). Also, in July 1983, HCFA released coverage instructions to Medi- care contractors that limits payment for a technology to its use in a specific setting and by specific providers. Closed loop blood glucose control devices will be paid for only if used in a hospital inpatient setting under the direction of specially trained medical personnel for insulin dependent diabetes during crisis intervention (383). 76 e Medical Technology and Costs of the Medicare Program denied coverage of apheresis for other indications. Three additional disease indications were added in 1983 (349). The Coverage Process Medicare's coverage process is depicted in fig- ure 2 and described in detail below. The cover- age process is generally (except for details) the same at the national and contractor levels. First, a new technology or new use of a covered tech- nology is identified. Second, a decision is made about covering the identified technology for Medi- care payment. The third and final step of the proc- ess, implementing the coverage decision, is mainly the contractors’ responsibility. Identification of New Technologies and New Uses of Technologies The identification of a new technology for a coverage determination may be done by a Medi- care contractor, by one of HCFA's 10 regional of- fices, or by HCFA's central office. Medicare con- tractors use general guidelines distributed by HCFA. In the last few years, the guidelines have been made broader and more general. HCFA as- sumes that Medicare contractors are familiar with medical and hospital practices and thus relies on the contractors’ knowledge and experience (173). Medicare contractors use various methods for identifying new technologies. In the recent past, claims review appeared to be the primary method Figure 2.—Model of Medicare’s Coverage Process for Individual Medical Technologies Stage I: Identification of new technologies and new uses of technologies National decisions Stage II: : about technologies Coverage decisions Request for y full assessment OHTA Assessment of the technology and coverage advice based on criteria of safety and efficacy and experimental status Stage III: Implementation of coverage decisions Local decisions about technologies on individual claims Request for y coverage advice Contractors’ medical consultants” Determination of whether the technology was medically necessary and appropriately delivered 8sources of coverage questions submitted to HCFA are identified in table 15. Sources of assistance for Medicare contractors’ medical consultants in making coverage decisions are identified in table 14. SOURCE: Office of Technology Assessment. Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs © 77 for identifying uncovered technologies. A 1983 survey of Medicare contractors, the results of which are presented in appendix E, however, found that most contractors learn about new tech- nologies for which coverage questions might be raised through inquiries from providers and man- ufacturers prior to the submission of claims. Other important sources of information about new tech- nologies are the drug and device approval lists from the Food and Drug Administration (FDA) and announcements from HCFA. Some contrac- tors keep informed about new developments through the medical research literature, or con- tacts with medical specialists, or supplier demon- strations and mailings. The claims form, despite the use of other tools, is still an essential, albeit imperfect, identification tool. The claims review process, described in ap- pendix E, was established for paying bills, not for identifying technologies. Furthermore, under HCFA'’s current allowances for administrative costs, Medicare contractors are financially con- strained to limit their review process. The claims form for hospital services under Medicare's traditional cost-based hospital reim- bursement method? is not designed to identify new technologies. This claims form groups services under broad headings, such as radiology and pathology, and provides little information about specific technologies (291). Furthermore, under cost-based reimbursement since 1981, HCFA has required intermediaries to examine only a 20- percent sample of inpatient hospital claims (384). On the claims form for physician services, a Part B service, the physician is required to supply information about the use of specific surgical and medical technologies. New technologies are rec- ognized by the absence of code numbers, by the presence of codes that are not recognized, or occa- sionally, by the excess charges for a type of serv- ice (54 and app. E). Because of the nature of the claims form and the fact that almost all the forms for physicians’ services are reviewed, it is com- monly assumed that carriers, the contractors who sThe claims form and the claims review process under Medicare's DRG payment method are not sufficiently established for comment at this time. As noted in ch. 2, some hospitals and some hospital units will continue to be paid under Medicare's cost-based reimburse- ment method. administer Part B of Medicare, are more likely to identify uncovered technologies from claims forms than are intermediaries. It should be noted, however, that carriers may overlook new tech- nologies because of inefficiencies in the review process and a high number of coding errors. It is also possible for physicians and hospitals to code new procedures under codes for established procedures (54). Some observers indicate that intermediaries, the contractors who administer Part A, can identify high-cost technologies. If a hospital or other in- patient facility exceeds a set level of expenditure for a particular type of service, the intermediary may examine the medical record and identify an uncovered technology. Intermediaries may com- pare annual Medicare cost reports (MCRs) from year to year to compare expenditures for groups of services. At times, hospitals have recorded a specific uncovered technology on the MCR, as well (17). On the whole, however, specific tech- nologies are not identified on the MCR, which is reviewed for financial purposes and not techno- logical use. Most Medicare contractors in the 1983 survey presented in appendix E were reasonably well sat- isfied with existing methods for identifying new technologies and reported that this was not a seri- ous problem for them. Some contractors men- tioned a need for greater cooperation between na- tional medical and insurance associations and governmental agencies in supplying information about new technologies to facilitate the identifica- tion process. Coverage Decisions Coverage decisions are made by Medicare con- tractors and by HCFA's central office. Local deci- sionmaking is informal and has no standing in law or regulation. National coverage decisions are made informally, as well, and the decisionmak- ing process has no regulatory status (114). Medicare contractors, advised by their medi- cal consultants, decide most of the coverage issues that are raised in their own geographic area. In- deed, less than 1 percent of the 250 million claims processed in fiscal year 1983 were sent to HCFA's central office for coverage decisions (88). Most 78 e Medical Technology and Costs of the Medicare Program questions are not of national interest. Further- more, the contractors view their function as pay- ing claims as quickly as possible and are disturbed by the delays in referring questions centrally (366). Although a special office dealing with coverage issues, the Office of Coverage Policy in the Bu- reau of Eligibility, Reimbursement, and Cover- age, was established in HCFA as a result of a 1979 memorandum, it does not appear that the pattern of decisionmaking has changed significantly (61). When considering coverage questions, most medical consultants to contractors appear to rely on similar sources of information, including HCFA regional offices, colleagues in other insur- ance companies or Blue Cross/Blue Shield plans, and State or national medical or specialty societies (see table 14). Most of the questions raised during claims re- view pertain to whether a particular technology was medically necessary in the case under review and whether the technology was furnished in an appropriate manner and setting. Sometimes, how- ever, the broader issue of general coverage arises, i.e., whether the technology should be covered under any circumstance. National coverage ques- tions are to be referred to HCFA's central office (435). Nonetheless, some contractors’ medical consultants make decisions about national cov- erage issues. Because of variation in the types of coverage questions that Medicare contractors consider and their decisions about any one question, the spe- cific package of covered services varies from con- tractor to contractor. The 1983 survey of Medi- care contractors found variation in coverage decisions made by medical consultants of Medi- care contractors (see app. E). In an attempt to modify differences and incon- sistencies in Medicare benefits in its region, HCFA'’s Boston Regional Office issued a bulletin in 1978 to Medicare intermediaries and carriers describing a “general approach that should be taken with respect to determining coverage of new or unusual procedures which the Medicare Bureau either has not categorized as covered or uncov- ered, or on which [it has] not advised [them] that a national coverage policy decision is currently pending” (388). The bulletin emphasized the expe- ditious use of medical consultants and suggested the referral of general issues to HCFA's regional and central offices. Although the guidelines de- scribed in the bulletin are not enforceable, the Bos- ton Regional Office believes that contractors in the Boston region have improved their coverage process and that the improvement has resulted in greater consistency in covered benefits in the area (73). HCFA's central office issued a similar directive to contractors nationwide in 1981 concerning its expectation that contractors refer coverage issues of national interest to the office (384). However, referral is not required by statute or regulation, and HCFA's request is not uniformly honored. The current locus for coverage questions within HCFA is the Office of Coverage Policy. If medi- cal advice is needed in order to arrive at a cover- age decision, the question is presented to HCFA's Physician Panel. The panel may then request tech- Table 14.—Sources of Information Used by Medical Consultants in Making Coverage Decisions Percent of consultants using source Source Yes No Colleagues in another insurance company or Blue Cross/Blue Shield Association plan.......... 78.6 21.4 HCFA regional office .......... 87.5 12.5 University medical Center ............ o.oo 44.6 55.4 PSRO (PRO) oe 26.9 73.1 National insurance association. ................. uuu 37.5 62.5 State or national medical or specialty association ..................uuurieen 75.0 25.0 Drug or device manufacturer. ................. iui 53.6 46.4 OMIT © viii norm 4505 4040 i om ow omen 9 om im = mn 3 mw 8 4 4 3 3 B® i 3 8 8 R$ LR 8 Enns a a 51.0 49.0 SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs ® 79 nology evaluations from the Office of Health Technology Assessment (OHTA),¢ which has taken over the coverage responsibilities of its predecessor, the National Center for Health Care Technology (NCHCT) (386). After conducting an assessment on the safety, efficacy, and clinical effectiveness of a technol- ogy (399), OHTA may recommend to HCFA that a technology not be covered by Medicare or that it be covered with or without restrictions. The coverage decision is made by HCFA, which subse- quently notifies its contractors and State Medic- sOHTA is located in the National Center for Health Services Re- search in the Public Health Service. It should not be confused with the congressional Office of Technology Assessment (OTA). aid agencies. In almost all cases, HCFA accepts OHTA's recommendation. In the late 1960's and throughout most of the 1970's, the vast majority of coverage questions received at HCFA's central office were submitted by the regional offices (381). From 1978 to 1982, however, the proportion of questions submitted by the regional offices declined (see table 15). After 1979, other parties, particularly manufac- turers, increased their participation in the cover- age process. In 1978 and 1979, there were no di- rect inquiries from drug or device manufacturers, but during 1981, 25 percent of the coverage ques- tions submitted to HCFA's central office were from producers of medical technologies (see table Table 15.—Sources of Coverage Questions Submitted to HCFA’s Central Office Inquiry source—by source, year, number, and percentage 1982 (through Source 1978 1979 1980 1981 November) HCFA regional office .......... 19 9 19 11 7 (61.3%) (32.1%) (59.4%) (34.6%) (24.1%) Manufacturers ................ 1 13 3 (3.1%) (25.0%) (10.3%) Hospitals/clinics .............. 1 1 1 1 (3.2%) (3.6%) (1.9%) (3.4%) Physicians ............... ou. 1 4 5 3 (3.2%) (14.3%) (9.6%) (10.3%) Medical schools .............. 2 1 (7.1%) (3.4%) Legislatures .................. 1 1 1 (3.2%) (1.9%) (3.4%) Professional societies ......... 2 5 2 2 (6.5%) (17.9%) (6.3%) (3.8%) BlueCross ........oovveennnn 1 2 2 1 (3.6%) (6.3%) (3.8%) (3.4%) Blue Shield .................. 2 (6.3%) Blue Cross/Blue Shield ........ 1 (3.4%) Private insurance ............. 1 2 (3.1%) (3.8%) Health Care Financing Administration. .............. Ba 1 2 7 (17.9%) (3.1%) 3.8%) (24.1%) Public Health Service. ......... 2 2 2 (6.3%) (3.8%) (6.9%) Attorneys ..........ouiiiinnn 1 1 (3.7%) (3.4%) Information not available. ...... 7 1 1 4 2 (22.6%) (3.6%) (3.1%) (7.7%) (6.9%) TOA) wo nomimmsmman mn ome suns 31 28 32 52 29 (100%) (100%) (100%) (98.2%) (99.6%) afFour of these questions arose from the Blue Cross/Blue Shield Medical Necessity Project. Data sources: Health Care Financing Administration, internal data sheets and HCFA staff. SOURCE: Office of Technology Assessment. 80 » Medical Technology and Costs of the Medicare Program 15). The now defunct NCHCT may have been partially responsible for the increase in manufac- turers’ questions, because it referred the manu- facturers’ inquiries it received to HCFA (258). HCFA files show that questions about coverage for a particular drug or device in some cases were submitted at approximately the same time by both the manufacturer and an interested provider. The Health Industry Manufacturers Association (HIMA) has spent considerable effort in educat- ing its members about health coverage and reim- bursement (385). Such attempts may be a signifi- cant factor in stimulating manufacturers’ interest in requesting coverage. For a few years, HIMA encouraged manufacturers to go directly to HCFA's central office to obtain coverage for their products and service. In the past year or two, HIMA has suggested that its members contact contractors, particularly carriers, because of its perception that the time required for making and releasing coverage decisions and the number of denials at the national level has increased. The decline in direct inquiries for coverage to HCFA's central office—from 25 percent of the total num- ber of inquiries in 1981 to 10.3 percent in 1982— may reflect the change in strategy. Implementation of National Coverage Decisions There is no formal mechanism for implement- ing Medicare coverage decisions made at the national level. For the most part, HCFA's func- tion is limited to disseminating the decision to con- tractors and providers through various sources, including HCFA's regional offices, instruction manuals, and transmittal letters. Government in- volvement is largely confined to cases of fraud and abuse. As noted earlier, Medicare contractors have no legal responsibility to adhere to the coverage deci- sions made by HCFA's central office. The man- ual instructions, including the coverage index ap- pendix, and the letters to the contractors are usually considered interpretive rules and thus not legally enforceable (366). Nonetheless, the contractors’ claims review process is an unofficial and limited means of im- plementing national coverage decisions. Claims requesting payment for noncovered services and claims with incompatible diagnostic and proce- dure codes are usually referred to the contractors’ nurse reviewers, and if necessary, to physician consultants (see app. E). However, no distinction can be made between those technologies that are not covered because a coverage question has not been raised and those technologies that are not covered because they have been denied coverage. Until recently, little empirical evidence was available about the contractors’ role in implement- ing national coverage decisions. A recent survey of the implementation of Medicare nursing home benefits by intermediaries found the use of skilled nursing facilities by Medicare beneficiaries to vary considerably from one State to another, a varia- tion that reflected wide differences in the inter- pretation and administration of rules governing nursing home coverage (314). The researchers concluded that the wide variation was due to the complexity of Medicare coverage rules and to Medicare's decentralized administration. The 1983 survey of Medicare contractors men- tioned above (see app. E) came to somewhat sim- ilar conclusions. The survey found considerable variation in the implementation of HCFA trans- mittals among contractors. This variation was ap- parently not related to certain characteristics of the contractors, including insurance type (Blue Cross/Blue Shield or commercial), geographic lo- cation, or claims volume. Part of the variation in implementation of cov- erage decisions appears to result from what is perceived as a lack of clarity in HCFA's cover- age instructions. Fifty-five percent of the Medi- care contractors surveyed in 1983 said they were always or almost always able to implement HCFA transmittals concerning the coverage status of par- ticular technologies without obtaining further in- terpretation. However, 45 percent of the contrac- tors reported that the transmittals sometimes, rarely, or never could be implemented without further interpretation (app. E). Some contractors also indicated they were not given sufficient time for implementing national coverage decisions, sufficient information about technologies undergoing assessment, or revisions in coverage policy. As noted earlier, HCFA's cur- rent policy is based on the premise that contrac- Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs 81 tors’ have sufficient knowledge and expertise to allow for general coverage instructions. Nonethe- less, some contractors are not content with the policy and said that the content of HCFA's in- structions could be improved by including more specific criteria and by eliminating ambiguous terms, such as “chronic” and “necessary” (app. E). The 1983 survey of Medicare contractors also examined decisionmaking by Medicare contrac- tors regarding the coverage of specific technol- ogies.” It found various degrees of variation 7 The study was exploratory and descriptive, with many metho- dological limitations (see app. E). The questions about individual technologies were intended to ascertain what the contractor's pol- icies generally were with respect to coverage. The telephone inter- viewers, researchers did not use the word “implementation” and did not check on the degree to which contractors adhere to HCFA pol- icy guidelines. Indeed, most coverage issuances allow for variation, depending on further investigation of the claim and the circumstances among contractors in their coverage of technol- ogies included in the survey (app. E). The study technologies were categorized ac- cording to HCFA coverage status: 1) explicit cov- erage by HCFA, 2) HCFA coverage with qualifi- cations, 3) no explicit HCFA policy, 4) implicit denial of coverage by HCFA, and 5) explicit denial of coverage by HCFA (see table 16). The varia- tion in coverage was least in instances in which HCFA had explicitly approved coverage. Table 16 shows that some contractors covered technol- ogies in this category with qualifications when of the individual patient. Variation among contractors was expected, because the coverage status of the technologies included in the survey are difficult to determine. Nonetheless, examining the contractors’ coverage decisionmaking process at one point in time does give some indication of the contractors’ compliance with national coverage deci- sionmaking. Table 16.— Reported Coverage Decisions by Medicare Contractors Decisions/policy of contractors Covered with Technology/HCFA coverage policy Covered Not covered qualifications Refer for advice HCFA explicitly covers: Lens implant. ........uueeinnnnnie aaa 98.2% — sr 1.8% Pacemaker: chronic second degree AV block ........... 70.9 — 20.0% 9.1 PTCA: single vessel procedure ........................ 61.1 3.7% 25.9 } 9.3 EEG monitoring: carotid endarterectomy ............... 68.0 14.0 10.0 8.0 Apheresis: hyperglobulinemias (multiple myeloma). ...... 81.1 5.7 7.5 5.7 HFCA covers with qualifications: Home blood glucose monitor ..................oennn. 11.8 7.8 72.5 7.8 External osteogenic stimulator: long bone fracture ...... 27.5 3.9 58.8 9.8 PUVA: PSOM@SIS «vive 38.5 5.8 50.0 5.8 Implantable chemotherapy infusion device: primary hepatic malignancy .............eiiiianiiiiannnnn 51.9 21.2 11.5 15.4 Implantable chemotherapy infusion device: cancer metastatic to liver .........c.coveuvnnnrnnencnsnnnnss 38.5 26.9 15.4 19.2 External insulin infusion pump .......... ci 10.6 46.8 27.7 14.9 No explicit HCFA policy—contractor decides (local option): Chelation therapy: rheumatoid arthritis ................. . 81.1 1.9 13.2 Streptokinase at cardiac catheterization: AMI ........... 30.2 45.3 15.1 9.4 Chemonucleolysis: herniated disc ..................... 64.0 10.0 14.0 12.0 HCFA denies, but not explicitly: ) Biofeedback: intractable pain ................... 0 9.3 55.6 31.5 3.7 PTCA: two or more coronary arteries... ................ 19.2 51.9 19.2 9.6 Apheresis: systemic lupus erythymatosis ............... 18.0 60.0 14.0 8.0 HCFA explicitly denies: Chelation therapy: atherosclerosis ..................... — 87.0 37 9.3 Pacemaker: sinus bradycardia without symptoms ........ 13.0 44.0 29.6 13.0 24-hour blood pressure monitoring: automatic (policy effective 7/83) ...........ccoviiiiiiiiiinn, 8.3 52.1 271 12.5 24-hour blood pressure monitoring: semiautomatic or patient activated .............. iii 4.2 774 8.3 10.4 EEG monitoring: open heart surgery ................... 15.4 71.2 9.6 3.8 Topical oxygen therapy: decubitus ulcers............... 7.8 86.3 5.9 —_ SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. 82 e Medical Technology and Costs of the Medicare Program there were no qualifications imposed by HCFA. However, this practice reflects not a lack of com- pliance with HCFA's policy, but rather caution on the part of contractors to assure that HCFA's criteria were met. From table 16, it would appear that there is sig- nificant variation among the technologies in the second category, i.e., those that HCFA covers with qualifications. However, most of the varia- tion exists because of the three infusion device technologies included in the category. When the survey was conducted, policies concerning infu- sion devices were undergoing review, and there was considerable uncertainty among contractors as to the current coverage status of the devices. If one eliminates the infusion device therapies from the category, the variation decreases con- siderably (app. E). Coverage decisions on technologies for which HCFA had explicitly denied coverage showed more variation than one might expect. That may be an artifact of the particular technologies in- cluded in the survey or may indicate a reluctance on the part of the contractors surveyed to flatly deny coverage without further investigation of the claim. Variation is, predictably, much greater for technologies for which contractors have made local coverage decisions and those for which HCFA intends to be denied but for which there is no explicitly stated policy (app. E). In most instances, the majority of the contrac- tors complied with HCFA's directives. Yet com- pliance was sufficiently diverse among the con- tractors as to result in variation. In general, the variation can be attributed to differing impres- sions on the part of the contractors about the cov- erage status of the particular technology, which may result from unclear or complex HCFA cov- erage policy, a change in policy, or a policy in the formation stage; the inherent complexity of clinical medicine and the difficulty of precisely matching a claim for a specific patient with a gen- eral policy written to cover many patients; and limitations of the study—e.g., the findings reflect responses to a hypothetical situation at one point in time. Coverage Policy Under Medicare’s DRG Hospital Payment System In October 1983, following enactment of Public Law 98-21, Medicare began phasing in prospec- tive hospital payment system using Diagnosis Related Groups (DRGs) as the case-mix measure. ® Federal regulations (114) state: . . . prospective payment legislation did not change Medicare coverage or eligibility rules cur- rently in effect . . . as a result, national coverage rules continue to be applicable. These rules will continue to be applied by intermediaries with assistance from PROs [utilization and quality con- trol peer review organizations] and PSROs [pro- fessional standards review organizations] where appropriate. DRG payment is not applicable to psychiatric hos- pitals, rehabilitative hospitals, pediatric hospitals, long-term hospitals, psychiatric and rehabilitative units operating as distinct parts of acute care hos- pitals, and physician services—provided in or out of the hospital. Thus, Medicare coverage for these institutions and services remains unaffected by the change in payment method. Although the regulations require that coverage rules remain, the structure of DRGs places little emphasis on individual technologies. Thus, under DRG payment for inpatient services, HCFA will rarely be able to discern the use of particular tech- nologies. Multiple combinations of drugs, devices, and procedures are possible within almost all DRG categories; specific technologies are not easily evident from DRG classification. For the most part, only a few of the 470 DRGs mention particular technologies. With few exceptions, spe- cific drugs and medical devices were not variables in the construction of DRGs as a patient classifica- tion system. Drugs are not specified in any of the 470 DRGs, and only one medical device, the pace- maker, is specified as or part of a DRG. Although the first major subdivision within most of the 23 major diagnostic categories (MDCs) of the classi- fication system is “the presence or absence of an See chs. 2 and 6 for further discussion of the DRG hospital pay- ment method. Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs * 83 operating room procedure” (389), the specific type or types of surgical procedures are not explicitly mentioned in most DRGs. Except for a few, such as heart transplantation surgery (DRG 103), cor- onary bypass (DRGs 106 and 107), and perhaps arthroscopy (DRG 232), the DRGs that describe most surgical procedures, e.g., pelvic procedure (DRG 334), are so general that many different sur- gical techniques could be used to carry them out (32). The inability to identify the use of particular technologies under Medicare's DRG hospital pay- ment method does not differ markedly from the situation under Medicare's previous retrospective cost-based hospital payment method. As noted above, inpatient hospital claims forms and MCRs under cost-based reimbursement do not specify the use of individual technologies. There are, however, several ways under DRG hospital payment to identify uncovered individual technologies that may raise hospitals’ costs. For example, outlier cases, cases involving either an extremely long length of stay (LOS outlier) or extraordinarily high costs (cost outlier) when com- pared to most discharges in the same DRG, will be reviewed in their entirety for noncovered or medically unnecessary or inappropriate days or services (114). Outlier cases may occur precisely because new and costly technologies were used in the care of the patient. If a new technology is not covered, outlier payments will be denied. New technologies will also be recognized dur- ing the process of adjusting DRG rates for all hos- pitals. Indeed, updating DRG weights appears to offer the most significant opportunity for identify- ing new technologies for coverage purposes. The decision to adjust DRG rates can therefore be con- sidered a quasi-coverage decision itself. Changes in DRG relative weights or prices will be made, in part, to reflect technological change. Because the DRG rate adjustment process includes identification of new technologies, it is reasonable that some of the techniques, including technol- ogy assessments, used in the process will be similar to those used for supporting coverage decisions. Indeed, the Prospective Payment Assessment Commission (ProPAC) has been given broad powers to assess medical technology and the appropriateness of medical practice patterns in de- veloping its recommendations for DRG rates. Pro- PACs role is only advisory; HCFA makes the de- cision concerning the appropriate payment rate for hospital services. Thus, both the coverage process and the proc- ess of adjusting DRG rates share a similar “ap- proval for payment” function. They differ in that a coverage determination focuses on specific tech- nologies, while adjusting DRG payment rates fo- cuses on the larger entity of a diagnostic group, which may include particular technologies. A more important difference is that the coverage process rarely considers costs, while the DRG rate adjustment process must include cost as an in- tegral issue. Nonetheless, the technology assess- ments performed for both processes may be sim- ilar. The potential for duplication is not to be ignored. The processes seem to be sufficiently sim- ilar to warrant coordinated Government effort. Whether technologies will be subject to a dou- ble review of safety and efficacy for payment pur- poses will depend on the approach chosen to up- date DRG rates. Irrespective of approach, it is reasonable to assume that hospitals’ adoption of cost-raising technologies will be made evident to HCFA for DRG payment and for coverage deter- minations. However, some approaches to up- dating DRG rates, such as through outlier cases, would not surface cost-saving technologies. In ad- dition, specific technologies will not be identified on the DRG hospital claims form, so the use by a hospital of a new, uncovered technology that lowers per case costs will not become known to HCFA through hospital claims review. HCFA in- tends to rely on physician claims and other sources for information to stimulate the initiation of a technology assessment solely for coverage purposes. As in the past, many technologies may go unnoted. 84 e Medical Technology and Costs of the Medicare Program EVALUATING TECHNOLOGIES FOR COVERAGE DECISIONS Current Activities® Before arriving at coverage decisions, both Medicare contractors and HCFA's central office have medical technologies evaluated. The evalua- tions performed for contractors by medical con- sultants are usually informal and have limited in- fluence on the diffusion of medical technologies. The evaluations performed for HCFA's central of- fice affect the diffusion of technologies nation- wide. The primary factors considered in such as- sessments are safety and effectiveness. Because cost criteria are not included as factors in assess- ments for Medicare coverage decisions, expensive technologies are eligible for coverage without re- gard to cost effectiveness. At present, the body that is responsible for evaluating the medical and scientific aspects of medical practice for HCFA is NCHCT's successor, OHTA. OHTA responds to HCFA's simple in- quiries about the regulatory and research stand- ing of particular technologies by providing infor- mation obtained from the responsible Public Health Service (PHS) agency. It also conducts “full” assessments at the request of HCFA with the objective of providing HCFA with the most current and scientifically valid information on which to base coverage decisions. The OHTA assessment process follows the process established by NCHCT. OHTA reviews the scientific literature and obtains opinions from experts in the public and private sector and then synthesizes the information it receives. In conducting its evaluations, OHTA uses numerous sources for information. If the evalua- tion concerns drugs or certain medical devices, prior evaluations by FDA provide some indica- tions of safety and efficacy; for procedures, how- ever, there is no comparable mechanism. For a drug to be covered under Medicare, FDA ap- proval is required. The use of drugs, however, is not usually questioned by HCFA. Drugs are covered for payment when provided in an inpa- *See app. C for a detailed discussion of the Government and pri- vate sector assessment activities. *°For a discussion of OHTA's method of evaluating medical tech- nology, see app. C. tient setting and their use is not monitored by HCFA; hardly any drugs are covered for payment when provided in an outpatient setting. OHTA's evaluations are confounded by defini- tional problems. The definitions of safety and ef- fectiveness, for example, differ among Govern- ment agencies. Thus, FDA considers a medical device to be effective when, on the basis of well- controlled investigations or other valid scientific evidence, the device is shown to have the effect claimed by the manufacturers under the manu- facturer’s specified conditions of use (21 U.S.C. 260c(3)). On the other hand, HCFA judges the ef- fectiveness of a medical device in terms of its ability to improve health. Thus, some devices ap- proved by FDA for marketing purposes are not covered by HCFA for payment (435). There are other definitional problems. Cover- age decisions about technologies of national in- terest are based on criteria of “general acceptance” and “stage of development.” If a technology is generally accepted by the medical community as being safe and effective (“general acceptance”) and is perceived to have moved beyond experimental status to clinical application (reasonable “stage of development”), then it is considered “reasonable and necessary.” However, the terms used in the criteria are not defined precisely. Applying the criterion of “general acceptance” to a new technology is difficult, because a new technology has usually been used by only a small fraction of the medical community. In such cases, coverage decisions are based on scientific evidence and professional judgment of safety and effective- ness. Yet standards for adequate proof of safety and effectiveness have not been established (435). The criterion of reasonable “stage of develop- ment” also creates problems in evaluating a new technology for coverage. Technologies do not progress neatly from research to development to clinical phases but more often are used simul- taneously as research and investigational tools and in medical practice (359). The distinction between an experimental and emerging technology may be arbitrary. Some contend, for example, that Medi- care reimbursement was approved for kidney Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs © 85 transplantation when the survival rate was less than it is now for liver transplantation, which is not reimbursed (43). Following its evaluation, OHTA sends its as- sessment and a recommendation based on the assessment to HCFA. The recommendation sum- marizes the evidence and PHS’ conclusions about the safety and clinical effectiveness of the tech- nology under review. The assessment is not made public routinely upon its completion; the recom- mendation is not released until HCFA has made its decision on the issue (58). For the most part, OHTA has not explicitly considered cost and cost effectiveness in its evalua- tions, although the Memorandum of Understand- ing between HCFA and PHS does not preclude this possibility (410). ProPAC is given powers to assess the cost effectiveness as well as the safety and efficacy of new and existing medical and sur- gical procedures, but its primary responsibility is to recommend changes in DRG payment rates. Although ProPAC is specifically mandated to assess medical practice in making its recommen- dations, it is unknown at this time how extensive its assessment activities will be. In addition to the Federal Government, the pri- vate sector is involved in assessing technologies. Insurers and other organizations use technology assessments for coverage and other determinations (e.g., payment, purchasing, and management de- cisions). Indeed, the present approach to medi- cal technology assessment is characterized by multiple participants from the public and private sectors and by uncoordinated activities (359). The private and nonprofit sectors have increased their involvement in the past 2 years. However, many of the assessments that are conducted are limited to specific organizational objectives and have limited value for national policy decisions. Safety and efficacy are usually used in technology assess- ments; economic, legal, social, and ethical criteria are sometimes used. Analytic Methods of Comparing Costs and Benefits Although evaluating the safety and efficacy of medical technology has protected patients from risky, unproven, and ineffective services (359), 25-337 0 - 84 - 7 the need for evaluation of the economic effects of medical technology is becoming increasingly important as Medicare and health care costs in general continue to escalate. In theory, the chance of containing Medicare costs through coverage policy would be increased by including not just safety and efficacy criteria but cost and cost-effectiveness criteria in Medi- care coverage decisions. The general value of formal analytic techniques for comparing costs and benefits, referred to collectively as cost- effectiveness analysis/cost-benefit analysis (CEA/CBA)," in decisionmaking about the use of medical technology was addressed in OTA's 1980 report The Implications of Cost-Effectiveness Analysis of Medical Technology (353). That re- port also identified the methodological strengths and weaknesses of the techniques and the poten- tial for initiating or expanding the use of CEA/ CBA in reimbursement coverage programs. CEA/CBA potentially can be more valuable for decisionmaking under a constrained budget, when tradeoffs have to made directly, than when con- straints are nonexistent or very indirect (as in most current reimbursement programs). In neither case, however, would CEA/CBA necessarily function as an effective cost-constraining mechanism or tool. Under a budget system, the budget itself would be the constraining mechanism. Under a nonconstrained system, since no direct tradeoffs are required, no direct limit on expenditures is set. Nevertheless, CEA/CBA might change the mix of expenditures (353). Medicare's DRG hospital payment system, while not a fixed budget, pro- vides more constraints than the previous cost- based reimbursement system. CEA/CBA can be conducted from a variety of perspectives, including that of the individual, the family, the hospital, the insurer, or society. Many researchers agree that societal perspective is desirable for policy decisions. When private or program benefits or costs differ from social bene- The main difference between CEA and CBA is the method of valuation of desirable consequences of a decision. In CBA, benefits (e.g., health outcomes) as well as costs are valued in monetary terms. In CEA, benefits are measured in nonmonetary units, such as years of life saved, days of morbidity avoided, or quality-adjusted life- years (which combine mortality and morbidity measures). The rea- son for a nonmonetary measure of benefits is either the impossibility or undesirability of valuing them in monetary terms. 86 * Medical Technology and Costs of the Medicare Program fits and costs (and if a private or program perspec- tive is appropriate for the analysis), the differences should be identified (353). The inherent conceptual and methodological strengths and weaknesses of CEA/CBA are ad- dressed at length in previous OTA reports (353, 359) and later publications (347,420). A metho- dological issue of particular importance to the Medicare program in performing CEA/CBA:s is whether to include discounted future medical care costs (due to longer life spans for patients resulting from the use of a medical technology) as a direct cost of the technology. Extending the life span of patients who are 65 years and older increases the chance that they will become high-cost consumers of medical care. Elderly people are particularly prone to chronic diseases.’> And, about 80 per- cent of health care resources in the United States are used for chronic disease (321). One can argue that future medical care expend- itures should be included in an analysis of a tech- nology’s influence on medical care costs. Both public and private insurers are interested in how the use of a technology will affect their future ex- penditures. It is important to recognize, however, that reducing the measured benefits of a technol- ogy by the extra medical costs attributed to longer life biases the analysis against the technology. People consume medical care and other goods and services as long as they live. “To reduce benefits by a part or all of the value of this consumption may lead to the erroneous conclusion that pro- longation of life is not worthwhile especially when consumption exceeds production” (417). Another methodological issue that is of special interest to the Medicare program is whether the human capital approach should be used in valu- ing life in the analysis of a technology. The human capital approach values life in terms of earning potential; health outcomes are valued in terms of the economic productivity they permit (420). Be- cause of its emphasis on economic productivity, the human capital approach values the lives of young people more than the lives of elderly peo- About 1 out of 10 40- to 50-year-old people has a chronic dis- ease; by age 65, approximately 2 people out of 10 have a chronic disease; after 75 years of age, the number increases to 4 out of 10; and by age 90, almost 9 of every 10 people have a chronic disease (164). ple. The bias of the approach is most explicit in CBA. CEA, however, has built-in value judg- ments, i.e., once money is allocated to save lives, the value of life is implied. In CBA, the analyst must choose a value to complete the analysis; in CEA, the policymaker chooses the value, albeit indirectly (353). It is important to bear in mind that CEA/CBA does not necessarily or easily take into account social values, moral judgments, legal implications or political realities. Thus, it does not easily or commonly address issues of equity and distri- bution. The power of CEA/CBA is diluted in many in- stances not only by methodological problems but by a lack of efficacy data on which to base cost- effectiveness calculations. Some of the difficulties associated with these techniques will diminish with time, but others will not. An analysis rarely can account for the vast range of applications of a specific technology and the technology's un- predicted effects. The setting of care, volume of use, and the practice of medicine also influence the cost effectiveness of a technology. Despite its limitations, CEA/CBA provides an analytical basis for integrating the economic aspects of a decision about medical technology with the health aspects. It can be very helpful in assisting the policymaker in structuring a prob- lem and understanding its ramifications. As Fuchs says, “given the will and the mechanism . . . CEA/CBA offers the most rational human basis for effective, efficient allocation” (129). Cost-Saving Technology Definitional Complexities The technical complexity of determining the cost effects of emerging and new technologies is compounded by the problem of defining a cost- saving or cost-raising technology. Differences in perspective impede arrival at a universal defini- tion of a cost-saving or a cost-raising technology. The disparity in perspective may bring about con- flict between parties because of the limited supply of resources for medical care.?® If resources are Costs of medical services are the economic resources (e.g., equip- ment, supplies, professional and nonprofessional labor, and the use of buildings) consumed in the provision of those services (423). Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs ¢ 87 used by one party, they are not available for use by another. The question is: From whose perspective should the cost effects of a technology be examined? Are the cost implications of a technology to be con- sidered from the perspective of the individuals, the hospital, the insurer, one Government pro- gram, the entire Government, society as a whole, or others? Each party has its own view of the cost effects of a particular technology, because health expenditures affect each differently. A technol- ogy that is considered cost saving by one may be considered cost raising by another. An associated question is: Is a cost-saving technology one that saves costs in the present or in the future? For purposes of this study, a cost-saving tech- nology might best be defined as a technology that saves Medicare program costs. Yet this definition does not distinguish between Medicare costs and societal costs—a distinction that may be required for policy decisions. Does the definition imply that the development, adoption and use of technol- ogies that save Medicare program costs but raise societal costs should be encouraged or discour- aged? One rationale for the Government's role in the health care system is “the promotion of the allocation of resources in the collective interest of the population of patients and potential pa- tients” (423). Those who act on the belief that “Government agencies that plan and regulate the distribution of medical services may be viewed as agents for society” (423) would discourage tech- nologies that decrease Medicare program costs but increase societal costs. Those who act on the belief that “Government agencies develop bureaucratic and organizational objectives that may not be consonant with the broader public interest” (423) would encourage technologies that decrease Medi- care costs but increase societal costs. Developing Criteria for Identifying Cost-Saving Technologies Considering the difficulties in defining a cost- saving or cost-raising technology, it is not sur- prising that research on developing criteria to identify cost-saving technologies is in an early stage. One type of technology that can be iden- tified as cost saving is a technology that substi- tutes exactly or very nearly for another and is also cheaper. The Shouldice method of hernia repair, for example, both substitutes for other methods of repairing hernia and is performed less expen- sively (166). On the other hand, while it is generally accepted that the automated clinical chemical analyzer sub- stitutes for previous nonautomated chemical tests, it is not clear whether the automated analyzer is producing more units of a formerly performed service or instead is producing a new type of lab- oratory analysis. Such distinctions, although often subtle and difficult to identify or measure, are crucial in analyzing the cost implications of tech- nology. Another problem in developing criteria is that the cost implications of a technology vary with the technology's stage of development: an assess- ment performed at one time may yield a different result when performed at another time. Technol- ogies change over time. The dosages of a drug may be refined. New generations of devices re- place old ones. Surgical procedures are modified. New, unanticipated uses, particularly for diag- nostic technologies, are discovered.'* Thus, the indications for using the technology change and its potential benefits and costs change over time. Furthermore, the cost implications of a particu- lar technology depend on the setting of care. For example, computerized energy management and bacterial susceptibility testing allegedly save hos- pital costs, but only if the size of the hospital and its volume are large enough to justify the invest- ment (418). Thus, a technology that saves costs in one hospital may raise costs in another. Because of the variation among hospitals, reports about the cost-saving nature of one or another technol- ogy may be only partially accurate and have to be interpreted cautiously. 14For example, the computed tomography (CT) scanner, a high capital investment instrument, was thought in the 1970's to increase the per unit cost of services (3). The current notion is that the CT scanner in fact may lower the per unit cost of health services by substituting for expensive, invasive procedures (420). If used appro- priately, the CT scanner has the potential for decreasing health care costs, but it may increase health care costs when used inappropriately (347). 88 ¢ Medical Technology and Costs of the Medicare Program Attempts have been made to categorize tech- nologies in hopes that distinctions can be made among types of technologies on the basis of ex- pected costs and benefits (283). If all technologies within a category had similar pressures on cost and delivered similar benefits, a coverage policy could be designed using a categorical approach (283). Technologies with a common medical pur- pose (e.g., therapy) have some characteristics in common, and CEAs are conducted with them in mind. For example, the analysis of a treatment DISCUSSION In the past, coverage policy has had important potential, but limited opportunity, to contain Medicare program costs by influencing the dif- fusion of medical technology. The Medicare pro- vision that a technology must be covered before it can be paid for, however, may have protected beneficiaries from the indiscriminate adoption and use of technology and possibly from some unsafe and ineffective medical technologies. The deliberate use of Medicare coverage to in- fluence the adoption and use of medical technol- ogy, however, has been limited by several fac- tors. These include Section 1801 of the Social Security Act, which prohibits Federal interference in the practice of medicine and the manner in which services are provided; the imprecise phrase “reasonable and necessary” governing the intro- duction of technology into the Medicare program; the decentralized mechanism for promulgating and implementing Medicare coverage policy; and the wide discretion allowed individual contrac- tors in making coverage decisions. Inadequacies in the current Medicare coverage process contribute to the circumscribed role of coverage policy in the rational diffusion of appro- priate technology (26). The current coverage proc- ess does not ensure identification of all important new technologies that are introduced into the Medicare program or those covered technologies whose safety and efficacy has not been proved. Indeed, the universe of new technologies that are introduced into the Medicare program through contractors’ coverage decisions is not known. The technology considers the effect of the technology on morbidity, disability, and/or mortality. But, it is difficult to identify the cost implications of such characteristics. There are, as well, charac- teristics not held in common among all technol- ogies in a medical purpose category that may af- fect costs. A method that is sufficiently sensitive to predict the cost implications of a particular technology based on its medical purpose category has not yet been developed. possibility that some new technologies are not identified but are paid for cannot be ignored. Medicare contractors vary in whether they cover (and then pay for) some particular technologies and in the extent to which they refer technologies to HCFA's central office for national coverage decisions. The absence of formal, legally binding requirements that contractors comply with na- tional coverage decisions leads to the lack of uni- form implementation of such decisions and a dis- parity in the coverage of technologies across the country. Tightening the coverage process would no doubt save money for Medicare and would pro- vide for a more rational diffusion of medical tech- nology. It might also ensure equal access to the same technologies by Medicare beneficiaries. With respect to centralization, however, caution is nec- essary. A nationally controlled coverage process might not take into account the unique needs of all patients and could be administratively expen- sive. Thus, a decision to reduce variation in cov- erage policy and increase the explicitness and uni- formity of Medicare benefits would require careful judgment and balance. Reconciling a more cen- tralized coverage system with the independent practice of medicine could be a potential politi- cal problem. The current process of evaluating medical tech- nology at the national level may need modifica- tion in order to achieve the more rational diffusion of technology. Although the safety and effective- ness of technologies that are brought to the na- Ch. 5—Specific Medical Technologies: Linking Coverage Policy and Technology Assessment To Contain Costs © 89 tional level for coverage decisions are evaluated more rigorously there than at the contractor's level, the information available for an evaluation is often limited. The amount and type of data available vary for drugs, devices, and procedures. New drugs are subject to FDA's premarket ap- proval requirements, and data on their safety and efficacy are usually available. For a drug to be considered for coverage, it must have FDA ap- proval. (There is, however, no monitoring of drug use in hospitals.) Medical devices, depending on their classification, are subject to general controls, performance standards, or premarket approval re- quirements for safety and effectiveness by FDA."® However, few clinical trials have been performed on new medical devices submitted for coverage approval. Also, FDA's definition of effectiveness differs from that used for coverage. Analytic data on medical and surgical procedures, which are not subject to FDA regulatory requirements, are even less available than data on medical devices. The guidelines used by OHTA to evaluate the safety, efficacy, and clinical effectiveness of med- ical technology stress the value of basing cover- age judgments on information derived from con- trolled clinical trials or other well-designed clinical studies. Unfortunately, there is a dearth of such information—information from clinical trials of any type was available for only 2 of 12 full assess- ments published by OHTA in 1982 (352). Further- more, even when such information is available, it often has limited value for coverage purposes. Very few clinical trials are designed to compare competing technologies. Most assess only one technology. Comparing the results of studies is often misleading because patient populations and study designs differ markedly. Few trials use in- 15For more information, see OTA's forthcoming report Federal Policies and the Medical Devices Industry (345). dividuals 65 years and older in the study popula- tion. Furthermore, few trials deal with questions about indications for use of a technology, and their results are not available soon enough to be included in a coverage decision. The lack of quan- titative data is one force behind the statute (Pub- lic Law 98-21) that allows HCFA to fund con- trolled clinical trials of concern to the Medicare program. Medicare may have further constrained the po- tential of coverage policy to influence the diffu- sion of technology by not limiting the coverage of certain technologies to specific providers and specific sites of care and by not explicitly consid- ering costs in coverage decisions. In theory, one way to use coverage policy for containing pro- gram costs would be to include cost criteria in the assessment of medical technologies. But it appears that incorporating cost criteria into an assessment does not necessarily lead to the identification of cost-saving technology. Not only is a cost-effec- tive technology not necessarily a cost-saving tech- nology, but many experts contend that the defini- tional and methodological uncertainties of CEA/ CBAs have not been sufficiently resolved for the use of these techniques as allocation tools. These techniques cannot provide decisions but can aid decisionmaking when used in conjunction with other kinds of information (353). Finally, the relationship between coverage pol- icy and DRG hospital payment needs further ex- ploration. The importance of coverage policy will be only marginally diminished by DRG payment. Furthermore, the newly created ProPAC will assess DRG payment rates in association with the technologies that might be incorporated into those DRGs—an activity that represents the first explicit merging of costs and effectiveness in the Medi- care program. 6. Medicare Hospital Payment and Medical Technology Probably every new and eagerly expected garment ever put on since clothes came in, fell a trifle short of the wearer's expectation. Charles Dickens Contents ap De bs We SOR Tain Ll rn SAR a i Las Prospective Payment of Hospitals ............... 0 0. cidade Evolution of Prospective Payment Programs. ....................... Key Features of Prospective Payment Systems ...................... Effects of Prospective Payment on Medical Technology .............. Medicare's DRG Payment System and Medical Technology .......... Per Capita Prospective Payment... v.......c.o. 0. iiiie ts biol Areawide Global Prospective Budgeting: The Rochester Area Hospitals Corp. ........;............ 0.000 Capital Payment Methods... .......c.. ........ 00 hs caidas History of Capital Payment Under Medicare ....................... Capital Payment Under State Prospective Payment Programs. ........ Capital Payment Alternatives Under DRG Payment ................. Limited Provider Contracting... oui... ova ri idii odibliden ea Increased Beneficiary Cost-Sharing for Hospital Services ............... ConCISIONS ee RE sl LIST OF TABLES Table No. 17, Sources of Hospital Revenue, 1080... ............. oo. . oo ii 18. Mandatory State-Level Hospital Prospective Payment Programs gsofJunedO83- 00 a ee Rene 19. Impact of Technological Innovation on Per Case Costs Under DRG Payment... .. 0h... ais ad iia FIGURE Figure No. 3. Options for Capital Payment Under Medicare's DRG Hospital Payment System .....................i vii ils fd a 101 6. Medicare Hospital Payment and Medical Technology INTRODUCTION In 1982, over 70 percent of Medicare's pay- ments was for hospital care (135). Furthermore, $34.6 billion (66.3 percent) of Medicare's $52.2 billion! expenditures was for inpatient hospital care (135,151), and between 1967 and 1982, Medi- care program expenditures for inpatient hospital services increased at an annual rate of 19.2 per- cent (151,392). Given the importance of expend- itures for hospital care in the Medicare budget, it is not surprising that both Congress and the ex- ecutive branch have had a longstanding interest in controlling the growth of Medicare expendi- tures for hospital care. Not only are hospitals important to Medicare, but the Medicare program and its methods of pay- ment are crucial to hospitals. In 1982, Medicare accounted for 35 percent of hospitals’ revenues (15). In 1980, as shown in table 17, Medicare and Medicaid together accounted for an estimated 42 percent of the revenues of a sample of short-term non-Federal hospitals. Because Medicaid hospi- tal payment has traditionally followed the Medi- care method in most States, the level and method of payment adopted by Medicare governs a siza- ble share of the total revenue of many U.S. hos- pitals. How Medicare chooses to pay hospitals— 1This $52.2 billion includes administrative expenses as well as ben- efit payments of $50.9 billion (135). Table 17.—Sources of Hospital Revenue, 1980 Revenue source Percent of total revenue MBOICEIE. .. .comissmivm ssa mmesmes ns 34.1% Medical cv: u:smermeimns ws men smeems 7.8 Blue CroSS ...ovviiiieiieaeiieeann 16.1 COMMBICIB css 555m 40s virn a mm sym sums 22.4 Other government sources ........... 2.3 All other revenue®................... 17.3 aBased on a sample of 4,989 non-Federal, not-for-profit, short-term general hospitals. bincludes operating, nonpatient care, and nonoperating revenues. SOURCE: J. Feder, J. Hadley, and R. Mullner, “Poor People and Poor Hospitals: Implications for Public Policy,” J. Health Polit. Policy Law, forthcoming. what it will pay for, how much it pays, and how it computes the level of payment—is therefore an issue of primary importance to hospitals, to the Medicare program, to communities, and to Medi- care beneficiaries. Furthermore, hospitals are the major provider of medical technology, particu- larly sophisticated capital-intensive diagnostic and therapeutic procedures. Hospitals provide these technologies to both inpatients and outpatients. Changes in Medicare's hospital inpatient policy are therefore likely to affect the availability of medical technology for both kinds of patients. As described in chapter 2, the Social Security Amendments of 1983 (Public Law 98-21) man- dated a change in Medicare's inpatient hospital payment system from a retrospective, cost-based system to a prospective system of payment based on per-case prices for patients in 470 separate Diagnosis Related Groups (DRGs). DRGs are a set of patient classes developed to reflect dif- ferences in resource needs among different kinds of patients. Medicare's DRG hospital payment system is to be phased in over a 3-year period beginning in October 1983. The initial set of DRG prices is based on the 1981 average inpatient oper- ating costs per case for each DRG in a 20-percent sample of Medicare claims. The prices will be up- dated regularly and will be adjusted for each hos- pital’s urban or rural location and area wage rate. They will apply to virtually all short-term acute- care general hospitals in the United States. Under DRG payment, the hospital-specific maximum limit on the amount of inpatient operating costs per case that will be reimbursed will continue to be designated by the Tax Equity and Fiscal Re- sponsibility Act of 1982 (TEFRA) (Public Law 97- 248). Capital expenditures (depreciation and in- terest and return on equity for for-profit hospi- tals), direct teaching expenses, and expenses for outpatient services will remain “pass-through” items (i.e., items not subject to controls) as they 93 94 e Medical Technology and Costs of the Medicare Program were under TEFRA until the end of the transition period. Congress contemplated, but did not spe- cify the method for, the incorporation of payment for capital into the DRG pricing system at the end of the 3-year period. This chapter describes and analyzes how the in- centives in Medicare's DRG prospective hospital payment system may affect medical technology adoption and use. Despite the recent establishment of the new hospital payment strategy for Medi- care, it is quite possible that other approaches to hospital payment will be actively considered in the future. Part of the reason is that pressure for cost containment at the Federal level may con- tinue, and part is that individual States may enact hospital cost control systems in which Medicare will agree to participate. Four alternative ap- proaches to hospital payment that have been sug- gested or applied by public or private payers which might be considered for implementation by Medicare are considered in this chapter: * alternative prospective payment programs and modifications of Medicare's current DRG system, ® capital payment methods, ® limited provider contracts, and ® increased beneficiary cost-sharing for hos- pital services. The alternatives discussed in this chapter per- tain only to payment of hospital care. They are not mutually exclusive, but are separated for ease of discussion.? Broader alternatives to the current Medicare payment methods, such as Medicare vouchers, which integrate payment for hospital services with that for other health care services, are discussed in chapter 8. Nevertheless, it is worth noting here that strategies based on the con- tinued independence of payment for hospital serv- ices from that of other health care services repre- sent a limited field of opportunity for reform of Medicare payment. *For example, capital payment must be considered for all hospital payment alternatives. PROSPECTIVE PAYMENT OF HOSPITALS Evolution of Prospective Payment Programs The Medicare program is not the only third- party payer that has used retrospective, cost-based payment for hospital care over the years. In 1976 and 1981, for example, about one-half and one- third, respectively, of the Blue Cross plans reim- bursed hospitals retrospectively on the basis of their costs (8,153). Furthermore, until 1981, State Medicaid programs were required to follow Medi- care’s principles of reimbursement for hospitals unless they applied for, and received, a waiver from the Federal Government for an alternative system. In the late 1960's, however, some States and private third-party payers began the search for alternatives to retrospective, cost-based reim- bursement, and alternative payment schemes, broadly termed “prospective payment,” appeared in State and voluntary programs throughout the 1970's. Prospective payment programs vary wide- ly, but they all have two features in common: * the amount that a hospital is paid for serv- ices is set prior to the delivery of those serv- ices; and * the hospital is at least partially at risk for losses, or is able to gain from surpluses that accrue during the payment period, or both. The litmus test of whether a payment system is prospective is the extent to which a hospital's own decisions will alter the payment rate. Medicare's DRG payment system for hospitals is a particu- lar type of prospective payment. The DRG sys- tem represents the culmination of years of experi- mentation with alternative forms of prospective payment by the States and private third-party payers. Ch. 6—Medicare Hospital Payment and Medical Technology * 95 The development by States of early prospec- tive payment systems was spurred primarily by rising insurance premiums and Medicaid budgets. In several States, severe financial crises prompted their immediate implementation (155). Federal support of State-run prospective payment exper- iments was authorized by the Social Security Amendments of 1972 (Public Law 92-603), which gave added impetus to their development. Four States—Maryland, Massachusetts, New Jersey, and New York—have been granted waivers from Medicare's current payment system. These waiv- ers give the State's rate-setting agencies the author- ity to set Medicare rates. The oldest prospective payment program, spon- sored by Blue Cross of Indiana, began in 1959 and remains in existence today. The first Blue Cross- sponsored prospective payment plans were volun- tary programs in which participating hospitals’ upcoming budgets were reviewed and approved by an appointed committee. The Blue Cross plan would then pay its share of the budgeted costs, rather than actual costs. Despite the fact that these systems were (or are) voluntary, hospitals were encouraged to participate because of the impor- tance of Blue Cross as a source of revenues. The negotiated hospital budgets sometimes covered self-pay and commercially insured patients as well as patients in the Blue Cross plan. As might be expected given their voluntary nature and limited coverage, however, the Blue Cross-sponsored budget review programs have not been particu- larly successful in moderating cost increases (311). State-mandated prospective payment programs have varied widely in their methods of payment and in the payers covered. The earliest State-run prospective program began in New York in 1970 and covered Medicaid and Blue Cross? (154). By 1979, there were 13 State-legislated prospective payment programs in effect throughout the United States (7), but the participation of hospitals was mandatory in only 10. These State programs evolved as their shortcomings were uncovered and as the crisis in hospital costs grew. Thus, current State-run prospective payment programs not only 3Medicare joined the program in 1980, and hospital charges made to self-pay patients and commercial insurance companies were fro- zen in 1978. vary among themselves at present but also do not resemble their earlier forms. In 1981, under the Omnibus Budget Reconcilia- tion Act of 1981 (Public Law 97-35), State Med- icaid programs were given expanded authority to deviate from Medicare reasonable cost reimburse- ment methods. As a result, several Medicaid pro- grams have recently enacted prospective payment systems. Some of these Medicaid programs are part of statewide systems; other Medicaid pro- grams have their own prospective payment systems. Table 18 summarizes the status of mandatory State-level hospital prospective payment pro- grams as of June 1983. At that time, 10 States had a State-legislated program, and another 17 States had a Medicaid-only hospital prospective pay- ment program. Other States had State or local . prospective payment programs in which partici- pation and/or compliance was voluntary. One unique approach to hospital payment, discussed later in this chapter, is used in the Rochester, N.Y., area. Bills for the establishment of hospi- tal rate-setting programs are currently under ac- tive consideration in six States: Florida, Pennsyl- vania, Ohio, Illinois, Indiana, and Wisconsin. Key Features of Prospective Payment Systems The details of program administration vary widely among prospective payment plans. Never- theless, there are five components of prospective payment design that are likely to affect the incen- tives of hospitals to produce services and use tech- nologies, and ultimately to influence the cost of hospital and health care. These are discussed below: ¢ Payers covered —The more payer classes cov- ered by the prospective payment program, the more likely it is that the system will mod- ify a hospital's decisions regarding the avail- ability and use of services. ¢ Unit of payment—All prospective payment programs set hospital rates either explicitly or implicitly on the basis of one of five units of payment: 1) per service rates; 2) per diem 96 ¢ Medical Technology and Costs of the Medicare Program Table 18.—Mandatory State-Level Hospital Prospective Payment Programs as of June 1983 Medicaid-only State-legislated Medicare participating in Medicaid participating in program program State-legislated program State-legislated program Alabama .......ovsamirnesnesis CAIOINIA, i» avs sine sms 5 500 5 5:0 0 308 Colorado ...................... Connecticut ................... FIORGOR +. cs 5m 5 5 we 5 win 5 50m 5 wis 4 00 Georgia... Idaho ......................... MINOIS coc viviis ius nn ovis ams aims ws Kentucky ...................... Maryland ...................... MAING . ; io uvio soins aoe s wim b 58 8 53 4 ain Massachusetts ................. Michigan ...................... Minnesota? .................... Mississippi... Missouri ............ aL. NEW JOrS8Y ... i vvissenmssncsmess New York ..................... North Carolina ................. Pennsylvania. .....« sswsvirnnss Rhode Island .................. MHAIY oz i oes 0 55.500 4 5 6 on 0 oui 6 dt 0 VIPQIR ss visi nan» iw ows wim» im 5 3 Washington... .................. X Wisconsin ..................... X XxX XxX XXXXX XXX xX XXX x xX XX XX XX Total number of States........ 17 10 @Minnesota’s system is mandatory in the sense that Blue Cross will not pay more than allowed rates. SOURCE: U.S. Department of Health and Human Services, Health Care Financing Administration, unpublished data, May 1982; and R. Lyman, U.S. Department of Health and Human Services, Health Care Financing Administration, personal communication, June 20, 1983. rates; 3) per case rates (per admission rates); 4) per case rates with adjustments for case mix;* and 5) per capita rates. Set prices for individual services give hospitals incentives to produce these services efficiently but also give them incentives to provide more serv- ices to each patient (389). A fixed payment for each day of stay establishes incentives to reduce the number of services offered per day but to increase lengths of stay as well. There is strong empirical evidence that these effects from a fixed rate of payment per diem actual- ly occur (431). A fixed rate of payment per admission encourages low use of resources per stay and short lengths of stay but also establishes incentives for hospitals to increase admissions and to engage in “cream-skim- ‘Case mix— The relative frequency of various types of patients, reflecting different needs for hospital resources. There are many ways of measuring case mix, some based on patients’ diagnoses or the severity of their illnesses, some on the utilization of services, and some on the characteristics of the hospital or area in which it is lo- cated (343). ming” to avoid more serious (and costly) pa- tient types. Per case rates that vary by type of case, as in the Medicare DRG payment system, reduce but do not eliminate the in- centive for cream-skimming but leave intact the incentive to increase admissions. Per cap- ita rates, on the other hand, reduce the in- centive to admit patients but encourage hos- pitals to agree to serve only healthier groups. Scope of coverage—Prospective payment programs vary according to the elements of hospitals’ costs covered in the rates. The in- clusion or exclusion of capital costs, teaching costs, and outpatient services in the prospec- tive rate are important determinants of how the hospital and other providers behave, and ultimately how medical technologies are used. Hospitals have an incentive to shift services from cost categories with prospec- tively set limits to those that are still treated as “pass-through” items. Extent to which a hospital's own costs deter- mine the level of payment—To be truly pro- Ch. 6—Medicare Hospital Payment and Medical Technology * 97 spective, a payment system should render a hospital unable to increase the level of pay- ment by increasing its own costs. If the cur- rent payment rate depends on the hospital's past costs, with a 1- or 2-year lag, then the system is not very independent of the hos- pital’s own cost experience, and the incen- tives in cost-based reimbursement will pre- dominate. On the other hand, if the rate is set on the basis of the hospital's own costs in a base year with infrequent updating ex- cept for general inflation indexing, the rela- tionship between cost and prospective price is effectively severed. Likewise, the use of comparative cost data from other hospitals to establish a hospital's payment limit also severs the link between the hospital's own past behavior and its prospective rate and will give the hospital more incentive to con- tain costs. Program stringency —Under any particular payment method, a program can be either generous or restrictive. A stringent program is one in which the payment level is low rel- ative to the cost of providing services. Pro- gram stringency is a function of numerous aspects of a program's design. For example, a program may or may not allow hospitals to keep any surpluses generated as a result of cost-saving behavior, or it may penalize hospitals more or less severely when costs run above the prospective level. Although over- all program stringency is difficult to meas- ure precisely, a rough indicator is the ratio of net revenues to total expenses from those patients covered under the prospective pay- ment program. If the ratio is high, then the program is relatively generous; if it is low, then the program is stringent. Under a strin- gent method, the hospital must cover oper- ating losses with cash from gifts, deprecia- tion allowances, or revenues from other sources. Some have argued that program stringency is the most important factor in the effective- ness of prospective payment programs in re- ducing hospital costs (64). The stringency of the program is likely to affect the ability of the hospital to generate capital from both in- ternal and external sources. Hospitals oper- ating under more stringent programs there- fore would be likely to face higher capital costs, which in turn would affect the adop- tion of capital-intensive medical technologies. Effects of Prospective Payment on Medical Technology There is evidence to suggest that in recent years, some State-level mandatory prospective payment programs have had a moderating influence on hospitals’ costs (36,64,309,310). A recent study of 15 State or areawide hospital rate-setting pro- grams found that the rate of increase in total ex- penses per hospital admission was reduced by ap- proximately 2 to 5 percent between 1969 and 1978 in seven States with mandatory programs (64). Apparently, prospective payment programs need time to mature before they begin to influence hos- pital costs (310). Using national data for the period 1969-80, Sloan estimated that mandatory State prospective payment systems ultimately reduce hospital expenses per adjusted admission by an average of 13 percent relative to States without such systems (309). That prospective payment can reduce hospital expenditures does not necessarily imply that the use of medical technologies will decrease. In fact, one would expect prospective payment to have different kinds of effects on different technologies (see ch. 3). One indirect method of examining hos- pitals’ technology choices under prospective pay- ment is to study the effect of prospective payment on a related measure: hospitals’ choices of labor and nonlabor inputs to the production of serv- ices. If prospective payment makes hospitals more efficient producers of services, both a reduction in the total value of inputs and substitution among different kinds of inputs should be observed. Studies of the impact of prospective payment on hospitals’ use of labor have documented both a decrease in the total number of hospital employ- ees and a substitution of less expensive for more expensive personnel in States with strong rate- setting systems (132,179,187). There is no evi- dence on whether hospitals reduce nonlabor inputs under prospective payment systems or whether such systems lead to substitution between 98 e Medical Technology and Costs of the Medicare Program capital- or material-intensive inputs and per- sonnel. There is some evidence regarding the effect of prospective payment on the provision of medi- cal services. As part of a comprehensive study of nine State-legislated hospital rate-setting systems, Worthington and Piro (431) found that programs that pay hospitals on the basis of a per diem rate all produced an increase in the average length of stay and the occupancy rate. This result would be expected from a per diem rate-setting system, because the longer the patients stay, the more rev- enue the hospitals receive. Although a per diem rate-setting program would also be expected to encourage increases in inpatient admission rates, no such admission effects were found in this study. These results suggest that manipulating hospital admission rates may be more difficult than increasing the length of hospital stay for pa- tients already admitted. Taken as a whole, how- ever, the results do suggest that decisionmakers in hospitals respond in predictable ways to finan- cial incentives for the use of hospital services. A related question with important long-run im- plications for medical technology is whether pro- spective payment influences the extent and speed of hospitals’ adoption of new technologies. Studies of the impact of hospital prospective payment programs on the adoption of new capital equip- ment or “technology-intensive” services suggest that prospective payment can have an effect on technology adoption and that the nature of the effect depends on both the specific attributes of the program and the characteristics of the new technology. Joskow (182), who analyzed the effect of rate- setting programs on the availability of computed tomography (CT) scanning in hospitals, found the number of CT scanners in a State in 1980 to be negatively related to the number of years that rate- setting had been in effect in the State.® Hospital rate-setting also led to a shift in the location of CT scanners to physicians’ offices. Cromwell and Kanak (77) recently analyzed the impact of 15 State rate-setting programs on the availability of 13 different services in the hospi- *The analysis controlled for possible impacts of other types of hospital regulation, including certificate-of-need (CON) laws. tal between 1969 and 1978. New York's rate- setting program, a restrictive program that pays per diem rates, had the most consistently nega- tive effects on the availability of all types of serv- ices. New Jersey's pre-DRG prospective payment system also appeared generally to reduce the availability of complex services. Other States’ pro- grams showed no consistent impact on service adoption. It is interesting to note that the service upon which rate-setting had the largest and most widespread negative effect is social work, a labor- intensive, not equipment-intensive, hospital service. In still another study of hospital payment and technology diffusion, Wagner, et al. (418), in- vestigated the impact of prospective payment in three States (New York, Maryland, and Indiana) on the adoption of five new pieces of capital equipment: electronic fetal monitoring, gastroen- doscopy, volumetric infusion pumps, automated bacterial susceptibility testing, and computerized energy management systems. The first three tech- nologies probably raise the daily cost of care, al- though their effect on the average cost per case is unknown. The latter two are investments in cost-saving equipment. The New York rate-setting system was found consistently to lead to adop- tion of fewer units of cost-raising technologies and to increase the probability of large hospitals’ adopting cost-saving equipment. However, the prospective payment programs in Maryland and Indiana showed no such consistent effects on hos- pitals’ adoption behavior. Together these studies imply that prospective payment does affect the adoption of new technol- ogy in predictable ways but that much depends on the strength and design of the program. New York's system, the oldest and most restrictive rate- setting program, has clearly altered the extent of availability of new technology. Other systems may be too new, too small, or too generous to show long-run consequences. Medicare’s DRG Payment System and Medical Technology Features of Medicare's DRG prospective hos- pital payment system currently being imple- mented create strong incentives for hospitals and Ch. 6—Medicare Hospital Payment and Medical Technology * 99 other providers to use and adopt technologies in ways that are different from those under cost- based reimbursement. Although the DRG hospi- tal payment system does not include payers other than Medicare, the per case payment approach provides incentives to hospitals both to reduce the amount of resources expended per stay and to selectively encourage admissions.® These incen- tives are strengthened because the payments hos- pitals receive for treating patients are essentially unaffected by the hospital's own costs and because the system puts hospitals entirely at risk for losses. During the 3-year period of the DRG system's implementation, several key elements of hospi- tal expenses, including capital, teaching, and out- patient expenses, will continue to be paid as pass- throughs. Inputs that are passed-through become less expensive to the hospital relative to inputs subject to controls and theoretically should be used more by hospitals. To discourage hospitals from increasing capital expenditures in antici- pation of future controls, the Social Security Amendments of 1983 (Public Law 98-21) delib- erately left uncertain whether capital expenditures for new projects begun in the first 3 years of the law's implementation would or would not be in- cluded in prospective per case limits at the end of the 3-year period. Although there is no empirical evidence on the effect of the new DRG inpatient hospital payment system, it is possible to describe how the incen- tives inherent in the system may alter the use of medical technologies. First, however, it must be stressed that financial incentives are but one of several influences on hospitals’ decisions to adopt and use technologies. DRG payment will not have a uniform effect on medical technologies, and in some instances, technologies will be subject to conflicting incentives. In general, the following can be concluded.’ The incentive to selectively increase admissions is somewhat mod- erated by the fact that the hospitals receive different payments for different types of cases. It is not entirely abolished, however, be- cause of the imperfections in the use of DRGs as the case-mix measure. ¢ Under DRG payment, the number and inten- sity of ancillary procedures provided to in- patients can be expected to decrease overall, but the use of ancillary procedures that can be shown to lower the cost per case can be expected to increase. The settings of technology use are likely to be influenced by DRG payment, but the in- centives work in conflicting directions and are sensitive to the key features of program design. For example, in the absence of a method for excluding very low-cost patients from the DRG pricing system, DRG payment encourages inpatient admissions for simple procedures. It remains to be seen which in- centive will dominate for which procedures. DRG payment will encourage the movement of technologies, particularly those for post- hospital care, into the home and other non- hospital sites of care. DRG payment is likely to influence the spe- cialization of services, but the magnitude and direction of these effects is unknown. The in- centives to reduce costs encourage concen- tration of capital-intensive technologies in fewer institutions. Conversely, increasing competition among hospitals for physicians and patients may create incentives for the widespread acquisition of some technologies. A change in technology product mix is likely to result from downward pressure on the price and quantity of supplies and, if capital 1 Photo credit: E. I. du Pont de Nemours & Co. "These conclusions are based on a detailed analysis of DRG pay- ment conducted as part of this project and published in a technical memorandum entitled Diagnosis Related Groups (DRGs) and the Medicare Program: Implications for Medical Technology (343). DRG payment will affect the settings of medical technology use. Some services provided formerly in the hospital will be provided in increasing numbers on an outpatient basis 100 ¢ Medical Technology and Costs of the Medicare Program is included in the DRG rate, capital equip- ment. Greater product standardization can be expected as more expensive models and procedures are eased out of the market through competition. It is also possible to describe the incentives regarding the adoption of new technologies and the discarding of old technologies under DRG payment. Many observers have pointed out that per case payment systems in which future levels of payment are largely independent of the hos- pital’s own historical costs create incentives for hospitals to adopt cost-saving technologies—i.e., technologies whose adoption decreases the hos- pital’s total cost per case. Given that technologies are neither cost saving nor cost raising independ- ent of the context in which they are used, how- ever, the same technologies are not likely to be adopted by all hospitals. The introduction of new capital-intensive cost-saving technologies in a DRG payment environment is likely to encour- age specialization among hospitals as small hospi- tals become unable to reap the cost-saving benefits of some investments. Some technologies that de- pend on high volume to be cost saving might be provided to smaller hospitals on a contract basis by large hospitals or independent laboratories. The feasibility of such contractual arrangements would vary depending on the specific uses of tech- nology and the geographical and competitive en- vironment in which the hospitals operate. The financial incentive to introduce new cost- raising technologies (i.e., technologies whose adoption increases the hospitals’ total cost per case) is lessened, but not eliminated, under DRG payment compared to cost-based payment. Under cost-based payment, the additional hospital costs of new technologies are covered; hospitals there- fore have no financial incentives not to adopt such technologies. Under DRG payment, a hospital's adoption of new cost-raising technology is not met ~ with an automatic increase in revenues to cover the additional cost. New technology will have to compete with alternative uses of hospital funds, such as employee wage and benefit increases or additional nursing staff. New technology may be at an additional disadvantage relative to other uses of funds because of the relative uncertainty about its benefits in the early stages of diffusion (282). The implications are obvious: with limited resources, hospitals will need to assess new tech- nologies more closely and ration resources more carefully. Nevertheless, the introduction of promising new technologies, even those that are cost rais- ing to the hospital, will be attractive to hospitals as they compete for physician loyalties and, ulti- mately, the admissions they represent. For exam- ple, despite its high capital and operating cost, nuclear magnetic resonance imaging, a new med- ical imaging technology still in the investigational phase, may be highly desirable to hospitals who seek to protect their admissions base from en- croachment by other hospitals. The importance of this incentive as a constraint on the incentive not to adopt new technologies is unknown. Thus, although DRG payment does not imply that tech- nological change will approach a standstill, the directions of such change are likely to be altered. Overall, the adoption of technologies that are cost raising to the hospital is likely to decline by an unknown quantity. Per Capita Prospective Payment Per capita prospective payment to hospitals refers to an array of methods for paying rates to hospitals in advance based on the number of en- rolled or covered individuals rather than on the services provided, days of care, or admissions. A single organization consisting of one or more hospitals takes responsibility for providing all covered hospital benefits to the beneficiary dur- ing a specific period of time in exchange for a fixed rate of payment. Beneficiaries may or may not be limited in their choice of place of hospitaliza- tion. Under free choice plans, the hospital receiv- ing the per capita payment would have its pay- ment reduced by the payments made to other hospitals for services to the beneficiary. Presum- ably, the hospital receiving the per capita pay- “ ment has a financial incentive to see to it that the amount and intensity and, therefore, the cost of hospital care is reduced. Per capita hospital payment is one of the oldest types of hospital payment mechanisms (303). This approach emerged from individual hospitals’ ef- forts in the 1920's to stabilize their revenue Ch. 6—Medicare Hospital Payment and Medical Technology © 101 sources. As the notion of beneficiary free choice gained prominence with nonprofit health insur- ance plans, however, payment of hospitals on a unit-of-service basis became the norm. Thus, for almost 20 years—from 1946, when the last in- dividual hospital per capita payment plan closed down, until 1964 when the Colorado Blue Cross plan initiated such a program in one county — per capita payment of hospitals was nonexistent in the United States (303). The Colorado Blue Cross program lasted 18 months but was abol- ished with the emergence of Medicare and Med- icaid (303). In 1980, the Blue Cross and Blue Shield Associ- ation, with funding from the John A. Hartford Foundation, initiated two demonstrations of per capita hospital payment involving beneficiary free choice (90). In 1982, 10 hospitals in 2 States (North Dakota and Massachusetts) began receiving per capita payments in an effort to demonstrate that per capita payment would lead to more efficient provision of hospital care (90). Individual hospi- tals in these demonstrations have no control over the beneficiaries “assigned” to them. The assign- ments are made either by the beneficiaries them- selves or by the Blue Cross plan. Per capita rates vary with beneficiary characteristics and are based on the hospitalization experience of each group. Thus, an attempt is being made to avoid the prob- lems of cream-skimming—the tendency of hos- pitals to avoid serving sicker populations—and adverse selection—the tendency of high-risk ben- eficiaries to seek out more comprehensive and costly care. Unlike per case prospective payment, per cap- ita hospital payment is designed to encourage hospitals to reduce the rate of admission to hospitals and to expand the use of pre- and post- hospital care outside of the inpatient setting. Since beneficiaries in the demonstrations maintain the right to choose freely their hospitalization site, the success of the program appears to hinge on the degree of cooperation between hospitals and physicians in managing the hospital resources effi- ciently. Data are not yet available on the effect of the demonstrations on the use of hospital and other health care services or on hospital costs. How- 25-337 0 - 84 - 8 ever, two observations can be made about the ap- plicability of such an approach to Medicare. First, implementation of per capita payment (especially with free choice of provider) depends on the avail- ability of excellent patient-based data systems through which estimates of per capita costs can be made. Second, the process of “assigning” ben- eficiaries to “home” hospitals may be feasible in certain communities but may not be universally feasible. Especially when beneficiaries maintain the freedom to choose their place of care, per cap- ita hospital payment may depend for success on the existence of specific hospital market envi- ronments. As with any prospective payment system that encourages hospitals to become more economi- cal in the provision of services, per capita pay- ment runs the risk of discouraging hospitals from providing needed care. In this case, the risk is to underadmit as well as to underprovide services while in the hospital. Finally, per capita payment would be likely to affect the adoption of medical technologies in ways that are similar, but not iden- tical, to the expected effects of DRG payment. For example, whereas DRG payment encourages hos- pitals to adopt new technologies which will bring in new admissions, per capita payment has the opposite effect. Areawide Global Prospective Budgeting: The Rochester Area Hospitals’ Corp. Nine hospitals in the Rochester, N.Y., area have been experimenting since 1980 with a global pro- spective budgeting approach to hospital payment (40,280). The nine member hospitals of the Roch- ester Area Hospitals’ Corp. (RAHC), accept a cap on the aggregate revenues they can receive each year. The areawide revenue cap is developed from the member hospitals’ actual costs in a base year (1978) with adjustments for inflation and techno- logical change in subsequent years. Individual hospitals’ budgets are allocated from the total in a similar fashion, but some of the revenues are withheld in a special fund to adjust for the cost of increases in admissions and for new equipment or facilities approved by both RAHC and the State. The RAHC board has the responsibility for distributing these special funds among hospitals 102 » Medical Technology and Costs of the Medicare Program (319). The hospitals are collectively responsible for keeping expenses within the global budget; if areawide admissions increase to the point that the special fund is exceeded, the hospitals must share in absorbing the excess costs. The experimental global budgeting program in Rochester clearly involves negotiation among hos- pitals in dividing up available funds for service expansion and modernization. Whether the con- cept can survive in the long run remains to be seen. In any case, Rochester has some special at- tributes that make it a particularly fertile envi- ronment for such an experiment. First, the hos- pitals’ market area is relatively self-contained and well defined. A global budgeting approach would not be feasible where market areas are not easily defined. Second, the Rochester area has a long history of business, community, and hospital sup- port for areawide planning (319). An earlier ex- periment funded by the Federal Government, which included a larger area of New York State CAPITAL PAYMENT METHODS The use of new medical technology often de- pends on hospitals’ making investments in capi- tal facilities and equipment. Hospitals’ willingness and ability to make these investments in “equip- ment-embodied technology” (233) are influenced by the third-party hospital payment system. Con- sequently, an important issue with respect to the diffusion of medical technology is how Medicare pays for hospital capital. It is important to recognize that the effects of any capital payment method depend on the larger payment system of which the capital payment method is a part. Thus, for example, a particular approach to capital payment would have different effects on the use of medical technologies under a cost-based reimbursement system than under a prospective payment system. The history of pay- ment for capital under Medicare's traditional cost- based hospital reimbursement system and under State prospective payment systems is described below. Also discussed below are capital payment options under Medicare's DRG prospective pay- ment system, which is currently being imple- as well as Rochester, collapsed (319), probably because these two factors were not present. If successful, Rochester's global budgeting ap- proach would represent an application of alloca- tion of health care resources through a political process. There is clearly an incentive for each hos- pital to increase its admissions at the expense of others, but without access to funds for new serv- ices, the member hospitals may be effectively con- strained from doing so. It is not clear how ration- ing resources on a community level would affect specialization of services or regionalization of fa- cilities. Evidence on the effects of the Rochester experiment is not available at this time. The po- tential for such an approach as a method for ra- tioning hospital technologies may be great in a few areas of the country where social, political, and demographic conditions are right, but for the reasons given above, it is not likely to be viable as a general approach to hospital payment. mented. The capital payment method under DRG prospective payment was left unresolved in the Social Security Amendments of 1983 (Public Law 98-21), so detailed examination of the implications for medical technology of various capital payment approaches under DRG payment is both impor- tant and timely. History of Capital Payment Under Medicare Medicare's cost-based hospital reimbursement system has had strong effects on hospitals’ capi- tal acquisition decisions. Traditionally, Medicare has reimbursed hospitals for interest and histori- cal cost depreciation expenses associated with all capital equipment purchases, regardless of wheth- er the equipment was purchased or leased. In theory, when combined with cost-based reim- bursement of operating costs, this approach to capital payment encourages hospitals to invest in new facilities and equipment and to finance as much as possible through debt. If a hospital per- Ch. 6—Medicare Hospital Payment and Medical Technology © 103 sistently “funds” payments for depreciation (i.e., establishes a separate fund that cannot be used for operating expenses), cost-based reimbursement provides the cash necessary to amortize whatever level of debt the hospital incurs. If all hospital revenues were derived from cost-based payers, capital, as well as all other inputs, could be used in unlimited quantities. The situation was more complicated in prac- tice. Some experts have claimed that Medicare's cost-based payment system did not pay its fair share of hospital costs, because it did not pay for a share of bad debts and charity care attributable to patients other than Medicare beneficiaries (71). Consequently, hospitals with high burdens of un- paid care and large numbers of Medicare and Medicaid beneficiaries have incurred deficits. Such deficits could only be accommodated by the cash flow generated through depreciation payments or other sources of cash such as unrestricted gifts or non-patient-care revenues. Hospitals with a high proportion of patients covered by insurance plans paying on the basis of charges have had an advan- tage over others, because they could recover the costs not covered by Medicare for these patients by raising their charges. Other factors being equal, hospitals with a high proportion of patients sub- ject to cost-based reimbursement are likely to re- ceive lower bond ratings® than other hospitals (70). Thus, while cost-based reimbursement under Medicare increased hospitals’ demand for capital, over time it has also made it more difficult and costly for some hospitals to obtain additional debt financing. Capital Payment Under State Prospective Payment Programs With exceptions noted below, those responsi- ble for designing State prospective payment pro- grams have been reluctant to deviate from cost- based reimbursement for capital. Even New York's otherwise restrictive rate-setting program, for example, continues to treat capital expenses as pass-throughs, much as in the Medicare sys- tem (154). Payment for capital in some States has sBond rating is an assessment of the credit worthiness of a hospi- tal by a rating agency such as Standard & Poor's Corp. or Moody's Investors Service. been even more generous than under Medicare. Washington, Massachusetts, and Minnesota have used price level depreciation rather than histori- cal cost depreciation (76). Maryland and New Jersey calculate a capital facilities allowance for buildings and fixed equipment that provides cash sufficient to pay off existing debt and to ac- cumulate a down payment for replacement or ad- ditions. In all of these cases, the hospital's pay- ment rate depends on its own capital expenditures. States have generally looked to the direct reg- ulation of capital expenditures through certificate- of-need (CON) laws as a capital rationing device. CON laws require that hospitals receive approval from a State health planning agency for major capital investments. Although there is a general consensus that capital expenditure regulation as implemented in the States has not been effective in reducing the level of capital expenditures (69,414), most rate-setting programs assume that the appropriateness of capital investments will be judged through CON. The only deviations from cost-based reimburse- ment of capital have been Maryland's and New Jersey's approach to payment for major movable equipment (e.g., beds, diagnostic instruments) and western Pennsylvania's approach to capital pay- ment. In Maryland, the depreciation on major movable equipment available in the hospital in a base year is adjusted in subsequent years by an appropriate inflation factor. The allowance for movable equipment is unaffected by the hospi- tal’s subsequent capital expenditure decisions, ex- cept for special cases in which the rate-setting commission may make exceptions (185). New Jersey's DRG system has a similar method for ma- jor movable equipment. The allowance is adjusted for inflation from a base year, which is updated periodically. Unlike Maryland's allowance, how- ever, the rate of reimbursement for major movable equipment is a blend of the hospital's own capital costs and those of other peer group hospitals (416). In the western Pennsylvania pro- gram, hospitals have a choice of historical cost depreciation not exceeding 4 percent of other allowable costs or any other method that results in less than 3 percent of costs (76). Thus, hospi- tals in western Pennsylvania are subject to an ef- fective ceiling on capital reimbursement. 104 » Medical Technology and Costs of the Medicare Program Capital Payment Alternatives Under DRG Payment What method of payment for capital will even- tually be adopted under Medicare's DRG hospi- tal payment system remains to be seen. The De- partment of Health and Human Services is required by law to report to Congress by October 20, 1984, on a recommended approach to capital payment. For the time being, capital expenditures by hospital will continue to be reimbursed as before—on a cost basis. Figure 3 presents the major alternatives for cap- ital payment under Medicare's newly created DRG hospital payment system. The fundamental issue under DRG payment is whether a hospital's cap- ital payment should or should not be subject to some kind of externally imposed limit. Pass-through reimbursement of capital could continue as a permanent feature of DRG payment. As shown in figure 3, there are three alternative pass-through approaches: ® payment of historical cost depreciation and interest expenses, ® payment by price level depreciation, and ® payment of debt service requirements. In all of them, the level of payment is directly linked to the amount of capital investment under- taken by the hospital. Payment of historical cost depreciation and in- terest expenses represents a continuation of the traditional Medicare method. Payment by price- level depreciation, with hospitals paid according to the current value of the capital assets used up in any year, would be more generous than his- torical cost depreciation in an inflationary econ- omy. Finally, payment by Medicare of its share of the hospital's debt service requirements would match the flow of capital payments over time more closely with the actual flow of debt pay- ments. In any of these cases, the hospital would receive cash sufficient to cover its debt over the lifetime of an asset.’ ’Investments using equity capital rather than debt instruments could be paid depreciation and a return on equity. The issue of which hospitals should receive a return on equity capital and how high this return should be is clearly important to the hospital industry, but it is not a central issue with respect to medical technology and is therefore not discussed in this chapter. Figure 3.—Options for Capital Payment Under Medicare’s DRG Hospital Payment System Controlled payments Pass-through payment ~~ Sl ital Historical Uniform Hospital- Sl cost Price level Debt service payments specific regulation depreciation play ot on requirements across payments (project-by- + interest > Minisy + return on hospitals project + return on tre un on equity approval) equity equity Flat Base year Percent of Plat [ote 8s Co Cos own vate per percent + inflationary operating bey one trend costs SOURCE: Office of Technology Assessment. Ch. 6—Medicare Hospital Payment and Medical Technology © 105 Methods of limiting capital payment fall into three categories: e methods that establish uniform rates of pay- ment across all hospitals (or all within a class), e methods that establish hospital-specific con- trolled rates, and e methods that condition payment on approval of capital expenditure projects. The uniform payment approach would treat all hospitals alike, regardless of their capital or oper- ating expenditures. The uniform payment could be calculated either as a fixed percentage of the DRG price or as a flat rate per bed. Hos- pital-specific approaches, on the other hand, would take the hospital's capital or operating costs into account in establishing a level of payment but would limit increases in the payment level over time. Thus, for example, capital payments could be limited to a percent of operating costs, so that hospitals with high operating costs would receive a higher capital payment than others; al- ternatively, the capital payment in any year could be tied to the hospital's actual capital costs (as measured by interest and depreciation) in a base year with adjustments for inflation in subsequent years. If capital payments were controlled through direct regulation of capital expenditures, only projects approved by a CON or other designated agency would be recognized by Medicare for cap- ital payment. Approved projects would then be paid on a cost basis. Areawide or statewide an- nual capital expenditure limits could be used to establish an upper bound on the value of ap- proved projects. The State of New York is cur- rently considering adoption of such a capital ex- penditure limit. The alternative capital payment methods de- scribed above can be evaluated on the basis of four general criteria: e Efficiency—The extent to which the ap- proach promotes the cost-effective use of hos- pital technology. An ideal method would not distort the relative costs to the hospital of capital and other inputs, would discourage needless duplication of capital-intensive serv- ices in the community, would encourage spe- cialization and regionalization of services when appropriate, and would minimize the cost of providing any given level of hospital care. e Equity of access to medical technology—The extent to which the method promotes equal access to capital-embodied medical technol- ogy. An ideal method would not deny peo- ple living in certain regions or with low in- comes access to medical technology that is available to others. e Fairness—The extent to which the method treats all kinds of hospitals alike. An ideal system in this respect would not reward or penalize hospitals according to their owner- ship status, location, or other factors that lie outside management's control. e Feasibility— The extent to which the method is administratively workable and politically acceptable. An ideal method would involve low administrative costs, minimize the prob- lems of transition from the old to the new payment method, and accommodate the in- herent cyclic nature of hospital investments. A permanent capital cost pass-through under DRG payment violates the efficiency criterion, be- cause it distorts incentives for hospitals to adopt and use capital-embodied technologies. Table 19 shows how a pass-through for capital expendi- tures influences hospitals’ incentives to adopt four different kinds of hospital technology under DRG per case payment. So long as the effect of medi- cal technology acquisition on a hospital's total cost per case is in the same direction as its effect on operating costs, for example, the method of capital payment under DRG payment will not alter the direction of the incentives for technol- ogy adoption. Thus, there are disincetives under DRG payment to adopt most quality enhancing, cost-raising (Type I) technologies regardless of the way in which capital is handled. Capital cost pass- throughs weaken the disincentive to adopt such technologies, but they do not remove it. New technologies with high capital costs but only small increases in operating costs would be affected less by DRG payment with a capital pass-through than by DRG payment with capital built into the rate. Of course, since DRG payment sets up incen- tives for hospitals to increase admissions, hospitals 106 » Medical Technology and Costs of the Medicare Program Table 19.—Impact of Technological Innovation on Per Case Costs Under DRG Payment Direction of effect on costs: Incentives for technology adoption Capital cost Operating cost Total cost With capital Without capital Type of innovation per case per case per case in rate in rate I. Cost-raising, quality-enhancing new technology ...................... ... + + + Il. Operating cost-saving innovations A. Raises total costs ................... - + B. Saves total costs .................... + . Capital cost-saving innovations A. Raises total costs ................... = B. Saves total costs .................... - IV. Servicelprocedure disadoption ............ — SOURCE: Office of Technology Assessment. can be expected to seek cost-raising technologies whose availability promises to bring in profitable admissions. A capital cost pass-through essentially subsidizes this kind of investment, leading poten- tially to wasteful duplication of these services among hospitals. In the case of operating-cost-saving (Type II) and capital-cost-saving (Type III) technologies, the incentives for hospitals to adopt may actually be reversed by the policy regarding payment for cap- ital. Of particular concern is the incentive under a pass-through to adopt expensive capital equip- ment that reduces operating costs but raises total cost per case. Automated hospital information systems, for example, could be evaluated in terms of their ability to reduce operating costs with in- adequate regard for their impact on total costs. The more labor-saving capital-intensive system would be preferred regardless of its net impact on costs. Over time, then, hospitals could be expected to become more capital intensive than efficiency would dictate. Despite the inefficiency inherent in a capital cost pass-through, this approach does well on the other three criteria. Its feasibility has been demonstrated through the years. It is inherently fair because all hospitals are treated alike in their payment. Final- ly, it poses no barriers to equal access to medical technology, although it does nothing to redress current inequities. Public and inner-city hospitals tend to have lower ratios of capital to total costs than other hospitals, because these hospitals have older facilities and possibly less equipment-em- ———— ——— -— | | f + + | + | f bodied technology (20,188,356).1° A cost-based system of reimbursement for capital would con- tinue to pay these hospitals relatively less than other hospitals. Because public and inner-city hos- pitals typically have a high burden of bad debts and indigent care, their reimbursement from de- preciation costs is often used to assist their cash flow to subsidize the operating costs associated with this uncompensated care. Since it is unlikely that the implementation of DRG payment will do much to change the situation, hospitals with low levels of capital assets will continue to receive low payment under a capital cost pass-through. Despite the low level of capital payment, some hospitals serving poor areas may find their ability to raise capital enhanced by DRG payment with a capital cost pass-through. These hospitals now have the potential to generate operating savings which could be used to offset operating cost losses on bad debts and charity care. Controlled capital payment is generally more efficient than pass-through capital payment, be- cause the hospital is encouraged to provide its care at the least possible cost. New technologies would be judged by hospitals in terms of their impact on total costs, not just on operating costs. Hos- pitals would be further encouraged to specialize and join in plans for regionalization of health serv- ices. However, it is difficult to devise a controlled 1°For example, the mean percentage of capital costs to total costs for non-Federal public hospitals in 1981 was 5.2 compared to 6.7 for private not-for-profit hospitals and 8.9 in for-profit hospitals (15). Ch. 6—Medicare Hospital Payment and Medical Technology ¢ 107 payment system that is fair to all hospitals. In a uniform payment system, hospitals that in the past have had lower ratios of capital to operat- ing cost would receive more than they had in the past, while those with high ratios would receive less. Thus, public hospitals would, at least in the near term, fare better under a uniform system than they had in the past. But uniform payment of cap- ital also favors multihospital systems, because it allows these affiliated hospitals to pool capital payments and smooth out fluctuations in capital expenditures across hospitals.’ A uniform rate of payment would also create a difficult and pos- sibly costly transition if hospitals that have made major investments in recent years are not to be unduly penalized. The American Hospital Asso- ciation has recently proposed a uniform capital payment system that would pay each hospital the higher of cost-based reimbursement or a fixed payment rate during a 10-year phase-in period (9). Andersen and Ginsberg have suggested a less gen- erous transition in which “budget neutrality”*? is maintained by gradually reducing the proportion of the capital payment that is a pass-through (15). Tying the capital payment to the level of capital costs in a base year or to the hospital's operating costs is efficient but may be unfair. This approach tends to reward the hospitals that were most cap- ital intensive in the past, leaving those with low levels of capitalization forever to receive lower payments. Moreover, this approach would not work well for hospitals requiring major capital expenditures in the early years of implementation. Perhaps for these reasons, support for this ap- proach has been limited to the movable equipment portion, which typically has shorter lifetimes and lower variations in asset values among hospitals. LAt present, approximately 26 percent of all community hospi- tals are members of multihospital systems. About 11 percent of these systems are proprietary; the remainder are government or not-for- profit systems (51). 12“Budget neutrality” means (as specified in the Social Security Amendments of 1983) that the aggregate payments for the operating costs of inpatient hospital services in fiscal years 1984 and 1985 will be neither more nor less than would have been paid under the Tax Equity and Fiscal Responsibility Act of 1982 for the costs of the same services. It is difficult to predict the effects of direct reg- ulation of capital expenditures through CON or other agencies. Direct regulation can occur with or without statewide or areawide maximum limits on the total capital outlays over a given period, and the effects can be expected to differ between the two. Although there has been much discus- sion in certain States about establishing capital expenditure limits or “pooling” capital, all experi- ence to date has been with CON and Section 1122 programs which do not operate with areawide or statewide limits. The experience with capital ex- penditures regulation in the absence of such limits has been disappointing, with most evaluations concluding that the level of capital expenditures has not been affected (65,69,251,414). Moreover, the distribution of medical technologies among hospitals does not appear to have been improved as a result of CON (65). The institution of an annual (or perhaps, long- term) limit on the level of capital expenditures that can be approved by CON agencies would, if it were strictly enforced, ensure that the program has an effect on the total level of capital expendi- tures. But there is no evidence, either theoretical or empirical, to suggest that the outcome of such a regulatory process would be either efficient or fair (417). A review of the literature on resource allocation decisions by committees revealed that the ultimate outcomes depend both on chance and on the composition of the committee and the pro- cedures governing the decisionmaking process (417). Moreover, the kinds of information needed to make informed tradeoffs among competing capital projects is likely to be unavailable, thus leaving the process even more exposed to politi- cal solutions. Regardless of whether or not an areawide limit is applied, direct regulation of capital expenditures is administratively feasible only for large proj- ects—construction and renovation projects and major new services. The current trend toward high thresholds for inclusion in capital expendi- ture controls (256) would probably continue, leav- ing an ever larger proportion of capital-embodied technology needing to be controlled in some other way. 108 Medical Technology and Costs of the Medicare Program LIMITED PROVIDER CONTRACTING Another approach to hospital payment has arisen following recent legislation. In 1981, under Section 2175 of the Omnibus Budget Reconcilia- tion Act of 1981 (Public Law 97-35), Congress gave States greater flexibility in restricting Med- icaid beneficiaries’ freedom to choose their pro- viders of medical care. Under this statute, State Medicaid agencies are permitted to apply for waivers from the freedom-of-choice provision of the Social Security Act. Most waivers to date have been for case management systems that restrict the providers from whom a Medicaid beneficiary can obtain primary care (332). The implementa- tion of a hospital-only contracting approach, whereby contracts are negotiated with selected hospitals for provision of inpatient care to Med- icaid patients, is nevertheless possible. California has recently adopted this approach (226). The two fundamental variables of hospital con- tracting are: 1) the rules used to determine which hospitals can serve beneficiaries, and 2) the method used to determine the level of payment for such services. A hospital contracting system could be administered by a negotiated bidding sys- tem where the bid price is computed on a per case, per day, or per service basis. Or, the third-party payers could select eligible hospitals on the basis of their per case costs (with or without adjust- ments for case mix) and either pay a flat prospec- tive rate or continue to reimburse the hospitals on a retrospective cost basis. In either case, by tying a hospital's receipt of revenues to its ability to constrain costs, contracting may encourage hospitals to keep costs low. The impact of contracting on technology use and on equality of access to medical technology depends to a large extent on the contracting pro- gram'’s design and administration. For example, California's Medicaid program uses a bidding sys- tem that encourages hospitals to choose a price for contracted patients that cover short-run, but not long-run, incremental costs of treating those patients. In essence, other classes of payers may subsidize the Medicaid program by bearing more than their share of overhead and other fixed costs. At the time that contracting was instituted in California, hospitals in the State were suffering from very low occupancy rates, a condition which encourages hospitals to offer to treat patients at rates below the long-run incremental costs of treating them (172). If the contracting agency selected low-cost providers and then paid a pro- spective rate based on fully allocated costs, the cost shift to other payers would be eliminated. Selected provider contracting can have serious implications regarding the equality of access of beneficiaries to medical technologies. If benefici- aries can receive hospital care only in low-cost hospitals, the availability of certain technologies, particularly newer ones, could be restricted. Moreover, as the revenues of contracting hospi- tals are held down relative to those with other kinds of patients, the discrepancies could widen over time. With Medicare accounting for about 35 percent of community hospital revenues nationwide, se- lected contracting with hospitals by Medicare would probably be highly disruptive and would greatly change the patient mix of hospitals. It would be difficult to contract selectively without turning some hospitals into predominately Medi- care hospitals, leaving the others to serve private sector patients. This kind of separation of care by payer class is a necessary condition for the de- velopment of a two-class hospital system and would represent the abandonment of the prin- ciples of equal access on which Medicare was founded. INCREASED BENEFICIARY COST-SHARING FOR HOSPITAL SERVICES Still another approach to affecting the use of medical technologies through hospital payment is to increase beneficiary cost-sharing. One exam- ple of this approach was found in the fiscal year 1984 budget request of the Reagan Administra- tion. The Administration proposed an increase in Ch. 6—Medicare Hospital Payment and Medical Technology 109 Medicare beneficiary cost-sharing for inpatient hospital services (224). Under the proposal, the beneficiary would continue to pay a deductible approximately equal to the cost of a day's care but would be responsible for additional payments of 8 percent of the deductible for days 2 to 15 and 5 percent of the deductible for days 16 to 60. In exchange, the beneficiary would no longer be re- sponsible for cost-sharing after 60 days. Increasing patient cost-sharing has the imme- diate benefit to the Medicare program of shifting the expenditure burden from Medicare to the ben- eficiary or other third-party payers. Proponents of increased cost-sharing for inpatient hospital services also contend that patients will have greater incentives to resist unnecessary tech- nology—admissions, long stays, and procedures or services offered as part of the stay. Thus, ac- cording to this argument, the Medicare program would benefit from these behavioral influences on the use of medical technology due to increased cost-sharing. Interim results from the Rand National Health Insurance Study, a well-designed experiment, in- dicate that the level of patient cost-sharing does influence the use of hospital services (243). This experiment randomly assigned 2,756 families whose members were not older than 61 with in- comes under $25,000 (in 1973 dollars) to one of six insurance plans with differing levels of deduct- ibles, coinsurance rates, and upper limits on an- nual out-of-pocket expenditures. Plans with high levels of coinsurance had lower admission rates per capita than plans with low rates of coinsur- ance. With coinsurance rates of 50 or 95 percent, hospital admission rates for adults were, respec- tively, about 60 and 40 percent below those with no cost-sharing. However, the annual expendi- ture per hospitalized patient showed no consist- ent or significant relation to the level of cost- sharing. Of the patients admitted to the hospital, 70 percent exceeded their catastrophic limit. Thus, while patient cost-sharing affects hospital admis- sion rates, it appears to be “a poor instrument for affecting costs once patients are admitted” (243). It should also be noted that the Rand experi- ment involved cost-sharing for all covered serv- ices, not just hospital care. Part of the decline in rates of hospitalization may have been due to a decline in ambulatory care visits that would other- wise have generated a hospital stay. A hospital- only cost-sharing provision with a catastrophic limit on out-of-pocket expenses might not result in the reductions in hospitalization rates experi- enced in the study. Moreover, applicability of the Rand study to the Medicare program is limited by its inclusion only of a nonelderly population. It is not known whether the elderly would respond to cost-sharing in the same way or to the same degree. Indeed, evidence from this and another short-term study of hospital cost-sharing indicates that cost-sharing’s effects on hospital use vary with the patient's age and sex (243,413). The effect of any beneficiary cost-sharing pro- posal on the use of medical technology must be considered in the context of a specific method of hospital payment. Under cost-based reimburse- ment, hospitals had no financial incentive to re- duce occupancy rates or the volume of technol- ogy use; increasing coinsurance rates, as opposed to increasing deductible amounts, could con- ceivably have made patients better consumers of care in the hospital. (However, available evidence does not support this contention.) Under the DRG inpatient hospital payment system, hospital ad- ministrators have incentives to implement policies that reduce the length of stay and the use of un- necessary ancillary technologies. Increasing hos- pital coinsurance rates would probably have lit- tle additional influence. Increasing the deductible for hospital admissions, on the other hand, might be more consistent with per case payment. In per case payment, the hospital has an incentive to selectively increase the number of admissions and readmissions. At present, the deductible of $356 upon hospitalization covers any readmission with- in 60 days of the original episode (Social Secu- rity Act, secs. 1861(a) and 1813(a)). To the extent that hospital-only cost-sharing can be expected to reduce the rate of hospital admissions in the elderly, requiring a second deductible for rehos- pitalizations within 60 days would counteract the incentive for hospitals to discharge and readmit patients for elective procedures that could be per- formed during a single stay. Whether cost-sharing for inpatient hospital care can moderate the use of hospitals and their tech- 110 » Medical Technology and Costs of the Medicare Program nology depends to a large degree on the will- ingness of Medicare beneficiaries to circumvent cost-sharing by purchasing private supplementary (“Medigap”) insurance covering deductibles and coinsurance. In 1976, approximately 63 percent of aged beneficiaries had some form of private supplementary coverage, and 14 percent were eligible for public support, mainly through Med- icaid (199). According to one study, most Medi- gap insurance covers deductibles and coinsurance (217), and joint Medicare/Medicaid beneficiaries have no cost-sharing requirements. These supple- mentary plans dilute the impact of present cost- sharing provisions. An increase in Medicare cost-sharing for hos- pital services would be expected to raise premiums of Medigap policies. The effect of such premium increases on Medigap enrollment has not been studied. Data are available on the correlation be- CONCLUSIONS The Medicare program has recently embarked in a new direction in hospital payment with its DRG prospective payment system for inpatient services. The implications of this approach for the use and adoption of medical technology are var- ied and to some extent uncertain. Much will de- pend on the way in which the program is imple- mented and the changes or refinements that may come in the future. The overall DRG price level and the rates of increase permitted over time will have a great deal to do with hospitals’ ability to adopt new medical technologies. If, as many have claimed, there is substantial room for increased efficiency in the provision of hospital care, and if the payment level is reasonably generous, DRG payment under Medicare could provide hospitals with substantial surpluses of funds that could be used to provide new technologies and services. If, on the other hand, increases in rates are re- strictive or set at a level that reclaims all the cost savings made in the previous years from hospi- tals, then hospitals would probably find it diffi- cult to finance new cost-raising technologies and services. The results cannot be predicted at this time. tween family income and the supplementation of Medicare with private insurance. One study found that, although purchase of supplementary insur- ance increases with family income, the differences are small across income groups when the avail- ability of public programs, particularly Medicaid, are accounted for (34). Another study found that Medicare families with low incomes were just as likely to purchase private Medigap insurance as families with higher incomes, a finding which may suggest that the elderly are relatively insensitive to premiums in their demand for Medigap cov- erage (216). If this is the case, increasing hospital deductibles under Medicare would serve mainly as a means of transferring the burden of expend- iture from the public to the private sector, with a particularly heavy burden on the near-poor, without substantially altering patterns of hospital use. The way in which capital is paid for under DRG payment is a critical issue for medical technology. Permanent continuation of a capital cost pass- through under DRG payment would be inefficient and would ultimately distort hospitals’ capital in- vestment decisions, making hospitals too capital intensive. Externally controlled capital payments, on the other hand, are efficient but are difficult to administer effectively or fairly. Approaches to prospective payment of hospi- tals other than DRG hospital payment are cer- tainly possible. Innovative prospective payment methods such as per capita hospital payment and areawide global budgeting may hold promise in some areas. Furthermore, the current Medicare law encourages States to experiment with these as part of all-payer systems. In addition, case-mix classification systems with more desirable prop- erties than DRGs may become available in the future (343). HCFA expects to hold statewide systems apply- ing for waivers to the cost-containment standard of DRG payment: for Medicare to join in a State system, the State must provide strong evidence Ch. 6—Medicare Hospital Payment and Medical Technology © 111 that Medicare's inpatient hospital expenditures will be at least as low as they would be under DRG payment (112). Although the available evi- dence supports the contention that the hospital cost increases under prospective payment systems implemented by individual States in the mid to late 1970's were lower than those under Medicare's traditional cost-based reimbursement system, there is virtually no evidence on the effects of per case prospective payment using DRGs. Thus, it is not known how DRGs perform relative to other prospective payment systems or whether the State prospective payment approaches can meet the DRG cost-containment standard. Other approaches to the control of hospital costs, including increasing the patients respon- sibility for cost-sharing and limiting providers through contracts between Medicare and hospi- tals, have significant limitations. Patient cost- sharing is not likely to be as effective as desired in altering the patterns of use of hospital technol- ogies because of the patient's relative lack of power and information to make informed deci- sions about the use of technologies in hospitals and the apparently strong preference of the elderly for supplemental medical insurance regardless of its cost. Finally, although contracting may save program dollars, it represents an abandonment of the principle of assuring beneficiaries freedom of choice of providers on which Medicare was built and forces subsidies of hospital care from other payers. ey = Oy PY, a al /. Medicare Physician Payment and Medical Technology As the ancients say wisely, have a care of the main chance and look before you leap, for as you sow, you are like to reap. Samuel Butler Contents Page IntroduCton tr in ee ee ER i 115 Physicians and the Use of Medical Technologies ..................o.. 00. 0 115 Methods of Enhancing Cost Consciousness Among Physicians ................ 118 Physician Education Programs. 0... o... 0 00 oi i oa aii 119 Utilization Review. Programs... ..c....... oii vo irri cain, 119 Second Surgical Opinion Programs... ......cc..ccvvvid ii vv iivinies 121 Changes in Physician Payment ............. cic viinvitn iiss oi inst 122 Changes in the Amount and Methods of Physician Payment ............... 122 Increased Beneficiary Cost-Sharing:...... .... cov iii civ iva dasa di 127 CONCIUSIONS ity tess sete ca ale iro A Ah i ise mde ma a Es a a a 130 7. Medicare Physician Payment and Medical Technology INTRODUCTION Although Medicare's method of payment for inpatient hospital services has recently been altered, its charge-based system of payment for physicians’ services remains (see ch. 2). Medicare expenditures for physician services continue to grow rapidly, however, and changes in payment are currently being considered. The recent changes in Medicare's system of payment for inpatient hos- pital services under Part A have been made pri- marily for the purpose of improving the long-run financial incentives for efficient provision of care. Although one objective of the changes is cost con- tainment, the payment policy does not provide for all savings to accrue to Medicare. Most of the payment changes proposed in the area of physi- cian services under Part B would result in imme- diate savings to Medicare. By changing physicians’ financial incentives to adopt and, especially, use medical technology, however, they would also affect program costs in the future. This chapter examines methods to change the incentives for the use of medical technologies through Medicare payment for physician services. The chapter discusses physicians’ influence on the use of medical technologies and analyzes several methods to enhance cost consciousness among physicians. Then it identifies possible ways that physician payment mechanisms could be used to reduce Part B costs related to the use of medical technologies. 1As noted in ch. 1, “medical services” is often used interchangeably with “medical technology” in this report. PHYSICIANS AND THE USE OF MEDICAL TECHNOLOGIES Physicians are the key determinants of the vol- ume and kinds of medical services and technol- ogies provided to patients. Because they control the decisions made and the resources used in pro- viding medical services, physicians can influence the demand for their services. Thus, it is impor- tant to consider how physicians behave in their adoption? and use of medical technologies and how that behavior is or can be influenced by pay- ment methods and programs in the health care system. It is generally accepted that the charge-based payment system used by Medicare and other third-party payers provides financial incentives 2As noted in ch. 2, there is little empirical evidence related to phy- sicians’ adoption of technologies, although it is known that physi- cians influence hospitals’ adoption decisions. Thus, the emphasis in this chapter is on the use of technologies. to physicians for the use of “technology-intensive” medical care, and these financial incentives exist even within a primary care field such as general internal medicine (295). Furthermore, the influ- ence of financial incentives is supported by em- pirical evidence from “natural experiments”? in which physicians might be expected to respond to the imposition of restraints on payment for the care they offer by providing larger numbers of “technology-intensive” services. One natural experiment involved the Economic Stabilization Program (ESP) put in place between 1972 and 1974 to slow the national rate of infla- tion. ESP imposed both general controls on price increases, including physicians’ fees, and admin- istrative controls on the amount Medicare would 3A “natural experiment” presents changing conditions that can help test a hypothesis. 115 116 ¢ Medical Technology and Costs of the Medicare Program pay for physicians’ services. An examination of the impact of ESP on Medicare payments to Cal- ifornia physicians found that growth in billed charges for individual services was successfully slowed by ESP to half that of the pre-1972 rate, and that when controls were lifted in 1974, the rate of increase more than doubled. During the period that ESP was in place, however, physi- cians’ gross payments from Medicare were largely unchanged. The reason was that physicians were able to increase the number of services they pro- vided and to shift to a relatively higher priced mix of services (153). In the second natural experiment, between 1976 and 1977, all Colorado physicians in similar specialties were grouped together in computing prevailing charges for Medicare reimbursement. Prior to this change, similar specialties had been grouped together, but prevailing charges had been computed for 10 separate regions in the State; urban physicians had had higher prevailing charges than nonurban physicians. After the change to a single statewide prevailing charge for each specialty, urban physicians’ prevailing charges were allowed to increase by less than 5 percent, while nonurban physicians had increases of about 20 percent. Subsequently, urban physi- cians provided more highly intensive services, while nonurban physicians provided less highly intensive services (275). General surgeons who had relative decreases in their reimbursement rates provided greater numbers of surgical services. Declining laboratory reimbursement rates resulted in ordering of more lab tests, and increasing reim- bursement rates resulted in ordering of fewer lab tests. Radiology services were not affected (275). Excessive use of medical technologies is some- times incorporated into medical practice through the habitual behavior of physicians and because the health care system contains few disincentives for such use. These practices continue even though the ordering physician personally may not bene- fit financially. Excessive use of medical technol- ogies occurs even within the norms of medical practice and is evident across the spectrum of tech- nologies available to physicians. Some examples follow. Lengths of stay of patients hospitalized with the same illnesses vary widely across geographic areas, and these differences are explained neither by regional differences in age, sex, or race distri- bution, nor by regional differences in severity of illness (350). Between 1972 and 1977, laboratory tests nearly doubled for both hospital and ambulatory care. The costs of hospital laboratory tests increased from $2.2 billion to over $4 billion, and the num- ber of out-of-hospital tests increased from 850 tests to 1,510 tests per 1,000 physician visits (134). Studies on specific tests have confirmed the in- tuition that much laboratory testing is unrelated to outcome and not used to assist in therapy (232). One study, for example, found a 27-percent in- crease in the number of laboratory tests but no decline in length of stay for patients hospitalized with diabetic ketoacidosis and pulmonary edema (150). Another study found that blood pressure control did not improve with more and costlier laboratory testing (79). Laboratory tests are often ignored by the order- ing physician, even when the tests are abnormal (90,150,427). And physicians often issue multi- ple orders for tests to be repeated, the results of which neither provide additional information nor are used by the ordering physician to change ther- apy (89,209,212,415). Increases in the number of tests and procedures have contributed to changes in clinical practice that have raised costs (212,298,300). Showstack and colleagues (302) have suggested that increased use plateaus over time, as the use of technologies becomes relatively standardized for individual diagnoses. They therefore suggest that methods to limit increases in use may best be focused on use shortly after technologies are introduced be- cause of the ease with which new technologies can become part of accepted practice and the expan- sion of the pool of patients in which certain ther- apies are applied. Physicians’ concerns for their patients’ health can lead to physician-initiated visits. An analy- sis of the magnitude and determinants of physi- cian-initiated visits (286,426) based on a 1977-1978 Ch. 7—Medicare Physician Payment and Medical Technology ¢ 117 Photo credit: National Kidney Foundation, Washington, D.C. The costs of unnecessary and redundant laboratory tests, which have no effect on the type of therapy prescribed or the health outcome of the patient, have contributed to the rise in Medicare expenditures National Medical Care Expenditure Survey found that physicians initiated 39 percent of all visits in the survey period. Physician-initiated visits in- creased with patient age and with poorer health status. Although most of the variations in physi- cian-initiated demand could be attributed to med- ical factors, other factors were also found to influence physician-initiated visits. Greater insur- ance coverage increased the probability of a phy- sician-initiated visit (but surprisingly, did not af- fect patient-initiated visits). Greater physician density (i.e., the physician-to-population ratio) also increased the probability of a physician-ini- tiated visit. The analysis concluded, however, that increasing the physician-to-population ratios should not result in large increases in physician- induced demand. The probability of physician- initiated visits also increased with younger phy- sicians, a finding that was attributed to the higher 25-337 0 - 84 - 9 incomes and more established practices of older physicians, who have less financial incentive to initiate visits. The substantial numbers of new, younger physicians expected to enter practice in the coming years, the analysis concluded, may have only temporary impacts on physician-in- duced demand. Physicians may also alter their medical prac- tices and change their patterns of technology use because they believe patients may sue them for malpractice if harmed. Thus, defensive medicine can lead to patterns of technology use that in- crease costs. Two major types of physician be- havior have been identified as defensive medicine. The most common type is known as positive de- fensive medicine and includes such actions as ordering extra tests and procedures, scheduling more followup visits, using more consultations 118 e Medical Technology and Costs of the Medicare Program and referring more patients to specialists, telling patients more about risks of procedures or tests, using more consent forms, and hospitalizing more patients (254). Less common is negative defensive medicine, which consists mainly of underutiliz- ing technologies, i.e., not performing certain tests and procedures because of possible risks to the patient, thus forestalling legal actions (370). Another behavior that would fit this category would be the refusal to treat certain patients be- cause of the physician's belief that they may be litigious. Identifying or measuring defensive medicine is difficult, in part because individual physician deci- sions may be considered either customary medi- cal practice or defensive medicine, depending on the motives for those decisions. Whether the use of individual tests or procedures constitutes defen- sive medicine depends on how physicians inter- pret their reasons for making specific decisions. Not surprisingly, physicians do not always iden- tify (nor do they try to identify) all the compo- nents in each clinical decision they make. Some physicians use the information on sensitivity and specificity of diagnostic tests, the prevalence of a particular disease, and other objective informa- tion before ordering tests for their patients (101,323). Some physicians consider cost in their decisions. Even physicians who use explicit deci- sionmaking criteria may not be able to identify the subconscious processes that may modify or override their conscious decision processes. The inability of physicians to specify their reasons for clinical decisions in all cases magnifies the prob- lem of determining which decisions should be classified as defensive medicine. Published data on the total cost of defensive medicine in the United States are sparse and poorly documented. Estimates of the annual na- tional bill for defensive medicine range from $1.5 billion to $8 billion (239). None of the published estimates refers to a source for these figures. Evidence of defensive medicine has been based on opinion surveys of physicians, which are nec- essarily subjective. One early survey concluded that defensive medicine is more likely to reduce than to increase use of medical technologies (165). Another survey found that fear of malpractice does influence physicians’ clinical decisions, espe- cially where positive defensive medicine was used. This study also found that defensive medicine is not practiced extensively (94), a finding which seems to conflict with the results of a survey of a random national sample of more than 4,000 practicing physicians (254). In the survey of 4,000 physicians, 35 percent of the surveyed physicians responded to questions regarding 15 specified practice changes. When asked whether concern over legal liability had induced any of these changes, more than 80 percent checked at least one. For example, 48 percent of the physicians who responded to the survey indicated that they ordered more diagnostic tests, citing fear of malpractice as the reason. Thus, the extent and effects of defensive medicine remain speculative. METHODS OF ENHANCING COST CONSCIOUSNESS AMONG PHYSICIANS Several methods of enhancing physicians’ awareness of cost-raising behavior have been used selectively in the past and continue to be suggested by policymakers. These methods include educa- tion programs, utilization review programs, and second surgical opinion programs. Physician edu- cation programs have not been particularly effec- tive in encouraging cost-conscious behavior. Un- fortunately, although cost-conscious behavior may have occurred during and immediately fol- lowing physician participation in education pro- grams, the effect has been quickly lost. Man- datory peer review programs have decreased inappropriate hospital admissions and lengths of stay, but have not been successful in addressing excessive use of tests and procedures. Further- more, the costs of administering these programs may have exceeded the savings that resulted from them. At a more limited level of peer review, second surgical opinion programs have had an Ch. 7—Medicare Physician Payment and Medical Technology * 119 impact on decreasing elective surgical rates, but only when the program has been mandatory, not voluntary. Physician Education Programs In 1978, the Association of American Medical Colleges (AAMC) surveyed 119 U.S. medical schools and found that over one-third of them had or had planned programs for teaching health care cost containment to undergraduate students, res- idents in postgraduate training, or both (175). The 1982-83 AAMC Curriculum Directory, however, lists only 15 medical schools that offered electives in cost containment. Curriculum development for cost-containment education in medical schools has been suggested by various authors. Some authors have suggested a preclinical and clinical curriculum that would attempt to make physicians aware of the economic consequences of their decisions to use medical technologies, while at the same time assuring quality of care (264). Others have suggested that misuse of tests might be decreased through more medical school courses in microbiology, principles of decisionmaking, cost effectiveness, bioethics, and health economics (101,149). Not only im- proved physician education, but better incentives for appropriate use of tests and more selective use of automated technology for tests have also been recommended (149). It is important to stress, how- ever, that although these recommendations ap- pear reasonable, they are not founded on actual evidence of effectiveness. When physicians begin ordering tests and pro- cedures during clinical training, they do not nec- essarily behave according to their knowledge (144, 232). Attitudes and personality traits are impor- tant factors in physicians’ clinical decisions. Phy- sician attitudes toward preventive medicine, for example, may be more important than factual knowledge (67). Several studies have reported a tendency for physicians to underestimate rather than over- estimate costs (93,306). Furthermore, the impact of information about costs on physicians’ behav- ior has been mixed. In one experiment, it was found that providing cost information to physi- cians led to an average reduction in the total cost of tests ordered per patient of almost one-third (78). In another study, two groups of physicians received feedback on costs of X-ray tests and two other groups received information on costs of lab- oratory tests. X-ray use stayed about the same for all four groups, but laboratory use decreased for the groups receiving feedback on lab tests (66). Peer review feedback alone might not ensure re- ductions in test usage (66). Long-range results of peer review feedback have also been questionable. Short-term changes in physician test ordering behavior have been demonstrated but have not persisted over the long term (197). Utilization Review Programs The purpose of utilization review is to ensure that patients receive appropriate medical services that range from specific diagnostic tests to hospital admission. Since physicians admit and discharge patients and provide and order other medical serv- ices, review processes are expected to have an im- pact on their behavior. The most relevant review programs for Medicare have been the Professional Standards Review Organizations (PSROs). PSROs, established under the Social Security Amendments of 1972 (Public Law 92-603), were areawide groupings of practicing physicians re- sponsible for reviewing institutional services pro- vided and paid for by Medicare and Medicaid. The purpose of their review was to help ensure that these services were: 1) medically necessary, 2) of a quality that met locally determined pro- fessional standards, and 3) provided at the most economical level consistent with quality of care (359). Thus, quality assurance and cost contain- ment were the two purposes of the PSRO pro- gram (140,167,313). PSROs were to accomplish these goals by con- ducting the following types of evaluations in in- patient hospital settings, long-term care facilities, and ambulatory care settings: 1) utilization re- view, 2) medical care evaluations, and 3) profile analyses. Utilization reviews were reviews of the necessity of each hospital admission and length of stay. Medical care evaluations were usually audits of patient records to monitor the appro- 120 Medical Technology and Costs of the Medicare Program priateness of tests, drugs, and procedures ad- ministered to patients, according to locally estab- lished criteria. Profile analyses were reviews of patterns of care to identify potential problems. They were intended to assist hospitals in focus- ing the utilization review and medical care evalua- tions on their specific problems. Procedures for the three types of PSRO evalua- tions were developed in the order listed. By 1980, however, most PSROs had worked mainly with review of inpatient care at short-stay hospitals; they had not worked much with care in long-term care facilities or ambulatory care settings. PSROs could penalize hospitals by withholding Medicare payments, but they have not penalized physi- cians directly, other than through occasional peer pressure. The PSRO program was evaluated by the Health Care Financing Administration (HCFA), the General Accounting Office (GAO), and the Congressional Budget Office (CBO). The basic question underlying all three organizations’ studies was whether the PSRO program paid for itself. HCFA reported that utilization review in the Medicare program paid for itself in 1977, but the number of hospital days saved per 1,000 Medi- care beneficiaries by different PSROs ranged from an 8.75-percent decline to a 1.95-percent increase (365). In 1979, HCFA compared geographic areas with and without PSROs and found a 1.7-percent reduction in days of care in the areas with active review (374). Physician cooperation differed among the regions with and without PSROs, so review was probably not the only cause of the difference. Using the same data but different assumptions and measures of effectiveness, both GAO and CBO disagreed with HCFA's findings and found the savings highly questionable (328, 329,334). HCFA maintained that the appropriate measure was the ratio of reimbursement savings to total program cost, which it calculated to be 1.27 to 1. CBO maintained that the proper meas- ure should have been resource savings instead of reimbursement savings, used different estimates of the effect of PSROs on utilization, made dif- ferent assumptions on savings on ancillary serv- ices, and adjusted the per diem rate used in cal- culating reimbursement savings. According to CBO, the resource-savings-to-cost ratio was 0.4 ~ to 1. Even if the ratio of reimbursement savings to total program costs were used, however, CBO'’s ratio (0.9 to 1) was lower than HCFA's because of the different assumptions used (329). Others have criticized using the costs saved per patient day (even when average variable costs per day were used) instead of marginal costs as the measure of success (50,60). Still others have crit- icized cost-benefit analyses that compared total costs of the PSRO program with the marginal ben- efits of PSRO review over claims review by in- termediaries (313). In addition, the use and cost of alternative forms of care for patients with shorter lengths of stay are usually ignored in anal- yses (60). Yet another problem is the fact that quality improvements resulting from PSRO re- view activities were not measured and considered in the evaluations (202), although HCFA did dis- tribute a quality review studies policy to PSROs in 1980 (376). More specific evaluations of individual PSROs and PSRO-type organizations were conducted in many parts of the country (50,98). The studies reported little or no cost savings, and in those studies where cost effects were observed, no con- trol groups had been assigned, so the cause of any cost savings found could not be conclusively at- tributed to the review programs (60,98). In one hospital study, while length of stay and average charges per patient generally decreased following institution of PSRO review, the decrease did not result in overall cost savings to Medicare and Medicaid (425). Following years of budget cuts and uncertainty over their continuation, PSROs were replaced statutorily in 1982 by utilization and quality con- trol peer review organizations (PROs) in the Tax Equity and Fiscal Responsibility Act of 1982 (Pub- lic Law 97-248). The Social Security Amendments of 1983 (Public Law 98-21) provided funding for PROs through Part A of Medicare and made PRO contracts a condition of participation for hospi- tal payment. PSRO areas will be consolidated into PRO regions. The PROs will be contract organi- zations that will set objectives to meet quality assurance and cost-containment goals at the be- ginning of each 2-year contract period and subse- quently be evaluated on the basis of those objec- Ch. 7—Medicare Physician Payment and Medical Technology * 121 tives. PRO functions will include review of the validity of diagnostic information provided by hospitals (Diagnosis Related Group verification); review of the completeness, adequacy, and quality of care provided to inpatients; and review of the appropriateness of hospital admissions and dis- charges. Evaluations of types of utilization review other than that performed by PSROs have been at- tempted. Few studies of ancillary service review focus on the review’s effectiveness, but several studies have shown that ancillary service review may not have the desired effect of reducing the number of tests and procedures used. One of the basic reasons for this result is that it is physi- cians, and not hospital administrators, who admit patients to hospitals, order diagnostic tests and therapeutic procedures for inpatients, and dis- charge patients from hospitals, while the finan- cial incentives or other requirements are often aimed at hospital administrators. There is also wide variation in ordering of ancillary services by physicians (98). Utilization review has also been performed in nonhospital settings, although to a limited extent. Until recently, cost problems for the Medicare program were not as obvious for ambulatory services as for inpatient hospital care, because the Part B program, which covers ambulatory serv- ices, includes patient copayments and deductibles. PSRO demonstration projects in which ambula- tory care was reviewed concentrated on the Med- icaid population, for whom the Federal and State governments pay the entire ambulatory care bill. Reviews of multiple ambulatory care facilities have been inconclusive (252). Second Surgical Opinion Programs The institution of second surgical opinion pro- grams for elective procedures represents another approach to enhancing cost consciousness among physicians and containing costs in both private and public sector health insurance programs. Al- though some second surgical opinion programs maintain that their purpose is to better inform pa- tients so that they can make appropriate decisions, the major purpose of such programs is to reduce the numbers of unnecessary surgical operations. Insurance groups were among the first to be- come interested in the potential cost savings of second surgical opinion programs. Members of several unions in New York City have participated in mandatory and voluntary second opinion pro- grams conducted by Cornell University Medical School/New York Hospital since 1972 (123,218, 219,220,221,288). Several Blue Cross and Blue Shield plans have voluntary second surgical opin- ion programs (225). At least one commercial in- surance company, The Prudential Co., has used a mandatory program (4). In addition, six State Medicaid programs (Massachusetts, Michigan, Missouri, New Jersey, Washington, and Wiscon- sin) use second surgical opinion programs to con- trol costs (281). The Medicare program itself has experimented with voluntary second surgical opinion demonstration projects in Detroit and New York City (127). Although second surgical opinion programs dif- fer in detail, they can be best distinguished by their mandatory or voluntary nature. Mandatory pro- grams generally require patients to consult with a board-certified surgeon, often from a panel des- ignated by the third-party payer, before elective surgery. If patients do not seek second opinions, they may not receive insurance benefits for those procedures (in some programs, however, patients still retain the right to receive insurance benefits for the surgery even if the second opinion con- tradicts the original recommendation for surgery). Consultation fees are usually covered by the payer when the second opinions are mandatory. Man- datory programs have averaged 18 to 20 percent nonconfirmed opinions (225). Voluntary programs make available lists of board-certified surgeons but do not require con- sultation in order for patients to receive reim- bursement for surgical procedures. Sometimes these programs will make the appointments for the patients’ second opinion consultations, and they, too, usually cover the consultant's fees. Few patients (about 2 percent) participate in these voluntary programs. Those who do are a self- selected sample and are not necessarily represent- ative of the group. Not surprisingly, patients in voluntary programs get more nonconfirming sec- ond opinions than do patients in mandatory pro- grams. The national average nonconfirmation rate 122 « Medical Technology and Costs of the Medicare Program in Blue Cross and Blue Shield plans’ voluntary program was 30 percent and one Ohio plan had a 42-percent nonconfirmation rate (225). HCFA has sponsored demonstration projects to test whether Medicare beneficiaries would use a voluntary second opinion program if they did not have to pay any out-of-pocket expenses. Two sites were chosen, and contracts were awarded to Blue Cross and Blue Shield of Michigan and of Greater New York. Both of these projects began in 1978 and ran for 3 years. Participation rates were very low. In New York, 1.2 percent of Medi- care beneficiaries who had surgery participated; in Michigan, 0.3 percent participated. HCFA esti- mated that the maximum reduction in surgery rates was 12 percent, and with the low participa- tion rates, the maximum reduction in the num- ber of surgical procedures would be less than 0.5 percent for Medicare beneficiaries (127). Mandatory second surgical opinion programs have usually been found to have greater cost sav- ings than expenditures. In a study at Cornell/New York Hospital, $2.63 was saved for each $1.00 spent on the program (288). The Massachusetts Medicaid program has been estimated to save $3.90 for each $1.00 spent, including the “sentinel” effect (the change in frequency of physician rec- ommendations for surgery when they are aware that their patients will need a second opinion) (213). Another study of the Massachusetts Med- icaid mandatory program, however, examined one procedure from other areas of Massachusetts and found that “speculative, simple benefit-cost analysis” yielded ratios of 2.27, 1.11 and 0.79, depending on the assumptions made and not in- cluding the sentinel effect (133). Voluntary programs, on the other hand, have high administrative costs relative to the low par- ticipation rates. Even though nonconfirmation rates are higher in voluntary programs, there is little potential for savings because of the low num- bers of patients getting second opinions. The use of second opinion programs, although assumed to decrease surgical rates, might increase ‘them. Researchers in both mandatory and volun- tary programs have indicated that patients tend to follow the second opinion consultant's recom- mendation (127,218). Since confirming second opinions outnumber nonconfirming second opin- ions, some of the patients with confirming opin- ions might decide in favor of surgery they might not have had without the second opinion (48). CHANGES IN PHYSICIAN PAYMENT The use of Medicare's physician payment method to contain Part B costs related to medical technology adoption and use consists of two approaches: ® imposing restraints on the amount and changing the method of payment to physi- cians, and *® requiring Medicare beneficiaries to assume more responsibility for their health care costs, either through increases in patient cost-shar- ing or reductions in the types of benefits covered. Although either approach could result in cost sav- ings for the Medicare program, each would have different effects on the use of medical technologies and on access to medical care by Medicare benefi- ciaries. Changes in the Amount and Methods of Physician Payment Restricting payment to physicians means the adoption of fee schedules or similar restrictions on the level of payment physicians will receive from Medicare. The inflationary nature of the “reasonable charge” criterion by which physicians are reimbursed by Medicare has been dampened somewhat by the imposition of the Medicare Eco- nomic Index, which limits the rate of increase in physicians’ fees to the rate of increase in their costs (see ch. 2). Further controls on physician pay- ment, such as indexing fee increases to the Con- sumer Price Index (CPI) or developing a system of fee schedules, would help to save some pro- gram costs and are discussed below. Such changes, however, would leave three problems un- Ch. 7—Medicare Physician Payment and Medical Technology 123 addressed: 1) as long as Medicare's voluntary assignment policy continues, physicians can at- tempt to recoup their unreimbursed charges from patients; 2) because physicians would still deter- mine the price they will charge (as opposed to what portion of that price Medicare will pay) and because no rational mechanism to set prices for new procedures exists, inflation in medical prices will continue; and 3) even if physician payment is limited, physicians may create demand for their services. Currently, physicians can respond to a reduction in the amount that Medicare will pay by refusing to accept assignment. If a physician refuses assignment, Medicare reimburses the ben- eficiary rather than paying the physician directly, but the physician can still anticipate payment. Limits on Reasonable Charges To slow growth in reasonable charges for phy- sicians’ services, indexes other than the Medicare Economic Index could be used. One of these is the CPI, which is projected to grow at a lower rate than the Medicare Economic Index (330). Rea- sonable charge levels could also be frozen tem- porarily and/or limited to modest yearly percent increases (412). Reasonable charges could also be selectively re- duced. Hourly reimbursement rates in 1978, for example, after standardization for variations in the complexity of different procedures, ranged from $40 for a general practitioner to $200 for sur- gical specialists (174). Thus, allowed charges for surgical procedures might be reduced by a speci- fied percentage (e.g., 10 percent) (330). Fee Schedules Fee schedules are set amounts of payment for particular services. Indeed, indexing the growth in reasonable charges to the Medicare Economic Index is slowly leading to a de facto fee schedule, but the base rate upon which allowable increases are calculated retains the historical specialty and geographic differences in fee levels that have de- veloped under the current system of payment. Fee schedules could be constrained on this historical basis of current fee levels or recalibrated on the basis of costs. Historical fee levels for existing technologies and services could be retained—with controls on future increases—and cost-based fee levels developed for new technologies and serv- ices as they are adopted and disseminated. But putting new procedures on a cost basis and leav- ing old procedures on a fee basis might encourage retention of older, less effective technologies. The adoption of fee schedules to achieve limited cost containment and provide incentives for ap- propriate technology use within the confines of the Part B program could be approached in a number of ways. First, de facto fee schedules could occur by simply freezing current allowed charges, by con- tinuing use of the Medicare Economic Index, by replacing the Medicare Economic Index with a less inflationary index such as the CPI or by impos- ing arbitrary limits (e.g., 5 percent) on the growth of reasonable charges. Second, selected specialties could be targeted for fee schedules, such as by imposing a 10-percent reduction in fees for surgical procedures (330). Alternatively, fee schedules could also be selec- tively or incrementally applied, perhaps starting with inpatient surgical services (330), which com- prise about 25 percent of physician payment under Part B (222). Inpatient surgical services would also be a logical starting point to complement the Diag- nosis Related Group (DRG) system of payment for hospital care under Part A. Another possibility relates to ways to encourage higher assignment rates without increasing costs. A study of changes in Medicare reimbursement rates and their relationship to changes in assign- ment rates has led to the suggestion that increas- ing reimbursement for medical services would lead to an increase in assignment rates, but that de- creasing reimbursement rates for surgical, labora- tory, and radiological services would not lead to significant decreases in assignment rates (276). The existing fee and price system provides financial incentives for the use of “technology-intensive” medical care (295). Still another possibility, there- fore, is to initiate movement toward an overall review of the relative values of all procedures and revision of the fee system, to value “cognitive” services more equally with technology-intensive services. 124 Medical Technology and Costs of the Medicare Program Third, fee schedules could be more broadly ap- plied, for example, in the conversion of physician payment to a parallel DRG-based system. This approach could be limited to inpatient and a few other (e.g., skilled nursing facility) physician serv- ices only, extended to ambulatory care for DRG- related services, or extended to all physicians’ services. In other words, physician payment could be included in the current DRG system for in- patient care, or a fee schedule might be devised that would consist of capitated payments for all services associated with a particular diagnosis. Whatever the approach to fee schedules, admin- istrative changes in the coding system that would have to precede specific changes in methods of payment may of themselves lead to significant cost savings, whether or not they are subsequently used to implement fee schedule changes. The ease with which physicians can use medical technol- ogies is in part a function of the present coding system that is used by Medicare to identify and pay for medical services. That coding system has developed in part in response to the payment pol- icies implemented under the Medicare program. Procedure codes are used to determine Medi- care’s reasonable charges. At Medicare's incep- tion, Medicare providers could use any coding system on their Part B claims and could change codes at any time. In 1979, providers were re- quired to continue with whatever coding system they were then using. Providers were subsequent- ly required to use a standardized coding scheme for Part B, the HCFA Common Procedure Coding System (HCPCS), which all providers are required to use by the end of fiscal year 1985. In HCPCS, coding for physicians’ services is identical to the American Medical Association's Current Proce- dural Terminology, 4th Edition (CPT4). The CPT4 code is augmented with compatible codes for nonphysician services, such as laboratory serv- ices, radiology services, durable medical equip- ment, orthodontic services, chiropractic services, and dental services. Coding for Part B remains completely different from Part A coding, as hos- pitals have been using the International Classifica- tion of Diseases (ICDA, Adapted for use in the United States), which they consider more appro- priate for hospitals (19). In 1966, at the time of the implementation of Medicare, the first Current Procedural Ter- minology (CPT1) contained 2,084 separate pro- cedures. By 1969, the second edition (CPT2) con- tained 3,440, up 65 percent; and by 1977, CPT4 contained 6,132 procedures, an increase of 78 per- cent over the 1969 edition and an increase of almost 200 percent in 10 years (228). The large numbers of procedures from which to choose not only make it easier for physicians to bill for their more expensive services, but also increase the possibility that physicians will in- advertently bill for the wrong procedures. Thus, for example, in an analysis of two types of oper- ations performed on Medicare patients—cholecys- tectomy (removal of the gallbladder) and total hip replacement—it was found that different mem- bers of the surgical team frequently billed for en- tirely different operations. The chief surgeon and the assisting surgeon billed for entirely different operations 19 percent of the time. The surgeon and anesthesiologist disagreed even more, in 40 percent of the gallbladder operations and in 55 percent of the total hip replacements (228). Similar problems have been found with hospi- tal discharge data. In two studies of the reliability of hospital discharge data (236,237), two types of data discrepancies were found: 1) “ordering” discrepancies, reflecting problems in determining which of several diagnoses or procedures should be regarded as the principal one, and 2) coding discrepancies, reflecting errors in assigning a diag- nosis or procedure code number. As a result, the discharge data were reliable in only 66.8 to 77.5 percent of the cases in the first study (236), and in only 59.1 to 64.1 percent of the cases in the second study, which examined data on Medicare beneficiaries (237). In both studies, the correct diagnostic code was a matter of judgment in about 5 percent of the cases. And in the Medicare study, in 70 percent of the cases in which discrepancies were found, the data on principal diagnosis in- cluded in the Medicare claim did not accurately reflect the patient's condition. Several groups of physicians have made at- tempts to create coding systems for common diag- noses encountered in office-based practice. These Ch. 7—Medicare Physician Payment and Medical Technology ¢ 125 include the United States Modification of the Royal College of General Practitioners Clas- sification (567 categories), the Canadian Modifica- tion of the International Classification of Health Problems in Primary Care (ICHPPC) (371 cate- gories), and ICHPPC-2 (362 categories) (294). Thus, it appears that proliferation of codes and their categories satisfies neither physicians nor those concerned with the use of these codes for payment and other purposes. A smaller number of billing packages could be constructed—num- bering in the hundreds, not in the thousands, and perhaps mirroring hospital-based DRGs—which would bear a more reasonable relationship to the services actually provided by physicians. Repackaging of physicians’ services could con- centrate on office visit and special procedure packages. Office visit packages could combine visits and ancillary tests instead of paying for them separately. Present codes could be collapsed for surgical procedures as well as for office visits, and an all-inclusive bill submitted by one physician (e.g., the chief surgeon) could avoid the problem of inconsistent billing by members of the surgical team (228). Special procedure packages, the inpatient ana- log to office visit packages, could be defined nar- rowly to include just specialist services or broadly to include all inpatient and nursing home costs as well. The more narrow packages for special procedures, a more feasible starting point, could be used to group all inpatient physicians’ services, i.e., not only specialist services but also services provided by the patient's primary physician for inpatient care (228). Assignment Physicians at present can decide whether to ac- cept assignment on a claim-by-claim basis. When physicians refuse assignment, they are released from their obligation not to demand the difference between their billed and allowed charges from their Medicare patients. Assignment rates fell from a high of about 60 percent of claims in the early years of Medicare to approximately 50 percent in the mid-1970s and have remained near 50 percent since then (80,118). These rates include claims from beneficiaries re- ceiving both Medicare and Medicaid benefits for whom assignment is mandatory and who com- prise about 10 percent of the total. Thus, the assignment rate for services where physicians have discretion in accepting assignment is about 40 percent. Assignment rates vary greatly by State, from a high of 83.9 percent of services for the aged in Rhode Island, to alow of 19.7 percent in Oregon in 1978 (222). Assignment rates also vary by med- ical specialty. The highest assignment rates (about 60 percent) are in the hospital-based specialties of pathology and radiology (222). This situation may change, however, because pathologists and radiologists are now paid 80 percent of reason- able charges (until 1983, they received 100 percent). As long as Medicare pays for nonassigned care, cost shifting to patients is likely to occur, because physicians can attempt to collect the unreimbursed portion of their bills from their patients. There are several possible methods of penalizing physi- cians for not accepting assignment. One option is for Medicare to institute a mandatory assign- ment policy in which no payment would be made for nonassigned care (228). A mandatory assign- ment policy would solve the problem of cost shift- ing to beneficiaries, but at issue is whether a substantial number of physicians currently treat- ing Medicare patients would refuse to participate under mandatory assignment. In a 1976 national survey, over two-thirds of primary care specialty physicians indicated that they would take no Medicare patients on assign- ment if assignment were mandatory. The great- est influence on the choice was the physician's cur- rent assignment rate—physicians with higher assignment rates were more likely to choose to accept assignment rather than not to participate. If assignment were mandatory, assignment rates were predicted to decline about 10 percent nation- wide, and the total supply of assigned visits to decline by almost 6 percent. Assigned visits were predicted to increase 11 percent for general prac- titioners but to decrease 12 to 25 percent for gen- eral surgeons, internists, and obstetrician-gynecol- ogists. Despite the survey results, the investigators 126 Medical Technology and Costs of the Medicare Program concluded that few physicians could probably af- ford to give up their Medicare clientele totally, despite their threats (228). More recent information supports the idea that physicians would find it difficult to cease treating Medicare patients. The number of physicians who treat Medicare patients is large. In 1982, 87 per- cent of active physicians treated some Medicare patients, and 80 percent of these physicians ac- cepted assignment on some of these patients (1). In a 1981 survey of physicians maintaining office practices, nearly 15 percent of their patient visits were funded by Medicare. Cardiology, internal medicine, urology, gastroenterology and several surgical subspecialties had levels of Medicare- funded visits well above the average. Moreover, while 23 percent of visits to all of the surveyed physicians were by patients 65 years and older, 47 percent of the visits to cardiologists were by patients 65 years and older; for internists, the fig- ure was 41 percent; for urologists, 39 percent; and for gastroenterologists, it was 36 percent (178). Physicians also tend to accept assignment in proportion to the size of their Medicare patients’ bills, the assignment rate rising to as high as 79 percent for bills over $2,500 for internists and 73 percent for general surgeons (222). If faced with mandatory assignment, the potential loss of in- come from their more expensive services may be enough to offset whatever misgivings they might have over accepting assignment for all Medicare claims. Another possible influence on assignment choices is the rapid growth in the number of phy- sicians, largely due to Federal policies beginning about the time Medicare and Medicaid were enacted in the mid-1960's. As a result of these pol- icies, the supply of physicians has grown much faster than the population has. There were 165.5 physicians per 100,000 population in 1974, in- creasing to 193.1/100,000 in 1980, and projected at 213.8/100,000 in 1985 and 231.3/100,000 in 1990 (346). Increases in the number of physicians relative to population growth have been positively correlated with increased assignment rates among general practitioners and internists, perhaps in- dicating that increased competition induced physi- cians to accept assignment on more claims to at- tract patients (276). Medicare is, therefore, already an important contributor to physicians’ incomes and will become even more important to physicians’ prac- tices as the size of the aged population continues to increase. Thus, existing conditions and future trends make it likely that most physicians would continue to treat Medicare patients. One option besides mandatory assignment is for Medicare to pay for nonassigned claims but at lesser rates. Nonassignment could be discour- aged by discounting the physician's charge on nonassigned bills in addition to the 20 percent coinsurance. This could be accomplished either by reducing the allowed charges for nonassigned care or by increasing the coinsurance rate (e.g., to 50 percent) (228). A variation of this option is to allow slower growth in allowed charges for nonassigned care (330). Changing assignment policy is closely linked to changes in physician payment. If further con- trols are placed on the rate of payment to physi- cians while assignment policy is left unchanged, it is likely that the bulk of program savings will be borne by beneficiaries if there is no change in medical technology use. Even though Medicare would reimburse the beneficiary and not pay phy- sicians directly when the physicians refuse assign- ment, physicians could still anticipate payment, subject only to delayed payments and bad debts. Thus, in the absence of a change to mandatory assignment, policies aimed at physicians ultimate- ly would be felt most by Medicare's beneficiaries. Therefore, no matter whether program savings come initially from reduced physician payment levels or increased cost-sharing by beneficiaries, in the absence of mandatory assignment, most Medicare Part B savings will still ultimately come at the expense of beneficiaries. Discussion of Physician Payment Changes Policies that place further limits on Medicare's allowable charges may themselves lead to further distinctions between physicians who treat Medi- care patients and those who do not. The price dif- ferential between what Medicare will pay and what physicians charge their non-Medicare pa- tients may become large enough that many phy- sicians may refuse to treat Medicare patients. The Ch. 7—Medicare Physician Payment and Medical Technology * 127 alternative is that most physicians would continue to treat Medicare patients. However, most pro- gram savings would again ultimately be borne by beneficiaries, because beneficiaries would still be liable for the difference between allowed and billed charges under current assignment policy. When any payment system is changed so that it changes the incentives to provide care, reviews of the appropriateness of the care provided are generally regarded as necessary for quality assur- ance. As noted earlier, most PSROs were never able to progress substantially beyond reviews of hospital admissions and continued stays, nor did they include reviews of ancillary and physicians’ services (109). And review activities have not in- cluded extended-care facilities or ambulatory care. PSROs were statutorily replaced by PROs in 1982 (Public Law 97-248). While it is too early to tell what kinds of review activities these new organi- zations will undertake, it is doubtful that many PROs will begin to review physician and ancil- lary services soon. Given the difficulties in assess- ing what is necessary medical care at the service- by-service level, it is even more doubtful that PRO reviews will be extended any time soon to ambu- latory medical care. Another possible review activity is second surgical opinions. The Inspector General of the Department of Health and Human Services rec- ommended that HCFA initiate legislation on man- datory second surgical opinion programs for Medicare and Medicaid (44). The types of surgery for which second opinions would be mandatory would be specified, and beneficiary copayments and deductibles would be waived for the second surgical consultations. HCFA has responded that these actions would be premature because of limited research on mandatory second surgical opinion programs and because of what HCFA considers questionable cost estimates made by the Inspector General (44). Increased Beneficiary Cost-Sharing Cost shifting to beneficiaries is expected to give beneficiaries increased incentives to be more pru- dent in seeking medical care. The expectation is not only for patients to reduce their use of physi- cians’ services (199,243,299), but also that physi- cians would reduce prices (307,309). Greater selec- tivity by patients in seeking medical care, resulting in fewer visits to physicians, it is hoped, will force physicians to compete by lowering their prices. It is important to note that different methods of cost shifting to Medicare beneficiaries can have different effects. Increasing Part B premiums should have little or no effect on the demand for services, because the increased cost is incurred regardless of whether beneficiaries reduce their de- mand for services. Raising premiums is an insur- ance mechanism for spreading the costs across Part B's entire beneficiary population. Increasing the deductible and coinsurance, on the other hand, may reduce demand, because more costs are incurred by the beneficiary as more services are used. In considering increases in cost-sharing by Medicare beneficiaries, it is important to recog- nize that the elderly already have more out-of- pocket expenses than the young. The results of 1977 and 1978 interviews with 14,000 households indicate that annual out-of-pocket expenses in- creased steadily by age, from $97/year for patients under 6 years of age, to $295/year for patients age 55 to 64, and to $326/year for those over 65 after Medicare's contribution (396). Another important point is that the Medicare population is not homogeneous in its use of med- ical services. About two-thirds of elderly ben- eficiaries use their Part B insurance in a given year (339). But less than one-fifth of beneficiaries ac- count for nearly 90 percent of costs, and these high users of medical services tend to remain high users over time (215). Whether these high users are consuming large amounts of unnecessary serv- ices is not known, but the answer is crucial to the question of whether increased cost-sharing would simply shift costs from the Part B program to ben- eficiaries, or whether it would also lead to reduc- tion of unnecessary technology use. Premiums As noted in chapter 2, beneficiary expenses under Part B are premiums, deductibles, and co- insurance. At the outset of the Medicare program, premiums financed half of Part B program costs. By 1978, following amendments limiting Part B 128 » Medical Technology and Costs of the Medicare Program premium increases to no more than the percent- age increase in social security cash benefits, how- ever, the percentage contribution of premiums to Part B costs had dropped below 25 percent (134). Under the Omnibus Budget Reconciliation Act of 1981 (Public Law 97-35), the limitation on Part B premium increases was suspended for a 1-year period between July 1, 1983, and July 1, 1984. During this period, premiums were to be increased so that they met 25 percent of Part B costs. How- ever, the Social Security Amendments of 1983 (Public Law 98-21) delayed the start of the suspen- sion period to January 1, 1984. Thus, premiums rose to $13.10/month on July 1, 1983, and rose again to $14.60/month on January 1, 1984. The previous method of calculating premiums is ex- pected to resume on January 1, 1985. In its fiscal year 1984 budget proposals, the Reagan Administration recommended increasing premiums until they reached 35 percent of Part B program costs by 1988 (412). CBO identified two similar options: 1) increasing premiums to 30 percent of costs; or 2) limiting the premium in- crease to higher income elderly, with a cutoff point in income at $20,000/year (330). Most of the elderly participants in Medicare Part B have some form of supplemental “Medi- gap” private insurance. In 1976, 63 percent had private insurance, and another 14 percent had Medicaid or some other public supplemental in- surance (199). CBO hypothesizes that additional medical services used as the result of extra first dollar coverage by Medigap policies will cost the Medicare program $3.2 billion in 1984, most of which CBO estimates could be recovered by a 30- percent premium tax on any Medigap policy that pays any part of the first $1,000 of Medicare cost- sharing (330). The Social Security Advisory Council to the Secretary of the Department of Health and Human Services has recommended that Part B premiums be doubled and the extra revenues be used to finance changes in Medicare's Hospital In- surance program. Under the Social Security Advi- sory Council's proposal, Medicare patients would have no limit on the number of days of hospital care each year after paying an initial deductible of $350 for the first day, plus a second deducti- ble of $350 if additional hospitalizations occur.* Out-of-pocket costs for Part B would be limited to $200 per year, a limit which would decrease the demand for commercial Medigap policies, be- cause premiums for these policies cost from $300 to $800 per year (372). For protection against the $200 deductible and for services not covered by Medicare, however, some demand for supplemen- tal insurance should continue. Part B premiums could be varied to reflect local costs. The reasoning follows analyses of plans for implementing a voucher system.* It has been pro- posed in those plans that the value of the voucher follow average reimbursements at the county or State level. Under such proposals, differences in the value of the voucher are meant to reduce the intermarket subsidies that result from large varia- tions in per capita expenditures for hospital care and reimbursement for Medicare and private in- surance enrollees. Otherwise, in some low cost markets, the value of the voucher would be more than current per capita reimbursement rates, and total costs would rise. If the price of health in- surance were adjusted to reflect local market con- ditions, more competition might result (424). Following that theory, Part B premiums could be similarly adjusted, with higher premiums for areas with higher program costs. For example, be- cause Part B participants already are subsidized by general revenues (such revenues cover 75 per- cent of current Part B costs), the lowest premiums could be set at the current rate. While this change would relate premium costs more closely with local medical costs, it would not be expected to lead to more competition. Beneficiaries would still have only one choice of an insurance plan and would not have the opportunity to shop among competing insurance plans as would be the case under a voucher system. Deductibles The current deductible, the amount Medicare beneficiaries pay before Part B insurance is ac- “The purpose in requiring the second deductible is to moderate the financial incentive under DRG hospital payment to increase ad- missions. See ch. 6 for more information. *A voucher is an entitlement of fixed value for the purchase of health insurance. For a discussion of vouchers, see ch. 8. Ch. 7—Medicare Physician Payment and Medical Technology * 129 tivated, is $75 per year. The deductible could be indexed to the Medicare Economic Index so that the deductible would increase in pace with medi- cal costs (412). It could, of course, be set at any other level. Unlike the premium, the deductible is related to initial use of services. In Medicare's elderly population, however, many people have chronic diseases and require continued access to care. For such a population, the deductible may often not be a deterrent to seek care, but a financial hurdle to overcome with each new calendar year. Thus, increasing the Part B deductible would shift costs from the Medicare program to its beneficiaries and might not result in any significant reduction in the demand for medical technologies. Coinsurance Coinsurance raises financial barriers to the use of medical services each time such services are sought. Thus, increasing Part B coinsurance re- quirements might reduce demand, because costs would be related to the amount of services used. However, Part B beneficiaries already have a co- insurance requirement of 20 percent, and for non- assigned care, beneficiaries incur additional liabil- ity for the difference between allowed and billed charges. Thus, one issue is the amount of addi- tional coinsurance requirements that might be rea- sonably imposed, and whether those additional beneficiary costs would be accompanied by a sig- nificant reduction in demand. In the past, suggestions to change Medicare's coinsurance requirements have been focused on Part A, where coinsurance is currently activated only after 60 days of hospitalization for a par- ticular illness. Together with the coinsurance of 20 percent of allowed charges under Part B, im- posing hospital coinsurance raises concern that out-of-pocket expenses could be substantial for hospitalized beneficiaries. Thus, for example, one CBO proposal to impose hospital coinsurance for the next 29 days after the day of admission (on which there is a deductible equal to the average cost of 1 day's hospitalization) also includes cov- erage of all charges after the first 30 days and a cap of $2,000 on total out-of-pocket costs for both Parts A and B for beneficiaries with annual in- comes below $20,000 (330). Discussion of Beneficiary Cost-Sharing Although a significant amount of cost-sharing is possible through changes in the Part B program, the likelihood of cost containment from such changes is limited at best. If significant cost con- tainment occurs, it may come primarily at the ex- pense of reduced access to medical care for Medi- care’s beneficiaries. And while reduced access includes a reduction in the provision of excessive technology use, it also implies a reduction in the provision of necessary and appropriate medical care. Of the three types of direct financial liability incurred by beneficiaries participating in Part B— premiums, deductibles, and coinsurance (a fourth liability is the difference in billed and allowed charges for nonassigned care) —changes in premi- ums have the greatest potential for reducing Medi- care program costs. The extra revenues from the Social Security Advisory Council's proposal to double Part B premiums were projected to allow unlimited hospitalization after an initial deducti- ble for each hospitalization, as well as to limit out- of-pocket costs in Part B to $200 per year. The $200 per year limit on out-of-pocket costs, in turn, was expected to make commercial Medigap pol- icies less attractive (372). Premiums, however, are insurance mechanisms for spreading the risk of the costs of illness, but are not particularly rele- vant in affecting behavior related to medical tech- nology use. This observation is particularly appli- cable with respect to the Part B program, where premiums purchase a single insurance policy — i.e., beneficiaries have no choice in the types of coverage they might purchase (as they would in the case of a voucher system). Regardless of their prudence or excess in seeking medical services, beneficiaries receive no feedback in terms of the medical costs they incur. Furthermore, premiums are deducted from Medicare beneficiaries’ Social Security checks before the checks are issued, so beneficiaries’ awareness of the relationship be- tween premium levels and the costs of medical care is even less than it might be otherwise. CBO hypothesized that, of all the Medicare- specific proposals to contain program costs, a tax on premiums for supplemental Medigap insurance would result in the greatest savings (330). Cur- rently, all persons who purchase health insurance 130 ¢ Medical Technology and Costs of the Medicare Program enjoy a tax subsidy, and elimination of that sub- sidy is a crucial element of proposals to increase competition in health care by making consumers more cost conscious (355). Despite the economic rationale, however, proposals that increase cost- sharing for Medicare beneficiaries and that also penalize them alone for seeking additional insur- ance against this greater cost-sharing by taxing their supplemental policies would be widely regarded as discriminatory. Furthermore, the elderly population has greater medical needs and more out-of-pocket expenses than the younger population (396). Thus, a tax on supplemental in- surance, while leaving the tax subsidy on general health insurance intact, not only would be dis- criminatory but also would be imposed on a class of persons who have the greatest need for medi- cal care. Finally, while a tax on supplemental insurance has been projected as resulting in the greatest sav- ings of the proposals identified (330), the projected savings are also the most hypothetical. In addi- tion to premium tax receipts, the bulk of the sav- ings are expected from a reduction in the use of Medicare-reimbursed medical services by benefici- aries who drop their Medigap coverage. Although such a reduction in use can be expected, quanti- tying the effect on the Medicare-reimbursed por- tion of these services is a tenuous exercise. Deductibles and coinsurance are methods with more promise than premiums for changing phy- sician and patient behavior and thus containing long-term Medicare costs, but even changes in these areas have limited applicability in the Part B program. CONCLUSIONS Strategies for changing the incentives for phy- sicians to provide medical technologies under Medicare's Part B have substantial limitations. These limitations fall into two categories. First, short-term Medicare program savings can be achieved, but these savings are likely to come primarily at the expense of beneficiaries and would not necessarily reduce excessive care. If A large deductible may be more appropriate for hospitalization than for ambulatory care, particu- larly for the Medicare population. Because of the higher prevalence of chronic conditions among the elderly than among the younger population, maintenance therapy for elderly people is more essential. Therefore, increasing the Part B deduct- ible would have a predominantly cost-shifting ef- fect without a proportional decrease in demand for services. Coinsurance has the effect of raising financial barriers each time medical services are sought. Thus, it would have a more significant effect on the demand for ambulatory care than a deducti- ble would. However, Medicare beneficiaries already have a 20-percent coinsurance liability under Part B, and for nonassigned care, the co- insurance is higher. Together with other out-of- pocket expenses, less than half of the health ex- penditures by the average elderly Medicare ben- eficiary was covered by Medicare Parts A and B in 1978 (203). Substantial increases in beneficiary cost-sharing would likely come at the expense of reduced ac- cess for Medicare beneficiaries. While reduced ac- cess would be less likely or of less magnitude with incremental increases, the effect would probably be a simple shift of costs from the program to ben- eficiaries. This effect would occur because mar- ginal changes in the Part B program cannot be expected to substantially alter the limited incen- tives for cost containment that are inherent in the Part B program. payment limitations are imposed on physicians’ services, beneficiaries would be affected either in- directly through cost shifting from physicians to beneficiaries or directly through increased cost- sharing. A substantial increase in cost-sharing through increased coinsurance might reduce de- mand significantly, but two factors argue against this approach: 1) there is already a coinsurance of 20 percent; and 2) further significant increases Ch. 7—Medicare Physician Payment and Medical Technology ® 131 (e.g., up to 50 percent) might reduce beneficiaries’ access to necessary medical care. Even if these Medicare-specific approaches to containing Part B costs resulted in an acceptable allocation of costs between beneficiaries and phy- sicians, the second limitation is that the gap be- tween fees paid through Medicare and those col- lected from non-Medicare patients would continue to widen. In 1978, the percent reduction of the average total billed charge per service used to cal- culate the average Medicare-allowed charge was 20.3 percent, up from 18.7 percent just 3 years earlier in 1975 (222). Further increases in the dif- ference between billed and allowed charges, which could be expected if additional restraints were placed on physician fees under Medicare, again raises the issue of reduced access for Medicare beneficiaries. Equalizing fees for all patients would solve the access problem and could be accom- plished either by raising public fees or by impos- ing a fee schedule on all physicians’ services, pub- lic and private. Raising public fees would lead to higher charges, but fee controls alone, under the current method of billing based on existing diag- nostic and procedure codes, may not be enough to control expenditures for medical technologies or to provide incentives for the provision of the most cost-effective technologies. Thus, there is still a need for methods other than changes in pay- ment, such as utilization review to monitor and evaluate the quantity, mix, and quality of the medical technologies provided (170), or more focused reviews, such as with second surgical opinion programs. rg i 8. Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use When society requires to be rebuilt, there is no use in attempting to rebuild it on the old plan. —dJohn Stuart Mill Contents Page Infroduchion.,....v. LL rh aE a Ta ave 135 Mechanisms To Foster Competitive Behavior by Providers ................... 136 Alternative Sites-of Health Care Delivery .............v.... 0 iil 136 Alternative Organizations for Health Care Delivery ........¢.......... .... 139 Administrative Changes in Medicare \...... 0. 0.0.0. 0 i iipiinsa 141 Mereing PartsiAand B-...... 00. Ld or did iii sane 141 Vouchers. co... ow... vais 8 la a oy aE en i hn Ta oR 142 DISCUSSION Cahn ii ee er bla vs Nn i am 143 8. Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use INTRODUCTION The strategies in previous chapters have focused on changes in Medicare payment methods— changes in coverage policy for individual technol- ogies and changes in hospital and physician pay- ment. This chapter explores policy mechanisms other than payment that could be instituted by Medicare in order to foster the appropriate adop- tion and use of medical technologies and ultimate- ly save costs. Such mechanisms include changes that must involve the general health care system but that Medicare could encourage or embrace and changes in the structure of the Medicare pro- gram itself. For discussion in this chapter, the mechanisms are divided into two broad catego- ries: 1) methods to foster competitive behavior by providers, and 2) administrative changes in Medicare. It is generally believed that costs to the Medi- care program and to the health care system in gen- eral can be contained by the rational adoption and use of medical technologies, which includes using them in appropriate settings. An important meth- od of stimulating the rational adoption and use of technology is to foster competitive behavior by health care providers. In most cases, it is through policies encouraging the use of the alter- native sites and organizations for health care de- livery that competitive behavior is expected to occur. This chapter presents an overview of some of the most prominent mechanisms to increase com- petitive behavior by providers. Alternatives to traditional fee-for-service, solo physician office practices and traditional inpatient hospital care include site alternatives such as freestanding am- bulatory surgery centers, emergency care centers, hospices, hospital outpatient departments, home health care, and nursing homes. They also include organizational alternatives such as health main- tenance organizations (HMOs), the use of primary care gatekeepers, and preferred provider organiza- tions (PPOs). Organizational and site alternatives are not precisely distinct entities, but they are separated here for the purpose of discussion. This chapter defines several alternatives, discusses available evidence on their cost and quality of care, and discusses Medicare's past experience or possible future involvement with them. The chap- ter does not provide an exhaustive list of alter- natives. More descriptive information on the selected alternatives is presented in appendix D. Alternatives to traditional modes and sites of care have been initiated and developed in response to a variety of factors. In many instances, such as the development of freestanding ambulatory surgery centers and HMOs, factors for change have included technological advances, perceived patient need, and potential financial reward for the entrepreneurs. Changes in the health care de- livery system have been influenced by, and had an effect on, the development and use of technol- ogies. As noted in chapter 2, some changes in tech- nology and in the health care delivery system have also been influenced by Medicare. It is the optimal blend of Medicare's adoption of pre-established health care system innovations and of Medicare's fostering innovations with which much of the dis- cussion in this chapter is concerned. Changes in the structure of the Medicare pro- gram itself are also examined in this chapter. These changes (vouchers, in particular) overlap with mechanisms to increase competition among providers but are presented separately for ease of discussion. 135 136 ¢ Medical Technology and Costs of the Medicare Program MECHANISMS TO FOSTER COMPETITIVE BEHAVIOR BY PROVIDERS In an October 1982 report (355), OTA noted that strategies to increase competition in health care generally fall into three major categories: 1) increased cost-sharing by patients when they use medical care; 2) greater competition among comprehensive care organizations (i.e., organiza- tions that provide health insurance and deliver medical care); and 3) increased antitrust activi- ties by the Federal Government. The first strategy, increasing cost-sharing when patients use medi- cal care, relies on the cost consciousness of pa- tients to deter their initiation of care and to temper their own use of technologies as well as use gen- erated by providers (355). Beneficiary cost-sharing proposals were discussed in chapters 6 and 7. The focus here is on the development of alternatives to traditional organizations and sites of health care delivery. Antitrust activities are excluded, because they are beyond the scope of this report. Proponents of greater competition in health care believe that the changes they propose have the best chance of moderating medical care use and costs in the near future. Proponents of com- petition among health care plans emphasize the importance of creating incentives for providers to perform efficiently. They point to the largely untapped potential to use medical technologies more judiciously and to hospitalize less fre- quently, and would rely on alternative delivery systems to rationalize technology use and to achieve lower medical expenditures (355). Under competition theory, providers’ behavior is expected to change as patients become more conscious of cost in deciding whether or not to use services and shop for services on the basis of cost and quality. Physicians are expected to con- tinue to guide patients as they do under traditional fee-for-service payment incentives, but their ad- vice is expected to reflect to a greater extent con- cerns about the effect on their patients’ finances. And hospital administrators are expected to be- come more conscious of costs in the management of their institutions (116). Because “the sick or worried patient is in a poor position to make an economic analysis of treat- ment alternatives” (102), some proponents of competition conclude that the appropriate point for a rational economic choice is the annual selec- tion of a health insurance plan. Attempts to in- crease competition among comprehensive care organizations, therefore, would place the critical choices by consumers at the point of insurance coverage rather than at the point of use of services. Physicians strongly influence the adoption and use of medical technologies. For example, they may purchase sophisticated diagnostic equipment for their office use, or they may persuade hospi- tal administrators and boards of trustees to pur- chase it. In their decisions, hospital administrators or boards may consider the importance of indi- vidual physicians in admitting patients and the various specialties competing for the technologies, as well as the cost of the new equipment and its benefits to patients. They may also consider the extent to which the physicians use the technol- ogies already available. To increase competitive behavior by providers, Medicare could be used to encourage further de- velopment of alternative delivery methods. Alter- natively, Medicare could be restructured to em- brace alternative methods of delivery (instead of providing exceptions for demonstration and eval- uation of alternative methods), with the expec- tation that the health system already has the capacity to provide the preferred modes of deliv- ery. Conclusions as to which is the preferred ap- proach will depend on the answers to two closely related questions: 1) what is Medicare's leverage in promoting or requiring that alternative deliv- ery methods be substantially available? and 2) what is the capacity of the health care system to provide these alternatives? Their answers will af- fect both the substance and pace of change in moving Medicare to a competitive system of in- surance. Alternative Sites of Health Care Delivery Patients can obtain different types of medical care in a variety of locations. Examples of alter- natives to inpatient hospital settings are ambula- Ch. 8—Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use ¢ 137 tory surgery centers, home health care, nursing homes, and hospices. Examples of alternatives to physicians’ offices for the delivery of primary care are hospital outpatient centers and emergency care centers. Each of these alternative sites is discussed further below. Ambulatory Surgery Centers Units to accommodate ambulatory surgery were developed in the early 1970's in response to overcrowded operating room schedules and in- convenience to patients and physicians (125). These units could not have been established with- out the technological improvement of fast-acting anesthesia and the development of the practice of encouraging patients to walk soon after surgery (125). Some of the units are affiliated with hos- pitals and are located either in the hospitals or at other sites. Others are not associated with hos- pitals and are known as freestanding ambulatory surgery centers. The latter are often physician- owned. Surgical procedures that are appropriate to these centers are procedures using general anes- thesia but requiring only a few hours of post- operative monitoring of the patient. Patients are carefully screened. In recent years, third-party payers have accepted claims for surgery per- formed in these centers, and some now require that certain procedures be done on an ambulatory basis for coverage. In 1982, Medicare began to pay 100 percent of a fixed fee for the facility and a surgeon's fee (if the physician accepts assignment) for 100 specific surgical procedures if they were performed in free- standing ambulatory surgery centers (108). His- torically, Medicare had paid 100 percent of costs (after the deductible and copayments) for in- patient, hospital-based tests and procedures including surgery through Part A but only 80 per- cent of an allowable charge for outpatient tech- nologies through Part B. The 1982 change represented an overt attempt by Congress to en- courage Medicare patients and their physicians to use the less costly ambulatory surgery centers when the quality of care they provided was at least as good as, if not better than, that provided in a hospital inpatient setting. The effect the Medi- care coverage change has had on the use of in- patient or ambulatory surgery in the first year is thought to be minimal (177). Medicare's prospective hospital payment sys- tem based on Diagnosis Related Groups (DRGs) provides conflicting incentives for the use of ambulatory surgery. On the one hand, the first criterion for categorizing patients into DRGs is the presence or absence of a surgical procedure, and the strongest financial incentive of DRG pay- ment is to increase hospital admissions. On the other hand, hospitals may try to shift some of their patient care to their own outpatient depart- ments, including ambulatory surgery depart- ments, where the DRG system is not in effect. Home Health Care Another care setting to which Medicare patients may be discharged from hospitals is their homes, where home health agencies can provide care. The specific aspects of home health care have changed over time. Currently, the basic services include part-time or intermittent nursing care by or under the supervision of a registered nurse; physical, oc- cupational, or speech therapy; medical social serv- ices; and part-time or intermittent services from a home health aide. Certain medical technologies (e.g., intravenous antibiotic therapy) that used to be administered only on an inpatient basis are now part of home health care (248). Home health services are usually provided by independent pub- lic or private home health agencies but can also be provided as an outreach service by hospitals. Continued growth in the home health indus- try is expected in response to the financial incen- tives for shorter hospital stays provided by the DRG payment system. The number of agencies providing home health care services has greatly increased since 1966 when Medicare began cover- ing skilled nursing care and physical and speech therapy to homebound elderly people (207). The purposes of providing those services was to lower the hospital length of stay for acutely ill patients, thus cutting costs to the program. Studies of home health care in the 1970's seemed to indicate that home care made early discharges from hospitals possible. Recent studies have ex- amined overall nursing home and hospital use and 138 Medical Technology and Costs of the Medicare Program found no reductions in length of hospital stay, although patient satisfaction and life expectancy were improved. These studies did not address re- admission rates or length of stay (333). Thus, the long-term effect of the early discharges on the sub- stitution of home care for hospital care is not evident. Nursing Homes The Health Care Financing Administration (HCFA) has categorized nursing homes as either skilled nursing facilities (SNFs) or intermediate care facilities (ICFs). More types of medical tech- nologies for treating inpatients and more inten- sive levels of care are available in an SNF than in ICFs. Medicare covers 100 days of care in an SNF following an acute episode of illness, but it does not cover care in ICFs. Not all SNFs accept Medicare patients and pay- ment. This is partly because of some financial risk posed by the possibility that Medicare intermedi- aries (contractors who administer Part A) may deny payment and partly because Medicare pa- tients may require more intensive nursing care and more use of medical technologies than other SNF patients (106). There are shortages of SNF beds for Medicare patients in parts of the country (106). As an alternative to inpatient hospital care, care in SNFs is less costly. The intensity of care dif- fers between hospitals and SNFs, so their patient populations also differ. SNF patients have usu- ally spent time in a hospital before being admit- ted to a nursing home. Under the old cost-based hospital reimbursement system, Medicare paid as much as four times the necessary cost of care for patients waiting in hospitals for SNF beds (106). Under the DRG payment system, if a patient re- mains in the hospital because SNF-level care is unavailable, the days are counted just like other inpatient days. If the length of stay exceeds the DRG average by a specified amount, Medicare will pay the hospitals a per diem outlier rate. In part because DRG payment provides incentives for hospitals to discharge patients to SNFs, Medi- care costs for SNF-level patients in hospitals may decrease. Discharges of sicker patients to SNFs could affect the need for more technologies in the SNFs. It could also increase Medicare's costs for SNFs, which may offset the Medicare hospital savings. Hospices Hospice care has been available to terminally ill patients in this country since 1971. Treatment consists of palliation of the patient's symptoms and psychosocial care from a multidisciplinary team of physicians, nurses, social workers, clergy, psychiatrists and psychologists, dieticians, law- yers, and specially trained volunteers. By allow- ing most patients to remain in their homes rather than the hospital, hospice care saves costs. It also is more pleasant for the patients. Hospice care is a recently enacted Medicare benefit. Legislation to cover hospice care was passed in 1982, while the National Hospice Study, which was to assess the costs and quality of care in a national sample of existing hospices, was still in process. At the time of the benefit addition, preliminary results seemed to show that hospice care would be cost effective for Medicare. The in- centives for hospitals to encourage their patients to use their hospice benefits depend on whether the hospital runs its own hospice and on whether the patient might be an outlier case for whom the hospital might be paid some additional, marginal costs. The effects of the new benefit for hospice care on the quality of life for terminally ill pa- tients are unknown. Also unknown are the costs of the benefit to Medicare and to the beneficiaries. Technology use in hospice care, whether home or hospital based, was significantly lower than use in conventional treatment of terminal cancer pa- tients in the National Hospice Study, while quality of life seemed to be about the same in the different sites (146). Hospital Outpatient Departments Outpatient departments of hospitals may be an alternative site to both inpatient hospital care and private physician office-based care. Many hos- pitals, particularly teaching hospitals, have long had outpatient services, including primary care. Furthermore, in recent years, one of the ways hos- pitals have responded to financial pressures has been to expand hospital services, including pri- mary care in outpatient departments (136). Other Ch. 8—Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use ¢ 139 reasons for the increased use of hospital ambula- tory care are low access to private physicians, particularly for some inner-city residents (2), in- creasing prevalence of chronic diseases (2), greater patient expectations regarding hospitals (2), and advances in medical technology that have allowed movement from inpatient to outpatient settings (136,157,248). Under Medicare's current payment system, hos- pital outpatient departments receive payments through Part B, based on the hospital's costs. Be- cause hospital overhead costs are high, Medicare payments for outpatient department visits are usu- ally higher than physician office charges. How- ever, unlike physicians, hospitals must accept assignment to participate in Medicare, so Medi- care beneficiaries’ costs for outpatient visits may be lower than those for physician office visits. Costs to Medicare are generally higher for out- patient department care than for office care. There is evidence that increasing numbers of patients 65 and older are using outpatient visits for primary care (210). Emergency Care Centers Emergency care centers are alternatives to hos- pital outpatient departments, to some emergency room care, and to primary care in physician of- fices. Such centers are generally equipped with some emergency technologies but do not treat life- or limb-threatening situations, so the name “emer- gency” may be misleading (325). They usually have more diagnostic technologies on location than a physician's office. Emergency care special- ists and some family practitioners have opened emergency care centers to make medical care more accessible to patients who have no primary care physician or who cannot find a physician after hours. The centers have extended hours during evenings and weekends when physicians’ offices are closed, and some are open 24 hours a day, 7 days a week. No appointments are necessary, so care is more convenient for some patients, al- though patients may not experience desired con- tinuity of care. Emergency care centers usually compare their charges with those of hospital emergency rooms rather than with fees for physician office visits. This comparison is not necessarily a good one, because the care provided is often more like of- fice than hospital care. Nonetheless, the National Association of Freestanding Emergency Centers estimates that center charges are 30 to 50 percent lower than hospital emergency rooms for compar- able services (238). If an emergency care center is hospital- affiliated, Medicare will reimburse for visits as though the center were a hospital department. If the center is totally independent, Medicare will pay as though the visit were a physician office visit (55). As noted earlier, hospitals must accept assignment to participate in Medicare, but phy- sicians need not. Thus, if elderly patients were in- formed about which centers were hospital-affili- ated (and thus accepted assignment), they would be more likely to choose them over the independ- ent ones if total prices were comparable. A 1979 study showed that most of the emergency care centers’ revenues came from private insurers or patients who paid directly, with only a small frac- tion coming from Medicaid and even less from Medicare beneficiaries (55). The 1983 followup, although limited in sample size, showed more centers accepting Medicare funds but some centers specifically excluding Medicare cases (248). Alternative Organizations for Health Care Delivery Organizational differences among providers allow patients choices and increase competition in the health care market. Two examples of alter- natives are described below, HMOs and PPOs. Also, primary care gatekeepers are discussed. Health Maintenance Organizations An HMO is a defined set of physicians and other health care providers who provide services for a voluntarily enrolled population that pays a prospective per capita amount. HMOs provide both insurance benefits and comprehensive but specified medical care, and are often cited as the cost-effective mode on which competitive care could be built. In a series of laws, the Federal Government has encouraged the development of HMOs in the be- 140 ¢ Medical Technology and Costs of the Medicare Program lief that health care costs will be contained by this organizational type. Even with the special gov- ernmental treatment, however, HMOs cover only about 5 percent of the U.S. population (392). HMOs cover an even smaller proportion of Medi- care beneficiaries: approximately 2 percent of the Medicare population is enrolled in an HMO-like organization (392). The evidence on why so few Medicare beneficiaries are HMO members is in- conclusive, but some possible reasons include a lack of interest in or knowledge about HMOs on the part of the Medicare population and a lack of interest in enrolling an older, sicker group of people on the part of HMOs. Another reason may be that beneficiaries need incentives to join HMOs, because they would probably have to change physicians and hospitals. Finally, another consideration is that HMOs operate on a per capita payment basis, and as long as Medicare re- quired cost-based data, the HMOs incurred sig- nificant administrative costs. The cost effectiveness of HMOs has been thoroughly studied. The empirical evidence on costs and quality of care for HMOs verifies the predicted behavior: patients are not constrained from seeking necessary care, and, at the same time, physicians are constrained from using un- necessary medical technologies, including tests, procedures, and hospitalizations (206). HMO physicians are usually salaried, and because they often share in the HMO's surplus revenues, they have financial reasons to keep general costs and, specifically, the number of office visits low. HMOs provide incentives to providers to use fewer laboratory and radiological tests on an am- bulatory basis. HMO physicians hospitalize pa- tients less frequently than non-HMO physicians (204). Office visits that have doubtful cost effec- tiveness, such as annual physical examinations for healthy individuals, might be discouraged by HMO physicians, but in many cases, patients ini- tiate such visits themselves. Studies have shown that although HMOs use fewer medical technol- ogies, they do provide care that is at least as good quality as fee-for-service care, and their costs are lower (204,206,429). Primary Care Gatekeepers One of the methods HMOs use to save costs is to have each patient choose a primary care pro- vider who acts as a gatekeeper to specialists and other types of care (316). These primary care gate- keepers not only have the responsibility for re- ferring patients to others, but they also coordi- nate all facets of a patient's medical care. The gatekeeper’s coordinator role is especially impor- tant for elderly patients, who often have multi- ple diseases and must take a variety of drugs that may interact dangerously. The gatekeeper is neither a new concept nor confined to a particular payment method (316). The “traditional” family physician who took care of most medical problems and referred patients to specialists when necessary was an informal gatekeeper. Several years ago, the SAFECO In- surance Co. experimented with a primary care network in Seattle that used physicians as gate- keepers. The experiment showed that the gate- keepers must have financial risks for referring too many patients to specialists, or there are no cost savings to the insurers (277). Great Britain's Na- tional Health Service uses general practitioners as gatekeepers. Their effectiveness is questionable, however, because there is a growing private med- ical care sector for people who do not want to wait for specialists’ care and can afford to pay for it privately. The evidence on technology use by physicians as gatekeepers comes from the literature on HMOs and Great Britain's National Health Service. Physicians in HMOs hospitalize patients less often and use fewer medical technologies in their roles as gatekeepers (204). Evidence of constraints on technology use in Great Britain include long wait- ing periods for elective surgery and an age cutoff for new hemodialysis patients (316). The evidence from Great Britain must be viewed with caution, because there are substantial differences between the health system of Great Britain and that of the United States. Clearly, the gatekeeper performs a rationing function. How this rationing of health care serv- Ch. 8—Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use ¢ 141 ices would be interpreted in light of the freedom of choice enjoyed by Americans in the health care system is unknown. Could the concept be intro- duced in the Medicare program, thereby limiting choice for elderly and disabled patients only? A demonstration project of case management (i.e., gatekeeper) for Medicaid patients in Massa- chusetts was ready for implementation but was canceled by the new governor for political rea- sons. Patient groups could be appropriate targets for experiments with gatekeeper approaches, but such special treatment might result in two-class medicine. Preferred Provider Organizations PPOs are new entities that are designated to combine some of the features of HMOs with those of fee-for-service medicine. A PPO is an agree- ment among providers (usually hospitals and phy- sicians), patients, and insurers that medical care will be delivered at a discounted price as long as the patients use the “preferred providers” (i.e., providers who are among the contractors). Dis- counted prices and utilization review agreements should result in the use of fewer medical technol- ogies by PPOs. Since payment is on a fee-for- service basis in PPOs, however, providers’ finan- cial incentives to use more ancillary services may be as great as in the traditional mode of care. The effect of PPOs on the use of medical technology will depend in part on the effectiveness of the organizations’ utilization review programs. The PPO concept is relatively new, and there are no reliable data on which to base predictions about how PPOs will interact with Medicare or the over- all health care delivery system. Currently, Blue Cross and Blue Shield of Michigan is developing a PPO for Medicare recipients in Detroit under a grant from HCFA (59). ADMINISTRATIVE CHANGES IN MEDICARE The alternatives addressed in this section focus on Medicare's structure. There are definite over- laps between the administrative changes discussed here and the competitive mechanisms presented above. Vouchers, for example, would encourage competition among providers and change Medi- care’s makeup. The separation is to facilitate dis- cussion. Merging Parts A and B It is well known that separate income and pay- ment mechanisms for Parts A and B of Medicare have led to inefficiencies. Some medical technol- ogies have been covered under both Parts A and B, but because of differences in which part paid for their use at which time, there has been unnec- essary duplication of equipment in adjacent fa- cilities. This effect could be avoided if there were one type of coverage and one payment source. Thus, merging Part A and Part B would thwart efforts to shift costs from one part to the other and should decrease the supply of some of the ex- cess technologies. The incentives to open a new freestanding facility for diagnostic tests across the street from a hospital equipped to do those tests would be reduced by the elimination of the cost- shifting possibility. There is no sound fiscal reason for separating Part A—the Hospital Insurance (HI) program — from Part B—the Supplementary Medical Insur- ance (SMI) program—which covers physician services. There is also no sound health reason for the separation. Merging the two parts could allevi- ate the financial problems of the Medicare pro- gram and improve quality of care for patients. Davis and Rowland (84) have proposed that a comprehensive, integrated set of benefits be sub- stituted for those under Parts A and B and be paid from a single trust fund formed from the HI and SMI funds. Everyone eligible for Part A would be covered; no benefits would be optional. According to Davis and Rowland, their sugges- tions for Medicare revenues would guarantee the future solvency of the Medicare program and the availability of medical care for its beneficiaries. Revenues would come from: 1) the current payroll tax contributions to the HI trust fund; 2) general revenues currently projected for SMI expendi- 142 e Medical Technology and Costs of the Medicare Program tures; and 3) new beneficiary premiums that would be related to income and administered through the personal income tax system. The pro- posal assumes that cost-containment efforts and incentives for providers to improve efficiency would continue. The merger proposed by Davis and Rowland focuses on financial solvency through revenue reforms for Medicare, but such a merger would also provide incentives for new organizational in- novations fostering competitive behavior among providers. Medical technology adoption and use would be more directly affected by the efforts fol- lowing the merger to increase provider efficiency than by the actual merger described. It might be possible, for example, to initiate a gatekeeper ap- proach that would penalize the primary physician for inappropriate hospital admissions, thus put- ting the onus on the decisionmaker. The merged Medicare program would be better able to admin- ister such a system than the current separate data systems. Vouchers Vouchers are seen by some policymakers and analysts as an important alternative to change Medicare and contain costs. A voucher system would allow each eligible person a set amount of money with which to purchase medical care and/ or health insurance. In some voucher systems, people who did not spend their entire amount would be able to keep the remainder. Under most systems, though, more benefits would be added to basic coverage to spend up to or over the voucher amount. Any costs of insurance benefits over the voucher amount would be paid by the patient. Medicare vouchers have been proposed in Con- gress several times as a capitation payment meth- od. A Medicare voucher system is attractive for a number of reasons. First, a fixed-dollar subsidy is a capitation method of payment, which would make it easier for HCFA to predetermine and con- trol the program's expenditures. Second, a vouch- er system could substitute for or complement the service-by-service (e.g., ambulatory surgical centers, hospices) and provider-by-provider (e.g., HMOs, PPOs) revisions in current Medicare pol- icy that attempt to fine tune the program in its search for cost-effective alternatives to traditional methods of delivering medical services. Third, such a system would provide Medicare enrollees with a greater choice of insurance plans, in con- trast to the present single Medicare program with its increasing amount of cost-sharing and perhaps decreasing access to physicians of the enrollee’s choice. And fourth, Medicare's enrollees’ ability to enter the general marketplace for medical serv- ices could lead to significant competition for their care and accelerate the development of cost-effec- tive methods of providing medical care. The entry of Medicare enrollees into the medi- cal marketplace through a voucher system raises three questions. First, will a voucher system gen- erate cost savings to Medicare? Second, at what pace and to what extent should the Medicare pro- gram be integrated into general health insurance plans? And third, to what extent could Medicare patients adapt to such a new system? The answers to the latter two questions are pri- marily philosophical. Cost savings to Medicare will depend on the voucher’s initial value and future increases in value. Currently, Medicare pays for hospital and physician services at lower rates than the private sector does. If the value of the voucher is set at average per capita expendi- tures per Medicare enrollee, insurers that enroll Medicare beneficiaries may have to reduce bene- fits or raise premiums to cover their actual costs. Insurers also incur costs that Medicare does not, such as advertising, reserve requirements, pre- mium taxes, as well as profits, all of which would have to be built into the premiums. Voucher proposals link future increases in the voucher’s value to indexes that have increased at lesser rates than medical costs, such as the Con- sumer Price Index. If medical costs continue to increase at a faster rate, premiums would have to be raised above the vouchers value or benefits would have to be curtailed (336). Thus, in order for Medicare enrollees to have the same level of benefits as under the current Medicare insurance program, initial expenditures would have to in- crease, and if medical costs continue to outpace general inflation, future costs would be compar- able to increases in costs under the present pro- gram. If the initial value of vouchers is kept at Ch. 8—Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use ® 143 the level of current expenditures, and if future values increase less than medical costs increase, average benefits would be less or the voucher would not be sufficient to cover premiums for most beneficiaries. Most legislative proposals to date would allow voluntary participation in a voucher system, al- though some of the proposals would allow bene- ficiaries to reenroll in Medicare and others would make the decision to participate permanent. A voluntary system could initially increase per cap- ita Medicare costs if low-cost users selected private plans while high-cost users chose the traditional Medicare option. If benefits for the low-cost users decreased in their plans, they would choose to go back to Medicare's regular benefits. One of the alternatives is a mandatory voucher system if and when more than half of the beneficiaries choose vouchers (336). Under a mandatory voucher sys- tem, which would replace the present Medicare program and in which Medicare beneficiaries would be required to purchase insurance from the marketplace, Medicare program expenditures could be kept the same as current expenditures. Beneficiaries, however, might have to pay more premiums or have their benefits reduced. None of the legislative proposals to date have included mandatory vouchers, though the conditions vary. Voucher problems for both insurance com- panies and Medicare beneficiaries deserve consid- eration. Individual policies would have high administrative costs, and many insurers have in- dicated they would not sell to the Medicare mar- ket (156). Many of Medicare's elderly and disabled DISCUSSION Medicare has fostered certain patterns of care because of its payment policies and program struc- ture. The program has influenced how much and where specific medical technologies are provided. For the most part, Medicare has fostered inpatient hospital care and adoption and use of technol- ogies. The End-Stage Renal Disease (ESRD) pro- gram (see ch. 2) also illustrates the influence of Medicare with the growth of the dialysis center beneficiaries may not understand the differences in insurance policies and may be excluded or dis- advantaged by policy exclusions and preexisting condition costs. Furthermore, beneficiaries’ mobil- ity problems may hamper comparison shopping for the best deal. Technology incentives would depend on the method of payment from the insurance companies to the physicians. If physicians accepted capita- tion payments for Medicare patients, they would have financial incentives to use the fewest possi- ble and least costly medical technologies while still providing good quality care. If they continued to be paid on a fee-for-service basis, their financial incentives would remain much as they are now. It is unlikely that insurance companies would choose the latter path. The choice between mandatory and voluntary voucher systems raises several issues. A volun- tary voucher system would present Medicare ben- eficiaries with the choice of joining other insurance plans on a test basis to see if there would be more cost-effective services available to them than under the present Medicare system. A mandatory voucher system would take advantage of the mar- ket power of Medicare enrollees as incentives for providers to develop more cost-effective ap- proaches, but would place beneficiaries at greater risk for increased cost-sharing and reduced bene- fits if providers failed or were slow to respond. A voucher system might also weaken the influ- ence of Medicare as a large payer for hospital cost control. industry. It is doubtful whether, without Medi- care coverage, the market would have stimulated the research and development for some ESRD technologies such as continuous ambulatory peri- toneal dialysis. Provided that the alternative sites and modes of care are truly believed to have a beneficial impact on costs and on technology adoption and use, can Medicare stimulate their development? 144 e Medical Technology and Costs of the Medicare Program The Medicare program has not been used to its full extent to encourage or discourage alternative sites and organizations of care, in part because of the original political agreement that the Fed- eral Government would not use Medicare and Medicaid to interfere in the practice of medicine. Despite the noninterference policy, however, Medicare was certainly intended to influence, and in fact has influenced, the practice of medicine through the conditions of participation and other quality of care standards provided by the Profes- sional Standards Review Organizations (see ch. 7). The key is that Medicare has influenced medi- cine in a politically acceptable manner, because it has traditionally included the medical profes- sion in its decisions. The separation of Medicare payment for inpatient hospital care (Part A) and for physicians and other types of care (Part B) has had an effect, though not purposeful, on the changes in medicine. As a large single buyer of hospital inpatient care, Medicare has always had a significant im- pact on the availability and use of hospital-based technology. The DRG hospital payment system is explicitly designed to provide incentives for more efficient provision of care, and because of Medicare's size, changes in hospital behavior are predicted to be realized. The actual effects of DRG payment on the adoption and use of medical tech- nologies by the hospital industry and alternative sites and organizations of care will provide pol- icymakers with necessary information for future change. Medicare's leverage for initiating changes in alternative sites or organizations of care is not as great as it is for initiating changes in hospitals. If Medicare used its purchasing power prudently, competition and alternative delivery systems could be either fostered or hindered. It would probably be more appropriate for Medicare pol- icies to be neutral for sites and organizations of care until the evidence on cost effectiveness is more conclusive. For example, the removal of freedom of choice of providers from one segment of the population by requiring case management of Medicare patients may raise ethical questions. Would a voucher system increase freedom of choice of benefits, and would the insurance in- dustry participate in such a program? Whether it is reasonable to expect Medicare to pay for the most expensive, optimal level of health care or to pay less for an adequate level of care is also at issue. If Medicare chooses the latter route, a possible result would be a system of two-class medicine. More targeted Medicare program changes to stimulate the use of alternative cost-effective modes of care may add to Medicare's costs in the short run, until these alternative modes are more firmly established in the health care system. HCFA's efforts to stimulate the use of cost-effec- tive modes of care are reflected in recent changes in the ESRD program (see ch. 2) and in Medicare demonstrations on risk-contracting for HMOs (see app. D). In the ESRD program, the imposition of com- posite reimbursement rates to stimulate home di- alysis care makes Medicare payment rates for home dialysis higher than the actual costs. Home dialysis costs may rise because of the need for home health aides for patients with little family support. If costs do not rise significantly and the use of home dialysis does increase greatly, Medi- care will be paying at a rate much above costs. Whether there will be a savings to the ESRD pro- gram with the redistribution of dialysis patients between home and center dialysis is not clear. Also not clear is whether the large difference be- tween costs and payment levels (estimated at near- ly 50 percent (344) will continue to be justified. HCFA’s HMO risk-contracting demonstration programs involve the same issue—paying more initially to establish alternative sites or modes of care, yet leaving unresolved for future consider- ation how these alternative modalities can be sus- tained at payment levels lower than originally needed to stimulate their participation in Medi- care. Payments in HCFA's risk-contract demon- strations were set at 95 percent of average adjusted per capita costs of providing fee-for-service care to these enrollees in the HMO's service area. Thus, payments in these demonstrations were set above the HMO's cost levels, with the extra payment to be used to induce Medicare beneficiaries to enroll through extra benefits or decreased pre- miums (97). Again, left for future consideration is the payment level to HMOs if and when they Ch. 8—Alternative Approaches to Changing Incentives for Medical Technology Adoption and Use e 145 gain a significant share of the medical market for Medicare enrollees. If the DRG payment system continues to be the only change in Medicare, what effects are pre- dicted for the rest of the system? Care in SNFs, ICFs, hospices, and in the home will probably in- crease in response to the financial incentives to shorten hospital lengths of stay. In some cases, therefore, sicker patients will be treated in these alternative sites. To the extent that the alternative sites have the facilities and staff to give appro- priate care, such treatment might not lower the quality of care. If the patients are so much sicker that their treatment in alternative sites necessitates the hiring of new staff or the purchase and use of new technologies, however, the cost of care in the alternative sites will increase. In addition, the current DRG prices reflect average lengths of stay in hospitals. If patients move to other sites for the final convalescent days of care and the DRG prices do not reflect the change, Medicare would essen- tially be charged twice for the convalescent care. Policymakers concerned with the Medicare pro- gram have shifted their emphasis from making mainstream medical care available to the elderly and disabled to a search for more cost-effective methods of providing care. Detailed, specific changes have been made, which still preserve the basic framework of the Medicare insurance pro- gram and its separate hospital and physician reimbursement parts. The development of alter- native modes of care, and step-by-step revisions in the original Medicare legislation in order to adopt and nurture these alternatives, are gradually transforming Medicare away from a cost- and charge-based retrospective system of payment. The direction of these changes is clearly toward a total, prospectively determined system of pay- ment and toward providers with both financial and service responsibilities. The transition to cost-effective modes of care raises issues that can only be resolved with ex- perience. At least for the short run, alternative methods of care may add to costs—as when home health care supplements instead of replaces hos- pital and nursing home care—or they may require payment levels similar to the fee-for-service sys- tem to build up their presence—as in the case of HMOs serving the elderly. In the case of vouchers, cost-containment objectives have to be balanced against the use of Medicare enrollees to test eco- nomic theory in practice and the probability that at least some of the Medicare program’s cost sav- ings will come at the expense of increased cost- sharing and/or reduced benefits for Medicare enrollees. Medicare's leverage in the health care system is variable and depends on which segment of the system is examined. If the Medicare program is to be changed, is it sufficient for Medicare to try reforms, or must the alternatives exist in the sys- tem so that Medicare can incorporate them? Alter- native sites and organizations for health care de- livery are currently available for experiments or revisions in the Medicare program. The extent to which these alternatives could adapt quickly and adequately to a major change in Medicare policy is being tested by DRG hospital payment. Clearly, Medicare policies can provide small steps, as they have by the special coverage for am- bulatory surgery in freestanding centers and for hospice care. Parallel developments of alternatives in the health care system outside of Medicare will continue. Policymakers should watch these sys- tem developments and modify Medicare policy to take advantage of new cost-effective modes or sites of care when available. Finally, the decision must be made for the Medicare program either to keep and strengthen its purchasing power by continuing to cover beneficiaries in a large pro- gram or distribute its beneficiaries into the mar- ketplace by means of a voucher system. Either of these actions has implications for technology in- novation and diffusion and for cost control. Part Three 9. Findings and Policy Options Nobody can really guarantee the future. The best we can do is size up the chances, calculate the risks involved, estimate our ability to deal with them, and then make our plans with confidence. —Henry Ford Il Contents : Page IntroducHON «Soa a a Sa a ae Ee Te as 151 Coverage of Specific Medical Technologies... .c.... co. iis ivirivisvimsiins 151 Broader Incentives Toward Technology... .... ...... con, a rev J daiiiiiivvh 156 HOSPItAHPAYINGIN «sass nner ir +s oisian sis niv ian wine Wi a iw vlns wid wn eines 156 Physician PAyment .. .... aha ngs ririos snes ssa roams » vas sn bv evvins vais 164 Alternative Approaches to Changing Incentives ........................... 168 9. Findings and Policy Options INTRODUCTION The dual relationship between medical technol- ogy and the Medicare program has been explored in previous chapters: Medicare policies affect the adoption and use of medical technologies, and patterns of use of medical technologies signifi- cantly affect the types and levels of Medicare costs. This chapter identifies several options for changes in Medicare policy that could be used to influence medical technology adoption and use and to restrain Medicare program costs. Medicare policy can influence the adoption and use of medical technologies in order to cut pro- gram costs both through policies directed specif- ically toward individual technologies and through policies that provide broad incentives for rational adoption and appropriate use of medical technol- ogies. This chapter presents findings and policy options by issue area within each of these two broad policy areas. The first section of this chapter focuses on policies directed at individual technol- ogies and contains options pertaining to Medicare coverage policy. The second section focuses on policies that provide broader incentives toward technology and contains options directly concern- ing Medicare hospital and physician payment and approaches encompassing interactions between Parts A and B of Medicare as well as interactions between Medicare and the general health care system. The options identified in this chapter generally fall into the following categories: ® changes in Medicare's coverage policy for specific technologies; ® changes in the methods of Medicare payment to hospitals; ® changes in the methods of Medicare payment to physicians; and ® approaches to changing the incentives for the adoption and use of technology that do not directly involve, but may be related to, Medicare payment mechanisms (e.g., en- couraging the development of alternative cost-effective health care delivery systems). Some of the options involve direct legislative ac- tion. Others are oriented to the actions of the ex- ecutive branch but would involve congressional oversight or encouragement. The order in which the options are presented is not meant to imply priorities among them. Fur- thermore, the options are not, for the most part, mutually exclusive. Adopting one option in any category does not necessarily imply that others are inapplicable within that category or within any other category. More often, a careful com- bination of options can better produce the in- tended effects. In some cases, an option may sug- gest improvements for more than one aspect of Medicare policy. It is important to keep in mind that changes made in one area have repercussions in other areas. COVERAGE OF SPECIFIC MEDICAL TECHNOLOGIES ISSUE: How can Medicare's coverage process for specific technologies be improved? It is generally agreed that third-party coverage policy has influenced decisions about the purchase of some expensive, visible medical technologies. Thus, the coverage of specific technologies is a potential method of containing Medicare costs through control of the diffusion (i.e., adoption and use) of medical technologies. In the present cost-conscious environment, the attention of pol- icymakers may become focused on expensive 151 152 eo Medical Technology and Costs of the Medicare Program technologies, i.e., both technologies with a high capital purchase price and technologies with lower capital purchase prices but high utilization. Long- term effects of other technologies, however, should be considered in Medicare coverage deci- sions. The precise relationship between coverage policy and purchase of other kinds of medical technologies or the use of any technologies re- mains speculative. The benefits for which Medicare will pay are designated in broad general categories such as in- patient services, outpatient services, and physi- cians’ services. With few exceptions, therefore, Medicare coverage policy is determined at the local or national level on a technology- by-technology basis. Coverage decisions by Medicare contractors or by the Health Care Fi- nancing Administration (HCFA) follow the statu- tory mandate which excluded Medicare pay- ment for items and services that are not “reason- able and necessary” for diagnosis, treatment, or improved functioning of a malformed body mem- ber. “Reasonable and necessary” items and serv- ices are interpreted as those that meet four cri- teria: general acceptance as safe and effective; not experimental; medically necessary; and provided according to accepted standards of medical prac- tice in an appropriate setting. Evaluation of the nonmedical effects, for example, economic and social effects, of a technology is usually not part of a technology assessment for coverage purposes. HCFA officials and individual Medicare con- tractors have had considerable latitude in deter- mining which technologies are to be covered for reimbursement. Furthermore, there is consider- able variation in the implementation of national coverage decisions by Medicare contractors. Uniform implementation of HCFA's coverage decisions might foster equal treatment of Medi- care beneficiaries throughout the country.! How- ever, although traditional local variations in the practice of medicine do mean that Medicare ben- eficiaries effectively receive somewhat different benefit packages, there is no evidence that local differences in standards of care affect patients’ health. Uniform implementation of coverage deci- It should be noted, however, that current national coverage policy is considered an interpretive rule and does not have the force of law. sions may discourage local differences, and some observers believe that it may interfere in the prac- tice of medicine. The strength of such interference would depend on the influence coverage policy has on the adoption and use of medical technol- ogy in the first place. The lack of necessary information on which to base assessments and coverage decisions is a seri- ous problem. Although the guidelines used by the Office of Health Technology Assessment (OHTA) to evaluate the safety and efficacy of medical tech- nologies stress the value of basing coverage rec- ommendations to HCFA on data from controlled clinical trials or other well-designed clinical studies, in many cases, such data are unavailable. Furthermore, timeliness of identification of new or outmoded technologies is important to Medi- care, because the assessment process is itself time- consuming. Some technologies, such as heart transplantation, are so expensive and visible that they have been identified, but a coverage deci- sion has been delayed. In the case of the artificial heart, the ostensible reason for the delay is to allow a thorough assessment of the technology's safety, efficacy, and other aspects. Some analysts have noted that the true reason behind the delay is cost containment (22,258). Such delays in cov- erage decisions may save Medicare program costs for a time. Sometimes the coverage decisions are slowed by the backlog of technology assessments faced by OHTA. In other cases, new techniques, such as coronary artery bypass graft (CABG), have been paid for by Medicare under existing procedure codes. By the time payment issues were raised, CABG was considered generally accepted medical practice. A new issue for the Medicare program is the role of coverage policy with respect to the new Diagnosis Related Group (DRG) prospective pay- ment system for hospitals. Although the cover- age process and the DRG-rate adjustment proc- ess share a similar “approval for payment” function, they differ in that a coverage determina- tion focuses on a specific technology, while ad- justing DRG payment rates focuses on the larger entity of a diagnostic group, which includes par- ticular technologies. Moreover, the DRG-rate ad- justment process must include issues of cost as an integral issue, while the coverage process at pres- Ch. 9—Findings and Policy Options ® 153 ent does not consider cost issues. Despite their dif- ferences, the technology assessments performed for the coverage and DRG rate adjustment proc- esses no doubt will have similarities, and their co- ordination should be encouraged. An idea related to the coverage process that may be worth exploring is the use of Medicare conditions of participation? to influence the adop- tion and use of individual medical technologies. A new condition of participation for hospitals, for example, could require that hospitals have technology assessment committees. The process of explicitly discussing safety, efficacy, and cost questions before a hospital purchases, and phy- sicians use, a particular technology may be worth- while. A major finding of OTA’s 1980 assessment of cost-effectiveness analysis (CEA) in health care was that the process of identifying and consider- ing costs and benefits in CEA might be more valu- able in decisionmaking than the rigid and formal application of CEA results in health care program decisions (353). There is no process of identifying and consid- ering costs and benefits in CEA for a hospital that is yet acceptable to all concerned, however, nor is there evidence of effectiveness of such processes in making technology adoption or use more ra- tional. In part, this is because a decision that is rational from a hospital's perspective may not be rational from society's perspective. Furthermore, Medicare conditions of participation are being simplified under the current Administration to re- duce the burden of administering detailed nonstat- utory requirements. Adding a new condition of participation would be counter to that objective. In addition, hospitals may decide to use technol- ogy assessment committees under the DRG pay- ment system even without the requirement be- cause of the complexity of the incentives. Many hospitals already have purchasing departments or committees that perform the function of assess- ment with variable rigor. 2Conditions of participation are requirements that must be met by providers in order to be allowed to receive payment for Medicare patients. An example is the requirement that hospitals conduct utilization review. Option 1: Amend the Medicare law to allow cov- erage for emerging technologies on an interim basis in exchange for data on their safety, effi- cacy, and costs. Implementing this option could improve the availability of data necessary for good coverage decisions and, ultimately, for the rational adop- tion and use of medical technologies. Defining “emerging” technology is difficult. The movement of any technology from the research phases into common clinical practice represents a continuum, and there is no specific point at which a particular device or technique would stop being “experimental.” Developing criteria would be necessary to ensure that interim coverage was not allowed too early in a technology's develop- ment in order to minimize the risks to patients involved in clinical trials. Coverage of emerging technologies in return for data would yield new information on elderly patients—few previous clinical trials have col- lected data on such patients. Involving elderly pa- tients in clinical trials would have positive and negative aspects. A disadvantage would be the in- herent risks to patients involved in clinical trials of emerging technologies, although subjecting Medicare beneficiaries to technologies tested only on younger patients also involves risk. Design- ing clinical trials involving elderly patients is very difficult, because such patients usually have more than one medical condition. Important decisions would have to be made. One such decision would be about how long to provide interim coverage. Time limits would have to be explicit from the beginning, if interim cov- erage were implemented. Another important deci- sion would be where and to whom the interim coverage would apply. Sites and providers to con- duct clinical trials would be selected as part of the research peer review process. Although the selec- tion of specific sites and investigators for interim coverage might be regarded as favoritism, limiting Medicare coverage only to sites and investigators with very specific protocols would seem to be a more prudent approach than paying for the use of emerging technologies on a more widespread basis. Once an emerging technology's safety and 154 e Medical Technology and Costs of the Medicare Program efficacy had been established, coverage could be extended, or, if Option 2 were implemented, other appropriate sites and providers could be identified for permanent coverage. The collection of cost data is a key element of this option. Thus, this option is closely related to Option 3 below. If Medicare payment rates, particularly for Part B services, are ever to be based on technologies’ costs instead of past charges, the Medicare program will need better cost data on which to set these rates. In the past, costs have often been unknown until technologies were already in widespread use. Coverage of emerging technologies in exchange for data would initially add costs to the Medicare program, but the information gained could very well be worth the investment. The availability and interpretation of better data on the safety, ef- ficacy, and especially costs of new technologies would provide a good mechanism for Medicare policymakers to decide how best to expand or contract benefits in a rational manner. Such cov- erage would shorten the usual delay involved in getting data from clinical trials on which to base coverage decisions. Although long delays may save Medicare costs in the short run, such delays may sometimes mean that some patients are denied efficacious technologies. Option 2: Amend the Medicare law to limit cov- erage of complex technologies to their provi- sion in selected sites by selected providers. Certain medical technologies involve highly complex equipment and supplies and require a skilled team of providers. Limiting Medicare cov- erage for specified complex technologies to their provision in particular sites could help control Medicare costs and might also improve quality of care. Theoretically, implementing this option would help control costs by reducing excess capacity. There would be a reduction in unused technol- ogies in a number of sites along with economies of scale produced by having larger capacities at only a few sites. Certain surgical procedures per- formed in high-volume hospitals have better pa- tient outcomes than the same procedures done in low-volume hospitals (124a). And for high vol- umes of complicated surgical procedures, the teams of surgeons, nurses, anesthesiologists, and medical technologists who work together would learn to work together more efficiently and effec- tively. While limiting sites and providers could help control costs by limiting the adoption and use of complex technologies, it might also limit the num- bers of patients who could be treated. If need ex- ceeded capacity, a method for rationing care would be necessary. Currently, private insurance companies cover exceptional technologies in selected sites: Blue Cross will pay for the heart-lung transplantation done at Stanford University. And Medicare itself has set some precedent by limiting payment for therapeutic apheresis to its use in specific settings and by specified providers (26). A potential problem with selective coverage by ‘Medicare is that such coverage could lead to two- class medicine. Nonselected hospitals, for exam- ple, could purchase major medical equipment that might not be covered by Medicare. Physicians in the hospitals might then use the equipment only for their non-Medicare patients. The probability of the occurrence of this situation would depend on the cost of the technologies and the availabil- ity of trained staff, as well as on the proportion of the hospital's patients who are Medicare ben- eficiaries. The specification of providers and sites for cer- tain technologies might be regarded as unequal treatment of providers. Yet in this option, as in the previous one, sites and providers could be selected on a peer reviewed basis to assure quality and to maintain acceptability within the medical profession. Peer review such as that undertaken by the National Institutes of Health (NIH) study groups would be one possibility. The DRGs encompassing the specified technol- ogies would have to be treated differently in selected hospitals (sites other than hospitals are not yet under the DRG payment system). Assum- ing that these complex technologies would be very expensive, Medicare hospital payment would somehow have to support their rational adoption Ch. 9—Findings and Policy Options e 155 and use (see Option 6). Otherwise, all cases using such technologies would be outliers. Option 3: Mandate that Medicare coverage deci- sions include cost considerations when appro- priate. Because cost containment is so crucial in the Medicare program and in the health care system, it may be necessary to explicitly include cost con- siderations in coverage decisions. At present, the adoption and use of medical technologies involves implicit rationing of scarce dollars. In today’s eco- nomically constrained environment, perhaps the tradeoffs among coverage decisions should be more explicit. Especially if Medicare covers high- cost technologies that yield few benefits, other services must be eliminated in order to decrease program expenditures. Because Congress provided little guidance on how it intended the statutory “reasonable and nec- essary” tests to be applied, the question of the ap- propriateness of using cost information in Medi- care coverage decisions has been raised by the U.S. Department of Health and Human Services (DHHS) several times. DHHS has asked its legal counsel to investigate the definition of “reason- able and necessary” in the Medicare law. No clear decision has yet been provided. If quality of medical care for Medicare benefi- ciaries is to be maintained, a method of determin- ing the most cost-effective technologies in health care is highly desirable. CEAs and cost-benefit analyses (CBAs) represent a possible group of methods. One strength of using cost considera- tions in general, or CEA/CBA specifically, in Medicare coverage decisions would be that the im- plicit rationing would become explicit. CEA/CBA methods still need to be enhanced and refined, but the process of analysis itself can help make as- sumptions explicit and can help identify as many costs and benefits as are feasible for considera- tion in decisionmaking in the health field. It is im- portant to note that the availability of cost data is essential for this option. A previous OTA study (353) found the use of formal, well-conducted, sophisticated CEAs or CBAs in decisionmaking in the health field is the exception rather than the rule. Adding it to the Medicare coverage decisionmaking process might be a helpful step. Although on the surface it would appear that technology evaluations including cost criteria would be more effective in allocating resources than those without, the relative effectiveness of the two types of evaluations has yet to be demon- strated. At this time, though, it appears that CEA/CBA can at least aid decisionmaking when used in conjunction with other kinds of informa- tion (353). Option 4: Conduct oversight hearings to improve the Medicare coverage process. Several administrative problems pertaining to the Medicare coverage process have been iden- tified in this report. Problems that need attention include the following: 1. the inadequate identification of emerging and outmoded technologies for coverage decisions; 2. the lack of uniformity in the implementation of national coverage decisions; 3. the timelag involved in the coverage proc- ess, including technology assessment; 4. the complex coding system and proliferation of codes; and 5. the incomplete dissemination of information. These problems all potentially raise Medicare's costs. Their solutions may save some money and reduce the disparity in Medicare beneficiaries’ cov- erage across the country. Numbers 2, 3, and 5, however, could actually decrease Medicare ex- penditures, so a detailed analysis of these prob- lems may be warranted. The timelag, for exam- ple, could save Medicare program costs but have negative consequences on patients’ health. If pa- tients use more services in the long run because of the adverse effects of coverage delays, costs to Medicare could increase. There are several ways of improving the iden- tification process. Contractors have recently reported that they are receiving inquiries concern- ing the coverage status of new technologies from manufacturers and providers. HCFA could mon- itor the Food and Drug Administration's (FDA's) processes to anticipate new medical devices and 156 e Medical Technology and Costs of the Medicare Program NIH's registries where new applications of pro- cedures (e.g., percutaneous transluminal coronary angioplasty) are followed. Similar efforts in the private sector could be scrutinized. Oversight hearings could be used by Congress to focus the attention of DHHS on problems in the coverage process. However, the law provides DHHS with the authority to make, but not to im- plement, coverage decisions. The coverage index appendix, other manual instructions, and in- termediary lessons are usually considered inter- pretive rules, and as such do not have the force of law. DHHS may need additional funds to improve its administration of Medicare coverage. Whether the savings to the program would justify the ex- tra cost of the administrative changes that might be necessary is speculative without further study. BROADER INCENTIVES TOWARD TECHNOLOGY Hospital Payment ISSUE: How can Medicare's hospital payment system incorporate appropriate incentives for generating effective and efficient adoption and use of medical technology? The retrospective, cost-based hospital reim- bursement system under which Medicare operated from 1966 until fiscal year 1983 was significantly altered first by the Tax Equity and Fiscal Respon- sibility Act of 1982 (TEFRA) (Public Law 97-248) and then by the Social Security Amendments of 1983 (Public Law 98-21). The latter mandated the phasing in of a prospective, per case, hospital pay- ment system based on DRGs. Prior to the implementation of Medicare's DRG per case payment system, hospital payment was based on utilization of medical technologies or on days of hospital care. Such payment may provide incentives for the inappropriate use of medical technologies in hospital settings (e.g., requiring chest X-rays upon admission to the hospital and admitting patients to the hospital on Friday before Monday surgery). Per case payment is an im- provement over per diem payment but still pro- vides rewards for certain types of inefficient behavior (e.g., unnecessarily increasing admis- sions). Another possible method of hospital pay- ment is capitation payment. Unlike per case payment, capitation payment offers incentives to decrease admissions. Paying by capitation would require data systems other than those currently available. Although it is too early to evaluate the effects of DRG hospital payment, Medicare's new hos- pital payment system gives financial incentives to hospitals to increase admissions and reduce lengths of stay. Under DRG payment, therefore, some patients may be admitted to the hospital un- necessarily, others may be discharged too early, and some may not get all their elective care in one hospital stay. In addition, under DRG payment, hospitals have a financial incentive to decrease length of stay as a way of both opening beds for new admissions and decreasing the hospital's cost per stay. To the extent that a hospital's occupancy rates are low, the incentive to shorten length of stay is weakened, because the hospital will have difficulty in filling the opened beds and covering its fixed costs. Medicare's DRG hospital payment system also provides incentives for hospitals to reduce the number and cost of ancillary services. Prior to the implementation of the DRG system, hospital ad- ministrators had financial reasons to encourage physicians to use available technologies. Under DRG payment, hospital administrators are likely to discourage physicians from using many high- cost technologies, particularly diagnostic tests that add only marginal information and may not change the course of treatment. In some cases, the substitution of low-cost technologies for high-cost technologies may result in a decline in quality of care. Thus, quality of care remains an important issue under DRG payment. Ch. 9—Findings and Policy Options ® 157 DRG prospective payment has changed the in- centives provided under cost-based reimburse- ment for the adoption and use of medical tech- nologies in hospitals. In general, hospitals now have greater incentives to adopt new medical tech- nologies that reduce their cests and lower incen- tives to adopt technologies that increase their costs, even when the latter are worth their added costs to society. But the incentives for the adop- tion of medical technology depend on the way capital costs are treated. Thus far, capital costs are not covered by the DRG hospital payment sys- tem and are treated as pass-throughs. Treatment of capital costs as pass-throughs does not alter the direction of the incentives governing technology adoption under DRG payment as long as the ef- fect of a new technology on total cost per case is in the same direction as its effect on operating costs. However, for certain medical technologies, namely, those which reduce operating costs but raise total costs per case, capital cost pass-through reverses the incentives for adoption inherent in DRG payment. Congress has recognized that cap- ital costs are still a problem for Medicare, and the law requires DHHS to study how capital costs should be paid in connection with the DRG hos- pital payment system. Innovations in medical devices, drugs, and medical techniques that improve the quality of care for the Medicare population but also increase hospital per case costs may not be as readily adopted under DRG payment as they were before. Quality is a difficult concept to define and its measurement is equally complex. Technology assessments may offer some assistance in com- parisons of the quality of care afforded by dif- ferent technologies. Comparisons of different measures of quality are important for national policymakers who must decide whether a particu- lar quality-raising technology is worth its cost to society and, thus, whether it should be adopted. Congress has provided some control over qual- ity of care by mandating the utilization and quality control peer review organizations (PROs). Hospitals must have signed agreements with these PROs in order to receive Medicare payments. Funding for the PRO reviews will come from the Part A Medicare trust funds. The PROs will be evaluated on the basis of their individual contracts with HCFA. One of the responsibilities of the PROs will be to monitor the potential admission/ discharge/readmission problem. The development of DRGs originally was not related to payment, but DRGs were refined once they were used for payment purposes. Although Congress has adopted DRGs for the Medicare hos- pital payment system, improvements of DRGs and of the payment system are still needed. Such refinements are anticipated in light of the series of congressionally mandated studies and the Pro- spective Payment Assessment Commission's re- sponsibility for recommendations on changes in DRG relative weights and categories. The success of the DRG payment system in con- taining Medicare's hospital costs remains to be seen, as does its impact on the adoption and use of medical technologies. Thus, experience with other approaches to hospital cost containment is still necessary. Furthermore, the effect of a Medicare-only payment system on health care de- livery is unknown. Whether a Medicare-only DRG payment system operating in the context of a largely pass-through system for other payment can bring about the desired changes in hospital and physician behavior on which ultimate cost savings depend is largely unknown. Thus, it would be useful to examine how all-payer systems perform. Option 5: Encourage DHHS to support further refinement and development of case-mix meas- ures other than DRGs.? Congress has recognized the need to refine the DRGs as the case-mix measure for hospital pay- ment by mandating several studies. Severity-of- illness measures, for example, will be studied for their applicability within the DRG system. Even with refinements, however, a DRG-based system may not be optimal, and case-mix measures that account for resource use more accurately than DRGs might be found. Examples of potential alternatives to DRGs are Disease Staging, the Severity-of-Illness Index, and Patient Manage- ment Categories (PMCs). 3The background information on this option was discussed in detail in the OTA technical memorandum entitled Diagnosis Related Groups (DRGs) and the Medicare Program: Implications for Medical Technology (343). 158 Medical Technology and Costs of the Medicare Program Disease Staging and the Severity-of-Illness In- dex were both designed to provide a framework for classifying diseases according to the relative severity of the patients condition. Both have re- quired extensive developmental work and testing, which are still underway. Use of either measure would require more data than are generally avail- able on Medicare claims at the present time, though the staging approach can be employed by using data that are normally included in hospi- tals’ computerized records. Neither measure has reached the point where it is suitable for wide- spread implementation in a reimbursement con- text. However, the existence of such measures, at a minimum, serves as a reminder that the pres- ent set of DRGs ultimately may not be the best system for classifying patients. PMCs are also in the developmental stage and are being tested now. This case-mix measure dif- fers from others, including DRGs, in that PMCs are normative. Physicians specify an optimal set of clinical care components based on a patient's clinical characteristics. This set is the basis for the relative cost weights of PMCs. The system ap- pears unique in that it recognizes that optimal pa- tient management should be the focus of a sys- tem that seeks to encourage efficiency. Development of alternatives to DRGs will re- quire continued interest and funding from DHHS. Research and demonstration projects in which these and other alternative case-mix measures can be studied and refined may need additional fund- ing. Other studies could be mandated if necessary. Option 6: Encourage DHHS to develop DRG price adjustment methods that result in higher DRG payment rates for those hospitals that purchase and use certain socially desirable but costly new medical technologies. Medicare's DRG hospital payment system pro- vides financial incentives to hospitals to purchase and use those technologies that reduce costs per case. Thus, specific policy might be required to encourage the adoption of socially desirable tech- nologies that raise costs. Making extra payment for a DRG conditional on a hospital's adoption and use of a new technology would encourage the technology's diffusion. Two possible mechanisms for making adjust- ments in DRG prices that would be conditional on the adoption of technology are reliance on hospital-initiated appeals of DRG prices and cre- ation of new technology-specific DRGs. Either of these mechanisms could stimulate adoption of desirable but cost-raising technologies. If a new DRG with a higher rate were created specifically to pay hospitals only if they actually adopt and use a technology, the adjustment in price would clearly be an incentive to adopt. Both of these mechanisms have drawbacks. First, reliance on case-by-case hospital appeals of DRG prices is likely to be administratively costly and cumbersome in comparison to other adjust- ment methods. Second, both the appeals mecha- nism and the creation of new DRGs would result in an effective increase in the number of DRG cat- egories, or DRG inflation. Eventually, if the num- ber of categories increased to significant levels, DRG payment would develop into a system re- sembling a fee-for-service system, because more and more categories would exist for specific pro- cedures or services. Such a system, in turn, would destroy the incentives inherent in per case pay- ment to minimize the utilization of services. Conditional DRG payment adjustments would work best if limited to a few very high-cost tech- nologies whose introduction would be strongly discouraged in the absence of such an adjustment, The majority of cost-raising technologies prob- ably would be adequately handled through peri- odic reestimation of the costs of DRGs. The timeliness of reestimation is considered in Option 7. Option 7: Amend the Medicare law to require an- nual reestimation of the relative costs of DRGs. Congress has recognized the need for periodic adjustment of DRG prices. The law currently calls for recalibration (assigning new relative weights to DRGs or establishing new DRGs) at least every 4 years. This option offers a refinement of that mandate and differs from it in two ways: 1) it defines reestimation as a type of recalibration based on the estimation of historical relative costs of technology; and 2) it suggests adjusting the rela- tive rates every year. Ch. 9—Findings and Policy Options ¢ 159 From year to year, some DRGs will experience cost-saving technological innovations; others will experience cost-raising ones. The relative prices of personnel, supplies, and other inputs will also change, with consequences for relative DRG costs. Relative DRG rates should reflect the relative costs of efficient and clinically optimal care, including appropriate use of technologies. The purpose of annual reestimation would be to keep DRG rates reasonably in line with the cost of efficient care. Annual reestimation of relative DRG rates would also encourage the rational adoption and appro- priate use of medical technologies. Under the cur- rent statutory requirement, up to 4 years may elapse after the emergence of a new, fast-develop- ing technology before the relative DRG rates would even begin to reflect the change, and the gradual adoption of new technologies would fur- ther delay the full capture of their effects in DRG prices. Implementation of annual reestimation would provide a quicker mechanism for adjusting prices to the efficient production of care. Financial incentives for hospitals to increase their efficiency would not be weakened by fre- quent reestimation of relative costs of DRGs. This option focuses on relative prices, so it would still be in a hospital's best interest to perform effi- ciently. Since the reestimation would be based on average prices across hospitals, the inclusion of a large number of hospitals’ cost reports in the reestimation is important so that there is a large sample size within each DRG. The potential gains from DRG creep, or delib- erately assigning a patient to a higher priced DRG to receive more payment, would diminish if DRG prices were restimated frequently. New DRG prices or weights would reflect the new distribu- tion of patients among DRGs and the new aver- age costs per DRG. Over time, reestimation of prices would cause the more profitable DRGs to become less profitable, and the less profitable ones more profitable. Thus, one could expect a gradual decline in the potential for “gaming” via DRG creep with periodic reestimation of DRG prices. One advantage of frequent updates of DRG prices would be the alleviation of the lag that is already built into the current system because fiscal year 1984 prices are based on fiscal year 1981 data. The 1981 data reflect accepted modes of practice in 1981 and do not account for new tech- nologies currently in practice. Furthermore, Medi- care cost reports are available on an annual basis, so annual reestimation would require no addi- tional data collection (at least until fiscal year 1988, when cost reports will no longer be re- quired). Frequent reestimation of DRG prices would capture gradual changes in hospitals’ costs which should result as hospitals seek technologi- cal efficiency. If the reestimation were too frequent, however, hospitals would not be able to plan their moves toward efficiency as well as they could if they were certain of particular payments for a segment of time. Because the year-to-year changes in rela- tive prices are likely to be small and gradual, an annual (or at most biannual) reestimation would seem to provide the optimal mix of certainty for hospital planning with the adjustment of prices for efficient production of care. Administering annual reestimations would be slightly more expensive for Medicare than reesti- mation every 4 years. Not only would the proc- ess be more frequent, but the data requirements for reestimation—cost reports and patient charges —would continue indefinitely. The requirements for data collection would increase administrative costs to the hospitals, although many hospitals are likely to use the same data for internal man- agement functions under DRG payment. Option 8: Amend the Medicare law to strengthen controls over hospital admission rates. Under Medicare's DRG prospective payment system, hospitals are paid by Medicare on the basis of the number of Medicare cases treated in each DRG. Thus, the DRG hospital payment sys- tem provides a strong financial incentive for hos- pitals to increase the number of Medicare patients they admit. One of the responsibilities of the PRO program is to monitor hospital admissions. The following suboptions are presented as possible support for the PRO regulations. 160 ¢ Medical Technology and Costs of the Medicare Program Option 8a: Amend the Medicare law to require a second deductible for rehospitalization within 60 days of the first admission. Implementation of this option could save the Medicare program money in two ways. Current- ly, Medicare beneficiaries do not have to pay a second deductible for rehospitalization within 60 days of the first day of the initial admission. By increasing beneficiary cost-sharing, adopting this option would save Medicare program costs. Im- posing a second deductible could also save costs by encouraging cost consciousness in physicians. Physicians would be aware of the financial bur- den of readmissions on their patients, and such patients might take a more active role in decision- making about readmissions. Identifying unnecessary discharges and readmis- sions is a PRO responsibility. It is important to recognize that some readmissions are necessary. Often, elective procedures are postponed until pa- tients are stronger and more able to withstand sur- gery or invasive procedures. Sometimes unrelated diseases may strike patients in short periods of time. A drawback to the second deductible is that Medicare beneficiaries already carry a heavy fi- nancial burden for their health care. Furthermore, the amount of control patients actually can have in making decisions about hospitalization is ques- tionable. Thus, a second deductible might de- crease access to inpatient hospital care for some elderly and disabled patients. On the other hand, its effect on admissions may be low because a high proportion of the elderly have supplemen- tal (“Medigap”) medical insurance. (Of course, Medigap insurance premiums are likely to rise in response to cover the increased costs.) The strength of the incentives is unknown, because there are no empirical studies on how the elderly would respond to additional cost-sharing. Finally, implementation of this option might ac- tually increase Medicare's costs. Physicians and patients might decide to keep patients in the hos- pital if there is a chance that a readmission might be possible in 2 or 3 weeks. Option 8b: Amend the Medicare law to provide a short-term outlier policy for DRG payments. With a short-term outlier policy, Medicare would not pay the full DRG payment if a length of stay were less than a particular number of days for each DRG. Such a policy would counteract the incentive under DRG payment for hospitals to admit patients for very short stays instead of treating them on an outpatient basis. A potential benefit could be that marginally ill patients who might not require the intensive services of an in- patient hospital stay would not be admitted. A potential problem would be that such patients might be kept beyond the outlier threshold length of stay to avoid the short-term outlier payments. The strength of the incentives under this kind of policy would depend on the marginal costs of ad- mitting a patient compared to the outlier price, as well as the hospital's influence with its admit- ting physicians. The effect of a short-term outlier policy on the volume of patients in each hospital or in hospitals in the aggregate is unknown. Option 8c: Amend the Medicare law to adjust DRG payments for patient volume changes. DRG prices are based on the assumption that hospitals’ annual volumes are predictable and vary only slightly from year to year. Adjusting DRG payments to hospitals for volume changes could directly balance the financial incentive that DRG payment gives hospitals to admit more and more patients. Unusual increases in the annual number of hospital admissions could trigger a pen- alty charge against the hospital's total Medicare payment, or each DRG payment could be de- creased by a certain amount. Thus, for example, if a hospital's volume increased 10 percent, DRG payments could be decreased by an amount re- flecting the marginal costs of those additional ad- missions. While this payment reduction would discourage hospitals from unnecessarily increas- ing their admissions, the net decrease in a hospi- tal’s revenues would be relatively small. A two-way volume adjustment imposing a pen- alty for unusual increases in admissions and giv- ing a bonus for unusual decreases could be insti- tuted. A penalty for unusual increases in patient volume could weaken the incentives for hospitals to overadmit patients. A bonus for unusual de- creases in admissions could protect hospitals from the financial problems caused by population Ch. 9—Findings and Policy Options 161 migrations. It should be noted, however, that some hospitals are inefficient or provide a lower quality of care and may experience decreased ad- missions for these reasons; the low-end volume adjustment might allow them to remain open when they actually should close. Another important consideration for deciding whether to adopt a volume adjustment is the fixed-cost-to-variable-cost ratio for various hos- pitals. Although low-occupancy hospitals may need to increase their patient volumes in order to cover their fixed costs, large volume increases may trigger high variable costs. Thus, even without volume adjustments, the fixed-cost-to-variable- cost ratio and its relationship to the DRG pay- ment and penalty are important. Volume adjustment also discourages the poten- tial for specialization of services within hospitals for the purposes of efficiency. Payment based on DRGs gives hospitals financial incentives to treat patients in those DRGs in which the hospitals are efficient and to avoid treating patients in those DRGs which lose money. Thus, consolidation of services is a potential result of the DRG payment system. Again, high- and low-occupancy hospi- tals would be affected differently, with fixed and variable costs important to their decisions about which patients to treat. Option 8d: Amend the Medicare law to establish financial incentives for physicians’ decisions about hospital admissions that are consistent with the incentives of DRG payment. Decisions to admit and discharge patients and to use medical technologies during hospital stays are primarily made by physicians, although pa- tients’ decisions are certainly important. Under DRG hospital payment, physicians will probably be pressured by hospital administrators to dis- charge patients earlier than they previously have and to readmit patients for elective procedures. While quality assurance and utilization review programs will reinforce physicians’ own inclina- tions to provide adequate care for their patients, financial incentives could be established to miti- gate any potentially harmful pressure from hos- pitals. One possibility would be to pay physicians only half their fees for rehospitalization within 60 days. A potential administrative and policy problem might arise when more than one physician is involved in a patient's care. In a case, for exam- ple, in which one physician admits a patient to a hospital for diagnosis and then, after dis- charge, refers that patient to a specialist for long- term treatment, it remains questionable whether a hospitalization should be counted against the patient (if Option 8a were implemented), the phy- sician (in the case of this option), or the hospital. Identifying readmissions would be a fairly sim- ple task, and monitoring their appropriateness is a PRO responsibility. Difficult judgments will have to be made about which ones are unnec- essary and what caused the readmission—not all complications can be predicted, nor can they nec- essarily be averted by additional hospital days during an initial admission. The difficulty in mak- ing such judgments could be compounded by ad- ministrative difficulties of combining Part A data on hospital admissions with Part B data on phy- sician payment. Option 9: Amend the Medicare law to control capital expenditures by hospitals by removing capital cost pass-throughs. Historically, capital expenditures by hospitals have been reimbursed under Medicare on the basis of depreciation and incurred interest expenses. Under the DRG hospital payment system, at least during the 3-year transition period, capital costs will continue to be treated as pass-throughs. The pass-through method of payment for capital, which directly links the level of payment to the amount of capital investment undertaken by a hospital, does not discourage inefficient capital purchases. However, its feasibility has been dem- onstrated, it is fair in the sense of treating hospi- tals alike, and it poses no barriers to equal access to medical services. The Social Security Amendments of 1983 (Pub- lic Law 98-21) mandated a DHHS study of alter- native methods of handling capital under Medi- care’s DRG hospital payment system. Three methods of limiting capital payment to contain costs for Medicare were discussed in chapter 6: 1) uniform payment (flat rate) approaches; 2) hospital-specific controlled rates; and 3) if the 162 » Medical Technology and Costs of the Medicare Program pass-through is continued, direct regulation of capital expenditures. If the capital pass-through is continued, capi- tal expenditures could be regulated on a project- by-project basis. Such regulation could be imple- mented through certificate-of-need (CON) pro- grams or other agencies. The political process in- volved in CON programs might, but would not necessarily, provide equity of access to medical technologies. Studies of the effectiveness of CON regulations in containing costs are inconclusive, and the efficiency of a process such as this is ques- tionable. Administrative costs would be very high, and some hospitals might be unable to af- ford the costs associated with the application proc- ess and, thus, be impeded from making capital investments. New hospitals might not be consid- ered because of the high cost of application and the possibility of denial. The following two suboptions are presented as possible alternatives to capital pass-through pay- ments under DRG payment. They are not in- tended to represent the entire range of methods of handling capital. Option 9a: Incorporate a flat rate for hospital capital into the DRG rates. The uniform payment approach would treat all hospitals or all those in a class alike, regardless of their level of capital expenditures. A flat rate for capital, whether calculated either as a fixed percentage of the DRG price or as a flat rate per bed, would encourage hospitals to provide care at the least possible total cost to the hospital. Since new technologies would be judged in terms of their impact on total costs, not just on operating costs, a flat rate would give hospitals more in- centives to be efficient than the current capital cost pass-through. Hospitals would be further encour- aged to specialize and join in plans for regionaliza- tion of health services. Despite the increased efficiency of a uniform rate for capital built into the DRG rates, it would be difficult to ensure that this type of system would be fair to all hospitals. Hospitals that in the past have had lower ratios of capital to oper- ating cost would receive more than before, while those with historically high ratios would receive less. Thus, public hospitals would probably fare better with a flat-rate system than they have in the past, at least in the short term. Multihospital systems, whose affiliated hospitals could pool cap- ital payments and smooth out fluctuations in cap- ital expenditures across affiliated hospitals, would also be favored. Implementing the uniform pay- ment approach might require a difficult and costly transition period if those hospitals that have made major investments in recent years or that face them in the near future are not to be unduly penalized. A uniform capital payment method such as a flat rate within the DRG payment would not dis- courage equal access to medical technologies and might help redress some current inequities because of the possibility of increased capital payments to public hospitals and others that serve low-in- come patients. It might, however, have a nega- tive impact on the regionalization and specializa- tion of services among hospitals because of the difficulty some hospitals would have in accumu- lating enough capital to specialize. If the popula- tion shifts, moreover, new inequities for opening and closing hospitals might appear. The question of separating capital expenditures for equipment from those for buildings should be addressed. Making the transition from the cost-based reim- bursement system to a flat rate capital payment could be difficult for hospitals. Phasing in the flat rate by reducing the proportion of capital pay- ment that would be a pass-through over a period of time could alleviate some of the financial dif- ficulties but would not necessarily reduce the amount of paperwork or data collection costs. Option 9b: Build hospital-specific capital allow- ances into the DRG system. To implement this option, hospital-specific cost information would be taken into account to estab- lish a base period level of capital payment, and the payment level would be increased by an in- dex over time. One approach would be to use the hospital's capital costs in a base year and then add a percentage for inflation. Another would be to limit capital payments to a percentage of a hos- pital’s operating costs in each year. Ch. 9—Findings and Policy Options ¢ 163 Tying capital payment to a hospital's level of capital costs in a base year or to the hospital's operating costs would be efficient but might be unfair. Hospitals that were most highly capitalized in the past would be rewarded, while those with low levels of capital would forever receive low payments. An increased administrative burden would be put on the hospitals, especially in the transition period from pass-through to hospital-specific con- trols. In the early years of implementation, this system would not work well for hospitals that re- quire major capital expenditures. Perhaps for these reasons, it might be better to limit this ap- proach to the movable equipment portion of cap- ital, which typically has shorter lifetimes and lower variations in asset values among hospitals. Option 10: Provide adequate resources or incen- tives for States to experiment with alternative hospital payment systems, especially those in- volving all payers. In States where Medicare is the only third-party payer using prospective payment, hospitals will have incentives at best to shift costs to other pay- ers and at worst to treat patients differently de- pending on their insurance. Savings to the Medi- care program may not offset these other social costs. Furthermore, a Medicare-only hospital pay- ‘ment system such as the current one may not pro- vide sufficient leverage to lower the annual rate of increase in hospital costs. Further experimen- tation with hospital payment systems would be desirable to learn which methods of cost contain- ment save the most money to Medicare and soci- ety as a whole. The Social Security Amendments of 1983 (Pub- lic Law 98-21) encourage States to experiment with payment systems that cover multiple third- party payers and differ from DRG payment by requiring the Medicare program to participate in any State-legislated prospective payment program that covers at least 75 percent of the State's population; makes provisions for competitive health plans; assures the Federal Government that access to hospital care for Medicare and Medic- aid beneficiaries will not decline; and assures the Federal Government that hospital costs will not be higher under the State program. Four States —Maryland, Massachusetts, New Jersey, and New York—already regulate all payers in their hospital prospective payment systems and have Medicare waivers. Other States will be examin- ing the effectiveness of the payment systems in these four States with Medicare waivers and others experienced in containing hospital cost in- creases. The goal of this option would be to en- courage the efforts of several State legislatures that are working on this issue. While all-payer systems would increase the pay- ment system's leverage over the hospital indus- try and reduce hospitals’ tendencies to shift costs among payers, such systems have been criticized for their potential to inhibit competition. Accord- ing to this viewpoint, regulation precludes mar- ket forces from exhibiting their desired effects and thwarts innovation. How much health care deliv- ery systems should respond to market forces is debatable. Changes in hospital behavior in response to the all-payer system incentives are predicted to range from increasing efficiency through specialization and interhospital cooperation to closing their doors (6). The viability of particular hospitals can be guaranteed or threatened, depending on the goals of the particular State payment system. For example, in New York, the capital pooling system was established to save a number of inner-city hospitals; and in New Jersey, several small hos- pitals have closed or merged at least in part be- cause they could not earn enough money through DRG payment (47). There is empirical evidence from New Jersey that vertical integration has been encouraged by the DRG all-payer system there (75). As long as only hospitals are under the all-payer system, there will be incentives for them to branch out and open separate home health agencies, nursing homes, and satellite outpatient clinics. They will have to continue to compete with each other for physicians and patients. All-payer prospective payment systems are not the only approaches that may be attempted by the States. Other approaches to the control of hos- pital costs, however, have significant limitations. Two of these methods, increasing the patient's responsibility for cost-sharing and limiting pro- 164 ¢ Medical Technology and Costs of the Medicare Program viders through contracts between Medicare and hospitals, were examined in chapter 6. Factors such as the patient's relative lack of power and information to make informed decisions about the use of technologies in hospitals and the apparently strong preference of the elderly for supplemental medical insurance regardless of its cost imply that beneficiary cost-sharing for hospital services is not likely to be as effective in altering the patterns of use of hospital technologies as desired. And al- though contracting might save Medicare dollars, such contracting would represent an abandon- ment of the principle of beneficiary freedom of choice of provider on which Medicare was built and would force subsidies of hospital care from other payers. Option 11: Consider ways to extend the DRG pro- spective payment system to payers other than Medicare (e.g., Medicaid) without relying on State waivers. This option differs from the previous one be- cause this option addresses the current DRG pro- spective payment system, whereas the previous option sought alternative systems. The desirability of this option depends on the effectiveness of DRG payment in encouraging the appropriate adoption and use of medical technology and in containing Medicare costs. The effectiveness of DRG pay- ment is currently unknown. As discussed in the previous option, a multi- payer system would increase the payment system's leverage over the hospital industry. A multipayer system would also diminish hospitals’ opportuni- ties to shift costs from one payer to another, al- though multipayer systems need not pay all payers the same price. The clearest technical problem with extending the current DRG-based hospital payment system beyond the Medicare program is that DRG prices have been based almost exclusively on Medicare data. Comparable Medicaid data bases are not available, and the Medicare and Medicaid popula- tions are so different that the use of Medicare- generated DRG prices for the Medicaid popula- tion would be unfair. Even if both data bases were available for recalculation of DRG prices, the problem of within-DRG variation in costs per pa- tient would be exacerbated because of the diver- sity of populations. It should also be noted that many State Med- icaid agencies are experimenting with methods of prospective payment. Potential refinements for the Medicare system could come from these ex- periments. Physician Payment ISSUE: How can Medicare's physician payment method be used to improve the incentives for appropriate technology adoption and use? Physicians can influence both the number of pa- tient visits and the use of a variety of technologies, especially diagnostic tests. Furthermore, the ways in which physicians are paid can influence phy- sicians’ adoption and use of medical technologies. Physicians who are paid on a fee-for-service basis have incentives to see more patients more often and to provide more technologies. Physicians (or practice plans in which they participate) paid on a capitation basis would want to increase the num- ber of their patients but would have incentives to keep the number of visits low (or nonexistent) and to use particularly cost-effective technologies. Physicians’ incentives under a fee schedule sys- tem would depend on the particular type of sched- ule adopted. Under fee schedules based on patient visits, physicians would have an incentive to schedule more visits but would have a disincen- tive to use a large number of technologies (par- ticularly those whose costs are high in relation to the fee per visit received). If the fee schedule were based on episodes of illness, physicians would have incentives to treat for more episodes but would want to keep patient visits for each episode and the use of costly technologies at a minimum. Excessive adoption and use of medical technol- ogies are sometimes incorporated into medical practice through habitual behavior of physicians and because the health care system contains few disincentives for these practices. Excessive use oc- curs within the norms of medical practice and is Ch. 9—Findings and Policy Options ¢ 165 evident across the spectrum of technologies avail- able to physicians. Physicians are motivated by their training to do all they can for their patients, and generally they have not—in the past—had to be concerned about the costs of the care they provide. Indeed, there have been economic incentives for physi- cians to increase demand for health care services. These factors, singly and combined, often result in overuse of medical technologies. Two types of changes in Medicare's physician payment method could contain costs for the Medi- care program and help rationalize the adoption and use of medical technologies: 1) requiring ben- eficiaries to assume more responsibility for their health care costs, either through increases in pa- tient cost-sharing or reductions in the types of services covered; and 2) imposing restraints on the amount and type of payment to physicians. The options presented below are grouped accord- ing to these two categories—cost-sharing by Medi- care beneficiaries and changes in physician pay- ment methods. Either type of change could result in cost savings for the Medicare program, but each type would have different effects on the adoption and use of medical technologies and on access to medical care by Medicare beneficiaries. The op- tions presented do not always relate directly to medical technology, but they are important be- cause of their indirect effects on technology. Changing Medicare's voluntary physician assign- ment policy could strengthen the effect of imple- menting some of these options. Option 12: Amend the Medicare law to increase beneficiary cost-sharing for Part B services. Several methods of cost-sharing were explored in chapter 7. Increasing the premium for Part B benefits would increase revenues for the Medicare program, but the evidence suggests that premium cost is too far removed from the use of medical services to alter patterns of use. This change would spread the burden of costs among many beneficiaries without regard to their use of the medical care system, and demand for medical technologies probably would not be affected. In- creasing the deductible for Part B, again, might not reduce the use of services. Medicare benefi- 25-337 0 - 84 - 12 ciaries often have chronic diseases and require multiple physician visits. Coinsurance raises financial barriers each time medical services are sought. Increasing Medicare's Part B coinsurance requirements would have a more significant effect on the demand for medi- cal technologies than would a deductible. Part B beneficiaries are already responsible for a 20- percent coinsurance payment for assigned care and even more for nonassigned care. Modest in- creases in coinsurance requirements would prob- ably have little effect, beyond the incentives already accompanying current coinsurance re- quirements, on patient behavior. Large increases, on the other hand, would probably result in fewer visits to physicians but might also result in a re- duction in access to necessary medical care, espe- cially for the lower income elderly who are not eligible for Medicaid. In summary, greater cost-sharing by Medicare beneficiaries under Part B could help contain Medicare program costs, in part by a shifting of costs to beneficiaries and in part through some resulting decrease in patient visits. It is unclear that all appropriate technologies would be pro- vided with greater cost-sharing, however, because Part B beneficiaries might have to forgo some nec- essary medical care. Option 13: Discourage Medicare beneficiaries’ purchase of private supplemental ( “Medigap”) insurance. Private insurance companies have offered, and many Medicare beneficiaries have purchased, sup- plemental (“Medigap”) insurance policies to cover, at least partially, patients’ out-of-pocket medical expenses. Noting that the type of extra first-dollar coverage that Medigap policies provide partially nullifies the intended effects of Medicare's deduct- ible and coinsurance requirements on use of med- ical services, some observers have suggested tax- ing Medigap policies to make their purchase less attractive. A principal objection against taxing Medigap insurance is fairness, because the gen- eral population's health insurance policies retain their tax advantages. If this option were adopted, the constitutionality of selective taxation would most likely be challenged. 166 ¢ Medical Technology and Costs of the Medicare Program Because elderly people already pay about one- third of their medical expenses out-of-pocket and need more medical care than younger people, this option raises two broad issues: 1) whether Medi- care should become the sole medical insurer for the elderly rather than providing a floor of insur- ance coverage as it does now, and 2) whether the Medicare program will provide adequate insur- ance coverage, especially in view of present cost- containment efforts. Option 14: Place further limits on payment to physicians under Part B of Medicare. Methods of limiting payments to physicians fall into two categories: 1) limits on allowable physi- cian fees, and 2) the use of fee schedules. Both methods raise the issues of assignment of claims and changes in coding of procedures. Option 14a: Amend the Medicare law to place a ceiling on allowable physician fees under Medicare. The Medicare Economic Index currently limits the rate of physicians’ fee increases to the rate of their cost increases. Other types of caps that could be imposed include freezes on physician payment levels for a specified period of time and percent limitations on the annual rate of allowable fee in- - creases. A cap on physician payment levels by Medi- care is unlikely to change overall reimbursement to physicians, because physicians could increase patient visits, increase the number of technologies provided to each patient, and shift to a higher priced mix of technologies. Because it is unlikely that physicians would charge less than the cap, such a cap would be effective in containing pro- gram costs only if it were set low. A low fee cap, however, would result in widening the gap be- tween fees paid by Medicare (allowed charges) and those paid by private patients (billed charges). Thus, fewer physicians would be likely to accept assignment. With less assignment, Medicare ben- eficiaries would have to pay an even greater share of their total health costs. Program savings would accrue at the expense of patients, and increased cost-sharing might result in decreased access to needed medical care. Option 14b: Move to fee schedules for physician payments under Medicare. Current limitations on increases in allowable charges are slowly turning Part B physician payments into de facto fee schedules, but histori- cal specialty and geographic differences in fee levels that have developed under the fee-for- service system of payment remain. Cost-based fee levels could be developed, but they would require a data base that relates costs to charges in some rational fashion. The difficulty in developing such a data base should not be taken lightly. Payment through fee schedules would necessi- tate a reformulation of the diagnostic and pro- cedural codes for physician services that are currently used by the Medicare program. The number of procedural and diagnostic categories in these codes has increased by the thousands since the onset of Medicare and Medicaid. The large number of categories increases the likelihood of incorrect coding, and the availability of numerous categories to choose from in the billing process makes it possible for physicians to bill for higher priced services than those actually provided. Furthermore, since the existing fee and price system provides financial incentives for the use of “technology-intensive” medical care, one long- range objective in developing fee schedules might be an overall review of the relative values of all procedures. The fee system could then be revised in such a way that technology-oriented services, such as performing diagnostic tests and surgical procedures, might be valued neutrally with cog- nitive services, such as taking medical histories and providing patient counseling. Current fees and payment methods favor the technological pa- tient services over the cognitive. Before fee schedules are developed, packages of physician services, possibly designed to com- plement existing DRGs for hospital care, could be developed. Inpatient surgical services would be one logical starting point. Other physician serv- ices could also be included. For example, DRG- based payment to physicians might be applied ini- tially only to acute inpatient care, then extended to include physician services in skilled nursing fa- Ch. 9—Findings and Policy Options e 167 cilities, and then to ambulatory care for those diagnostic categories where such an inclusion was found to be appropriate and feasible. Coordinat- ing physician and hospital payments is important, ‘because there is an incentive under DRG hospi- tal payment to move services out of the hospital settings. An advantage of a fee schedule for phy- sicians would be a probable reduction in inpatient hospital physician consultations and possibly even reduced length of stay. A disadvantage is that it would depend on inpatient care as the starting point, leaving other ambulatory visits outside the system, at least at first. Option 15: Change Medicare's claim-by-claim voluntary physician assignment policy. Medicare's current policy of allowing physi- cians to decide whether or not to accept assign- ment on a claim-by-claim basis allows costs to be shifted from the physician to Medicare benefici- aries. Although such cost shifting may decrease demand for medical technologies, it may also de- crease access to necessary medical care. Chang- ing assignment policy would strengthen Option 14 and suboptions. One type of change in assignment policy is to make assignment mandatory, so that physicians are not paid at all by Medicare if they refuse to accept assignment. Under mandatory assignment, beneficiaries would have greater incentives to seek out physicians who accept assignment, because they would assume all of the costs of nonassigned care (except for any portions covered by a Medi- gap policy). Physicians would also have to weigh the financial impact of losing their Medicare in- come altogether versus accepting the payment levels set by Medicare. Thus, mandatory assignment provides an in- centive for cost-conscious behavior in both pa- tients and their physicians, whereas under pres- ent assignment policy, most of the burden of cost savings falls on Medicare patients. Mandatory assignment could reduce access to medical care for Medicare's beneficiaries, however, because they would be responsible for payment of charges for nonassigned care. At least under current assignment policy, Medicare pays 80 percent of allowed charges for nonassigned care. An alternative to establishing a mandatory “all- or-nothing” assignment policy or maintaining cur- rent assignment policy is for Medicare to pay less for nonassigned than for assigned care. This might spread the burden of cost-sharing more equitably between patients and their physicians and provide significant incentives to both groups to be more conscious of costs. Reduced payment could be im- plemented either by further reducing allowable charges for nonassigned versus assigned services, or by increasing the coinsurance requirement for nonassigned care (e.g., from 20 to 50 percent). Either approach could be designed to achieve sim- ilar savings for the Medicare program, but re- ducing allowable charges would be directed at physicians, while increasing the coinsurance re- quirement would be directed at patients. Patients may ultimately bear most of this shift in costs in either approach, however, because of their liability for the difference between billed and allowed charges for nonassigned care. An addi- tional adjustment under this alternative could be to prohibit billing by physicians for the difference for nonassigned care, and payment could be made directly to the physician for allowed charges (minus the patient's coinsurance share). Under current payment for nonassigned care, allowed charges are computed and 80 percent paid to the patient, not to the physician, who must then col- lect from the patient. Most, if not all, payment changes in the Part B program are likely to place additional financial burdens on patients. Some observers suggest that the additional burden (which Medigap insurance moderates) is desirable, because it gives patients incentives to behave in a cost-conscious manner when seeking medical care. Patients must have the information on which to make such informed decisions. Thus, even if none of the financing and assignment options is undertaken, it would be desirable for patients to have more information. Medicare beneficiaries could be given information on their payment responsibilities, what assign- 168 » Medical Technology and Costs of the Medicare Program ment means, the relative charges of physicians in their areas, and names of physicians who accept assignment. Option 16: Require review of physicians’ services. Option 16a: Encourage the development of a re- view program for physicians’ services. Professional Standards Review Organizations (PSROs), which concentrated their review on the appropriateness of hospital admissions and lengths of stay, were statutorily replaced by PROs in 1982 (Public Law 97-248). Under Medicare's DRG hospital payment system, PROs will have the added responsibilities of reviewing the validity of diagnostic information provided by hospitals (DRG verification) and the appropriateness of moving patients from hospitals to less intensive care settings such as nursing facilities and home health care. DRG payment provides financial incentives to reduce the unnecessary use of ancillary services in hospitals. This option would address the fact that similar disincentives for the excessive use of medical services in physicians’ offices and other ambulatory care settings are lacking. While there is evidence that physicians do respond to restraints on physician payment levels by increasing the number of services they provide and shifting to a higher priced mix of services (153,275), how- ever, one problem in identifying excessive use on a procedure-by-procedure and physician-by-phy- sician basis is in differentiating between normal and excessive provision of medical care. Under Public Law 98-21, HCFA is required to study the possibility of extending DRG payment to physicians’ services. The resulting information will, in essence, reflect norms of care on which DRG-based prices could be computed. Such norms of care might also be used to extend PRO activities to inpatient physicians’ services. Simi- larly, information for extending DRG payment to ambulatory care might be used to develop re- view systems for office-based care. The costs of conducting such an extensive review program, however, would be substantial. Option 16b: Require or pay for second opinions in elective surgery. Voluntary second surgical opinion programs generally have had low participation rates. In HCFA's two demonstration programs, for exam- ple, only 1.2 and 0.3 percent of Medicare benefi- ciaries who had surgery participated in the pro- grams and sought second opinions, despite waiver of deductibles and coinsurance for the second opinion (127). Because of low participation rates, the potential savings from voluntary programs are not great. On the other hand, there is growing, though not comprehensive, evidence that mandatory sec- ond surgical opinion programs reduce the amount of elective surgery. The reduction takes place in part because the first surgeon is aware that the patient will need a second opinion and in part because patients tend to follow the second sur- geon’s recommendation. As an alternative to an across-the-board, man- datory program, a mandatory program could be initiated for a few elective surgical procedures. The procedures included could be slowly ex- panded if the original program leads to cost sav- ings. Monitoring the impact of this option, if im- plemented, would thus be very important. Alternative Approaches to Changing Incentives ISSUE: What broad approaches, other than those directly involving Medicare's payment mechanism, could be used by Medicare to encourage the appropriate adoption and use of technology? Most cost-containment strategies that rely on the existing Medicare program structure empha- size restraints on payments made to hospitals, physicians, and other providers, augmented by utilization and other types of review programs. Such approaches are complemented by efforts Ch. 9—Findings and Policy Options e 169 to increase beneficiary cost-sharing. Other ap- proaches that may be more feasible or desirable include changes that must involve the general health care system but that Medicare could em- brace and changes in the structure of the Medi- care program itself. Long-range cost containment in the Medicare program is limited by the kinds of health care de- livery systems available and the influence that Medicare financing can have on the settings and kinds of technologies provided. In recent years, the Medicare program has granted exceptions to specific alternative types of care (e.g., freestand- ing ambulatory surgical centers) and encouraged the development and evaluation of alternative de- livery methods (e.g., preferred provider organi- zations) through demonstration programs. Thus, Medicare's efforts in developing competition with the types of care predominantly available have been to identify and encourage other types of pro- vider practices and modes of delivery. In the long run, it is hoped, the use of alternative sites and organizations will lead to cost-effective health care by encouraging competitive behavior among pro- viders. A complementary approach to increasing com- petition among providers involves moving from the current Medicare program structure to a sys- tem in which a variety of types of health insur- ance coverage would be made available to Medi- care beneficiaries. The most discussed possibility is the use of vouchers, wherein persons eligible for Medicare would receive a specified amount of money to purchase health insurance from the marketplace. The assumption is that beneficiaries would be encouraged to select delivery systems that offer the best benefits for the least amount of money. To the extent that health maintenance organizations (HMOs) or preferred provider orga- nizations (PPOs) can achieve this goal, these orga- nizations would be selected. Competition can occur both at the point of in- surance and at the point of service delivery. In both cases, payment by capitation is believed to increase competition. A voucher, with its fixed dollar subsidy, is actually a capitation method of payment for total medical care (both inpatient and outpatient). The beneficiary who receives the voucher may benefit from competition among plans. The beneficiary could purchase traditional fee-for-service insurance or could select a plan such as an HMO that accepts a capitated payment per enrollee. Important decisions regarding com- petition for policymakers in the Medicare pro- gram include: 1) the relative emphases to be placed on the insurance versus the alternative de- livery system approach, and 2) the pace of adopt- ing the various competitive approaches into Medi- care. To increase the capability of Medicare to embrace the various competitive approaches, however, the program could undergo an admin- istrative change—merging Parts A and B. Option 17: Move toward a capitation payment system for Medicare. The extent and pace of changing the Medicare program to capitation payment depend on the capacity of the health care system to provide alter- native sites and organizations of medical care and on Medicare's leverage in promoting alternative delivery methods or requiring that they be sub- stantially available. In one sense, there is a chicken-or-egg question—i.e., must substantial changes in the health delivery system come first, or is it the financing leverage of programs such as Medicare (and its effects on the general health care system) that will lead to the desired health system changes? Under current Medicare policy, the implicit assumption is that health care system changes must come first. This assumption is reflected in Medicare-supported demonstrations of alternative delivery methods and Medicare's service-by- service adoption of alternative methods (e.g., am- bulatory surgery centers, special provisions for HMO participation). This assumption is also pres- ent in discussions among policymakers on Medi- care vouchers. A voucher program may involve mandatory or voluntary participation. A voucher could be completely voluntary and allow benefi- ciaries to reenroll in Medicare, it could require that the decision to opt out of Medicare be per- manent, or it could trigger mandatory participa- tion if and when more than half of the benefici- aries choose vouchers. It is believed, however, that the implementation of mandatory vouchers 170 Medical Technology and Costs of the Medicare Program for Medicare is not politically feasible (336). In other words, there is reluctance to end the cur- rent Medicare program per se and place the bur- den on beneficiaries to see if the market will respond. Under voluntary voucher proposals, the pol- icy is to provide enrollees with incentives to seek more cost-effective care, such as through HMO- or PPO-type organizations. If voluntary vouchers succeed in stimulating alternative systems, then the current Medicare program would slowly be replaced. A voluntary voucher system for Medicare could be implemented without fundamental changes in the basic Medicare program. Replacement of the current Medicare program would depend on the amount of use of the vouchers, which in turn would depend on the capacity of the health care system to provide cost-effective alternatives to present Medicare benefits. In sum, encouraging competitive approaches into the Medicare program can proceed by pro- viding enrollees with the opportunity to opt out of the basic program, or by transforming the basic payment program itself into a competitive mode. A Medicare program with capitation as the in- surance mechanism might be initially imple- mented in urban areas, particularly urban areas with competition for patients and with substan- tial availability of prepaid services. The pace at which a voucher-only approach might be imple- mented has already been explored by the several bills introduced in Congress. A cautious pace would be to implement voluntary vouchers as a first step with periodic opportunities for reenroll- ment in Medicare. Option 18: Merge Parts A and B of Medicare. The separation of the Hospital Insurance por- tion of Medicare (Part A) from the Supplemen- tary Medical Insurance portion (Part B) is ineffi- cient and allows incentives for the inappropriate provision of technologies to persist. Because of duplication of administration in the two parts, administrative costs to Medicare are probably higher than necessary. In addition, the fiscal separation, wherein Part A is financed through a payroll tax and Part B through premiums and general revenues, also seems wasteful. Merging Parts A and B could ameliorate the current revenue problems faced by Medicare. One proposal (84) would substitute a comprehensive, integrated set of benefits for current separate sets under Parts A and B. The benefits would be paid from a single trust fund formed from the Hospi- tal Insurance (HI) Trust Fund and the Supplemen- tary Medical Insurance (SMI) Trust Fund. Rev- enues would come from a combination of the current HI contributions, the general revenues projected for SMI expenditures, and new income- related beneficiary premiums. Currently, parallel data systems and adminis- trative mechanisms for Parts A and B do not allow easy cross-referencing by patient or provider. This problem is important because of providers’ efforts to shift costs from one part to the other (usually A to B). Some medical technologies have been covered under both parts, but because of dif- ferences in which part paid for their use at which time, facilities covered under one part or the other have duplicated equipment unnecessarily. Such duplication results in facilities and equipment that remain idle and raise prices in order to cover fixed costs. If there were one type of coverage and one payment source, at least some of this duplication and subsequent cost shifting could be avoided. A merger of Parts A and B would allow Medi- care beneficiaries to participate more easily in alternative organizations of care. A merger would also facilitate expansion of the DRG payment sys- tem beyond the inpatient hospital setting. For ex- ample, in the future, the DRG could be defined on the basis of an “episode” of care under the joint purview of Parts A and B. The definitional dif- ficulties could be substantial, but so could be the payoffs in efficiency, cost control, and appropri- ate medical technology adoption and use (201). The transition from Part A and Part B to an integrated system would be complex. Data sys- tems would have to be merged, and intermediaries and carriers would have to negotiate to be single Medicare contractors. Once integrated, however, the system could be more efficient, less adminis- tratively burdensome to hospitals, and less costly to society. Appendixes Appendix A.—Method of the Study and Case Studies This assessment of “Medical Technology and Costs of the Medicare Program” was preceded by a 2-month planning effort that identified areas on which to con- centrate and established a tentative study approach. The planning phase took place in April and May 1982, and resulted in a study proposal for the full assessment. The full assessment began on June 1, 1982. One of the first tasks undertaken was the selection of the advi- sory panel. Most of the studies undertaken at OTA rely on the advice and assistance of an advisory panel of experts. The advisory panel for a particular assess- ment suggests source materials, subject areas, case studies, and perspectives to consider; assists in inter- preting information and points of view that are assem- bled by OTA staff; and suggests possible findings and conclusions based on the accumulation of information produced by the study. The panel members review staff and contract materials for accuracy and vaiid- ity, discuss policy options of the study, and present arguments for and against the options and conclusions. They do not determine the report's final form, how- ever, and are not responsible for its content, direction, or conclusions. The advisory panel for the present assessment con- sisted of 20 experts with backgrounds in health policy, hospital administration, health economics, medicine, health insurance, State- and Federal-level Government, industry, and academia. Several panel members also represented consumers of the Medicare program. The panel was chaired by Stuart Altman, Dean of the Florence Heller School of Brandeis University (in De- cember 1983, Dr. Altman became chairman of the con- gressionally mandated Prospective Payment Assess- ment Commission). The first panel meeting was held on October 22, 1982. Panel members discussed the overall study plan for the assessment based on the proposal and prelimi- nary modifications and helped OTA staff refine the goals for the project. The panel examined the project boundaries and definitional issues and was key in sharpening the study's focus. The panel was also helpful in reviewing the primary issue areas to be cov- ered and in providing suggestions of individuals and organizations to contact for information and assist- ance. Case studies of four medical technologies that were specifically requested by Congress were dis- cussed, and the panel provided ideas for possible ad- ditional cases. The case study approach was intended to provide additional (e.g., efficacy, safety, and costs) information on specific medical technologies in order to analyze their possible effects on Medicare. The re- quested technical memorandum on the proposed use of Diagnosis Related Groups (DRGs) as Medicare's hospital payment method was also discussed. Following the panel meeting, contracts were let for some of the additional case studies. Drafts of the first three case studies were received by OTA staff and subsequently mailed out for the review. This process involved the advisory panel and 50 to 80 additional reviewers, depending on the case study. In addition, OTA staff prepared staff papers on the main issues of the assessment. A draft of the technical memorandum on DRGs was also prepared by the staff and sent to the panel for their review. Only 2 weeks after the first draft of the technical memorandum was completed, the Social Security Amendments of 1983 (Public Law 98-21) mandated a change in Medicare's hospital pay- ment system to a prospective system based on DRGs. A decision was made to focus the technical memoran- dum on implications for medical technology under DRG payment. Previously, the focus had been on whether a DRG system would be appropriate. Finally, during this period, the case study on alcoholism treat- ment was also completed. It was released by the Sen- ate Finance Committee, Subcommittee on Health in March 1983. The second panel meeting was held on March 22, 1983. At that meeting, progress of the study was re- viewed, and the panel explored modifications in the emerging conceptual approach of the project. Consid- erable time was spent discussing ways to analyze and synthesize the material that had been collected. The panel also provided comments on the technical memo- randum, the case studies, and OTA staff papers. In July 1983, the technical memorandum on DRGs, entitled Diagnosis Related Groups and the Medicare Program: Implications for Medical Technology, and the case study on therapeutic apheresis were completed and released by OTA. Additional case studies were received from contractors and mailed out for review. In August 1983, the case study on the variations in hos- pital lengths of stay was completed and released by OTA. The staff also prepared a first draft of the main report for the panel's review. The third and final meeting of the advisory panel was held on August 2, 1983. The primary focus of the meeting was on the draft of the final report prepared by OTA staff. The panel identified its strengths, weak- nesses, and omissions and also defined areas for devel- oping policy options for congressional consideration. The first draft of the main report was revised by OTA staff to reflect the extensive suggestions and com- ments of the advisory panel. The second draft was then sent for a further round of review by a much broader 173 174 e Medical Technology and Costs of the Medicare Program range of experts in a diversity of settings: Federal agen- cies, private and nonprofit organizations, academic in- stitutions, practicing health professionals, and other selected individuals. Altogether, more than 200 indi- viduals or organizations were asked to comment on drafts of the main report, the technical memorandum, or individual case studies of this assessment. The second draft of the main report, containing policy op- tions, was sent for review to approximately 90 in- dividuals. After appropriate revisions based on com- ments received were made, the report was submitted to the Technology Assessment Board. This project resulted in a number of documents: the main report, of which this appendix is a part; a tech- nical memorandum on DRGs; and six case studies on specific medical technologies: * The Effectiveness and Costs of Alcoholism Treat- ment: Leonard Saxe, Denise Dougherty, Katha- rine Estes, and Michelle Fine. Requested by the Senate Committee on Finance; Subcommittee on Health. ® The Safety, Efficacy, and Cost Effectiveness of Therapeutic Apheresis: John C. Langenbrunner (Office of Technology Assessment). Requested by the Senate Committee on Finance; Subcommit- tee on Health. ® Variations in Hospital Length of Stay: Their Rela- tionship to Health Outcomes: Mark R. Chassin. Requested by the Senate Committee on Finance, Subcommittee on Health. ® Intensive Care Units (ICUs): Costs, Outcomes, and Decisionmaking: Robert A. Berenson. ® Effectiveness and Costs of Continuous Ambula- tory Peritoneal Dialysis (CAPD): William B. Sta- son and Benjamin A. Barnes. Requested by the Senate Committee on Finance and its Subcommit- tee on Health. ® The Cost Effectiveness of Digital Subtraction Angiography (DSA) in the Diagnosis of Cerebro- vascular Disease: Matthew Menken, Gordon H. DeFriese, Thomas R. Oliver, and Irwin Litt. Several contractors’ reports were also prepared. The main report, the technical memorandum, and the case study on apheresis were prepared by OTA staff. The remaining case studies were commissioned by OTA, performed under contract by experts, and reviewed ex- tensively under the direction of OTA. The case studies are part of OTA’s Health Technol- ogy Case Study Series. The case study selection proc- ess involved OTA staff and consultations with the congressional staffs, the advisory panel for this assess- ment, the Health Program Advisory. Committee, and other experts in various fields. Four of the case studies were specifically requested by congressional commit- tees. The remaining two were selected to provide in- formation and ideas for the main report and to serve as individual analyses of particular issues and technol- ogies. Like this report, the case studies and the DRG technical memorandum are available through the U.S. Government Printing Office. Appendix B.—Acknowledgments and Health Program Advisory Committee The development of this assessment was greatly aided by the advice and review of a number of people in addition to the Advisory Panel and the Health Program Advisory Committee. OTA staff would like to express their appreciation especially to the following individuals: Joel Broida Health Care Financing Administration John P. Bunker Stanford University Marc Chassin Los Angeles, Calif. Dennis Cotter Prospective Payment Assessment Commission Philip Cotterill Health Care Financing Administration J. Richard Crout National Institutes of Health Helen Darling Government Research Corp., Washington, D.C. Palmer Dearing Blue Cross/Blue Shield Association Ginette Fowles Stanford University Paul Ginsburg Congressional Budget Office Mark Goodhart American Hospital Association Ruth S. Hanft Washington, D.C. Charles Helbing Health Care Financing Administration John C. Langenbrunner Health Care Financing Administration Joseph Lipscomb Duke University Bryan R. Luce Health Care Financing Administration Glenn Markus Health Policy Alternatives, Inc. Washington, D.C. John E. Marshall National Center for Health Services Research Thomas McElvogue Blue Cross of Greater Philadelphia Mary Lou McIntyre Health Care Financing Administration John Reiss Dechert, Price & Rhodes, Philadelphia, Pa. Thomas Rice University of North Carolina Sheila Ryan Health Care Financing Administration Leonard Saxe Boston University Marvin Shapiro U.S. Administrators Inc., Los Angeles, Calif. Bruce C. Vladeck United Hospital Fund of New York Irwin Wolkstein Health Policy Alternatives, Inc. Washington, D.C. Craig Wright Health Care Financing Administration James M. Young Blue Cross and Blue Shield of Massachusetts HEALTH PROGRAM ADVISORY COMMITTEE Sidney S. Lee, Committee Chair President, Michael Reese Hospital and Medical Center Stuart H. Altman* Dean, Florence Heller School Brandeis University H. David Banta Deputy Director Pan American Health Organization Carroll L. Estes** Chair Department of Social and Behavioral Sciences School of Nursing University of California, San Francisco 175 176 e Medical Technology and Costs of the Medicare Program Robert Evans Professor Department of Economics University of British Columbia Rashi Fein Professor Department of Social Medicine and Health Policy Harvard Medical School Harvey V. Fineberg Dean, School of Public Health Harvard University Melvin A. Glasser*** Director Health Security Action Council Patricia King Professor Georgetown Law Center Joyce C. Lashof Dean, School of Public Health University of California, Berkeley Margaret Mahoney **** President The Commonwealth Fund New York, N.Y. Frederick Mosteller Professor and Chair Department of Health Policy and Management School of Public Health Harvard University *Until April 1983. **Until March 1984. ***Until October 1983. ****Until August 1983. Nora Piore Senior Advisor The Commonwealth Fund New York, N.Y. Mitchell Rabkin* President Beth Israel Hospital Boston, Mass. Dorothy P. Rice Regents Lecturer Department of Social and Behavioral Sciences School of Nursing University of California, San Francisco Richard K. Riegelman Associate Professor School of Medicine George Washington University Walter L. Robb Vice President and General Manager Medical Systems Operations General Electric Milwaukee, Wis. Frederick C. Robbins President Institute of Medicine Rosemary Stevens Professor Department of History and Sociology of Science University of Pennsylvania Appendix C.—Selected Activities in Medical Technology Assessment Introduction Medical technology assessment in health policy today typically refers to an evaluation of the efficacy and safety, and sometimes costs, of medical technol- ogy. Medical technology assessments are a source of information needed by government officials in formu- lating legislation and regulations, by health profes- sionals in managing patients, by industry in develop- ing products, by private insurers in creating benefit packages, and by consumers in making personal health decisions (359). Furthermore, medical technology assessments can yield information for hospitals func- tioning under Medicare's new Diagnosis Related Group (DRG) hospital payment system.! Currently, there are numerous activities in medical technology assessment. Multiple participants from both the public and private sector perform or use med- ical technology assessments (359). However, there is little, if any, coordination among the various organi- zations, and because of a lack of funds, many organi- zations neither assess as many technologies nor per- form as comprehensive assessments as they desire. Assessments are used for different purposes, including coverage, payment, purchasing, and management de- cisions. The focus of this appendix is on assessments made or used for coverage purposes. In the past 2 years, the private and nonprofit sec- tors have increased their involvement in assessing med- ical technology. However, many of their assessments are limited to specific organizational objectives and have limited value for national policy decisions. Safety and efficacy criteria are usually used in technology assessments; economic, legal, social, and ethical cri- teria are sometimes used. Public Sector Activities OTA has estimated that Federal expenditures on evaluating health technologies in general are approx- imately $200 million a year (361). Only a small frac- tion of this amount is spent for evaluating medical technologies specifically to determine their eligibility for Medicare payment. By way of contrast, it is in- teresting to note that fiscal year 1982 reimbursement for Medicare services totaled $50.9 billion (135). Currently, the body with explicit responsibility for evaluating selected medical technologies to assist the For more information on DRGs, see OTA's July 1983 technical memo- randum Diagnosis Related Groups (DRGs) and the Medicare Program: Im- plications for Medical Technology (343). Health Care Financing Administration (HCFA) in de- termining what diagnostic and therapeutic techniques ought to be covered by Medicare is the Office of Health Technology Assessment (OHTA), of the Na- tional Center for Health Services Research in the Public Health Service (PHS). OHTA's budget for fiscal year 1983 was approximately $1 million. The Medicare program has called on PHS to pro- vide technical medical advice for making coverage decisions since the late 1960's. The coverage advice process established initially was a loosely structured one, relying mainly on informal contacts with experts at the National Institutes of Health (NIH) or medical specialty societies for opinions about the safety and effectiveness of medical technologies. In 1977, the Administrator of HCFA and the Assistant Secretary for Health of the Department of Health and Human Services (DHHS) formalized the PHS role in providing advice through the Office of the Assistant Secretary for Health. In recent years, the increasingly rapid development and use of sophisticated and expensive medical tech- nologies has increased the number and complexity of Medicare coverage determinations. Thus, Congress passed the National Health Services Research, Health Statistics, and Health Care Technology Act of 1978 (Public Law 95-623). That act established the National Center for Health Care Technology (NCHCT). One of NCHCT's mandated assignments was to pro- vide scientific/ medical assessments to HCFA on Medi- care coverage for specific medical procedures and tech- nologies. The agency's overall mission, however, was much broader: it was to “stimulate increased scrutiny of new and existing health care technologies to insure that their safety, efficacy, cost-effectiveness, social, ethical and economic impacts are more completely ex- plored” and to encourage the “rapid dissemination of newly developed health care technologies which have proved their worth in terms of safety, efficacy, (and) cost-effectiveness.” NCHCT's staff was officially lim- ited to never more than 20, but “creative management” by its director, Dr. Seymour Perry, enabled the cen- ter to obtain the services of 39 individuals (45). In De- cember 1981, however, NCHCT ceased to function be- cause of a lack of congressional funding. OHTA, formed as NCHCT's successor, has been assigned a variety of duties (see table C-1). Generally, however, its activities have been confined to evaluating Despite the similarities in their names, the Office of Health Technology Assessment (OHTA) and the congressional Office of Technology Assessment (OTA) are two different offices. 177 178 * Medical Technology and Costs of the Medicare Program Table C-1.—Activities of the Office of Health Technology Assessment (OHTA) * Provide national leadership, coordination, and administration of a comprehensive program for health care technology assessment and transfer to improve the quality and reduce the cost of medical care * Establish criteria for public and private organizations and individuals both within and outside OHTA to identify the critical technologies to be assessed * Administer a program of assessments of health care technologies which take into account their safety; efficacy; cost effectiveness; and social, ethical, and economic impacts * Make recommendations on health care technology issues in the administration of the laws under the Assistant Secretary for Health's jurisdiction, including preparation of the PHS position regarding appropriateness of Medicare coverage of health care technology * Publish and disseminate the information obtained as a result of activities supported by OHTA, and undertake programs to develop new and improved methods for making such information available ¢ Provide technical assistance and consultation to organizations and individuals within and outside DHHS engaged in or concerned with the results of health care technology assessments, research, evaluations, and demonstrations * Coordinate PHS research, evaluations, and demonstrations relating to the assessment of health care technology undertaken and supported by DHHS components. SOURCE: Federal Register 48(13):2444, Jan. 19, 1983. technologies in response to requests from HCFA. When HCFA has a simple inquiry about the regula- tory and research standing of a particular technology, the staff of OHTA provides background information obtained from other PHS agencies, including the Food and Drug Administration (FDA), NIH, the Centers for Disease Control (CDC), and the Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA). HCFA may also request full assessments from OHTA. OHTA's assessment process, which involves synthesiz- ing the available information about a technology and transmitting the results to HCFA, requires 12 to 18 months for completion. As of June 1983, OHTA had a backlog of 12 months or more on 23 technologies requiring assessment—some from 1979 (211). OHTA completed 26 assessments in 1982. The OHTA Direc- tor reported that as of June 1983, 16 assessments had been completed, 30 assessments were in progress, and 40 were expected to be completed by the end of the 1983 fiscal year (211). In performing a full assessment, the OHTA staff first reviews the coverage issue with HCFA, clarifies the original request, and defines appropriate questions. It also initiates a literature search and data collection ef- fort, gathering information from a wide spectrum of sources. Working under a philosophy that it should hear from all interested parties, OHTA advertises an impending assessment in the Federal Register. Al- though notices and advertisements regarding partic- ular coverage issues also appear in professional and trade publications, notices in the Federal Register are PHS’s primary access to the “general public” and are used to obtain the views of a broad mix of interested parties. Responses to such notices have generally come from industry and physicians. For available scientific information, clinical trial data, bibliographic materials, and other relevant ma- terials, OHTA solicits the advice and assistance of Fed- eral agencies such as NIH, FDA, ADAMHA, and CDC. Each agency has developed a formal plan for responding to such evaluations. Furthermore, PHS routinely announces the assessment through DHHS's Technology Coordinating Committee, which informs other interested Federal agencies. OHTA also contacts the Council on Medical Specialty Societies, the orga- nization representing all medical specialty societies, as well as the relevant specialty societies for information on the specific technology (or in some cases, a list of technologies) being considered. The OHTA staff then analyzes and synthesizes the medical and scientific evidence and professional opin- ions collected. OHTA's assessment, primarily of the safety, efficacy, and clinical effectiveness, of the tech- nology in question is conducted according to specific criteria (399). The types of acceptable information range from qualified medical opinions derived from personal experience, to well-designed clinical studies, to controlled clinical trials. Although OHTA's guide- lines emphasize the value of controlled clinical trials, few evaluations have had the benefit of such rigorous evidence (352). Like its predecessor NCHCT, OHTA provides rec- ommendations to HCFA about the appropriateness of providing Medicare coverage for a technology which it has assessed. OHTA does not release its assessment until HCFA has taken action on the recommendation. The assessments, but not the recommendations, are published and disseminated. The formation of the Prospective Payment Assess- ment Commission (ProPAC), mandated with the pas- sage of the Social Security Amendments of 1983 (Public Law 98-21), initiates a new Government in- volvement in medical technology assessment for pay- ment purposes. ProPAC members were appointed in November 1983, and the body's activities started in the early part of 1984. ProPAC is an independent advi- sory committee that is empowered to collect and assess information on hospital costs and productivity, tech- nological advances, and the cost effectiveness of hos- pital services. It is required to use existing informa- App. C—Selected Activities in Medical Technology Assessment 179 tion where possible but is also given authority to carry out research and to award grants and contracts. Pro- PAC will make recommendations to the Secretary of DHHS concerning the appropriate payment rate under Medicare's DRG payment system for inpatient hospi- tal services. ProPAC also has broad powers to assess medical technology and the appropriateness of medi- cal practice patterns. Private Sector Activities In the private sector, the Blue Cross and Blue Shield (BC/BS) associations often rely on evaluations of med- ical technology in arriving at coverage and reimburse- ment decisions. Like Medicare, the BC/BS associations make a majority of coverage decisions through a de- centralized and loosely structured process that places key emphasis on two criteria used in Medicare's cov- erage policy: stage of development and acceptance by the medical community. Some BC/BS member plans use informal methods to evaluate medical technologies. The medical advisors of individual plans may make coverage determinations based on immediate personal knowledge or may sur- vey the literature and consult with advocates of the procedure, local specialty societies, the county medical society, and other area insurers. If the plan functions as a contractor for Medicare, as an agent for Medic- aid, or as an agent for the armed services CHAMPUS program, the medical advisor will generally review the coverage determinations of these programs. Similarly, the medical advisor will usually study the coverage rec- ommendations of the national BC/BS Association and other member plans, although he or she may not fol- low any previous rulings (353). On the other hand, there are BC/BS plans that fol- low an extremely formal procedure for assessment. A particularly active State affiliate, Blue Shield of California, established a Medical Policy Committee in 1966 to assess the scope and limits of Blue Shield’s standard medical insurance policies with respect to new diagnostic and therapeutic procedures (52). This committee has continued its interest in evaluating tech- nologies in the interest of cost containment. Blue Shield of Massachusetts is actively involved in technology assessment, as well. A separate commit- tee, the Interspecialty Medical Advisory Committee, is incorporated into its decisionmaking structure. One of the advisory committee's main duties is to make rec- ommendations concerning medical, surgical, or thera- peutic procedures, i.e., to determine the “general ac- ceptability” of a new service or procedure. Only those services or procedures deemed “generally accepted” by Blue Shield are reimbursed. The “generally accepted” guidelines promulgated by Blue Shield of Massachu- setts are very similar to those used by OHTA in per- forming its assessments for coverage purposes (see ch. 5). According to the Blue Shield guidelines (41): . . a generally accepted surgical or medical procedure is a procedure which: a) has completed the research or experimental stage of development; b) does not involve, as an integral part of the proce- dure, the use of any drug, substance or device which has not been released by the Food and Drug Admin- istration or other governmental licensing agencies for general use by qualified physicians; ¢) is in general use for patient care by physicians qualified to perform the procedure; d) is of demonstrated value for the diagnosis or treat- ment of an illness, injury, or bodily function. There are very specific definitions and clarifications for the administration of the Massachusetts Blue Shield guidelines, including a requirement for factual mate- rial such as the rates of mortality and morbidity asso- ciated with the technology. A health care provider (or group of providers) contacts his or her specialty society committee concerning coverage of a new technology. There are 30 such committees. The specialty society committee then assesses the surgical or medical pro- cedure according to the guidelines just described. If the new technology is approved, the recommendation with all necessary supporting information is presented to Blue Shields Interspecialty Medical Advisory Com- mittee, which then decides on a fair reimbursement level for the service/procedure. After going through additional steps in the organizational structure, the rec- ommendation is formalized. If coverage is denied, there is opportunity for a request for reconsideration. The assessment performed by Blue Shield of Mas- sachusetts differs from that performed by OHTA in a number of ways. One important way is that atten- tion is paid in the Blue Shield evaluation to other estab- lished procedures that are equal to the procedure under consideration in merit or effectiveness as well as in their morbidity and relative cost. Another important consideration in performing a Blue Shield evaluation is the cost of the prosthesis/device/drug/substance which is an integral part of the proposed procedure. The coverage recommendations of the national BC/BS Association are made using two mechanisms: the Medical Advisory Subcommittee and the Medical Necessity Program. The Medical Advisory Subcom- mittee, composed of six medical directors from: in- dividual affiliates, meets four times a year to make coverage decisions. Requests for coverage determina- tions may originate either directly from a manufac- turer or provider or indirectly from one of the local plans. After the national BC/BS staff conducts re- search on the product or service, which may include 180 e Medical Technology and Costs of the Medicare Program collecting the opinions of professional medical asso- ciations, the Medical Advisory Subcommittee either reaches a coverage decision or postpones a decision until further information on the product or service is available for review (141). Cost considerations are more likely to be explicitly included in the evaluation of a technology by the Med- ical Advisory Subcommittee than they are in evalua- tions for Medicare. During the initial review for the Medical Advisory Subcommittee, the BC/BS staff gathers information about the reimbursement level for the product or services, or similar product or services, where possible. Furthermore, when a decision is reached, it is reviewed by the Providers Affairs and Cost Containment Review Committees and then by the National Blue Cross and Blue Shield Board Review. The coverage determinations, which go out as a rec- ommendation to the 106 local U.S. BC/BS plans, usu- ally concern new technology and are expressed in broad terms. However, sometimes they specify indica- tions for use. The Medical Necessity Program was developed in 1977 in conjunction with the American College of Phy- sicians (ACP), the American College of Surgeons (ACS), and the American College of Radiology. It was designed to curtail reimbursement for obsolete, dupli- cated, and outmoded procedures (41). In 1977, the Medical Necessity Program advised BC/BS plans to discontinue routine payment for a group of 42 proce- dures unless physicians provided special medical justification for their use. Since 1977, the original list of 42 procedures has been expanded to nearly 100 (41). In 1979, on the basis of the recommendations of ACP and ACS, the Medical Necessity Program advised BC/BS plans to pay for diagnostic tests performed at the time of admission to hospitals only when the tests had been specifically ordered by a physician. The stated objectives of the Medical Necessity Pro- gram—cost containment and the improvement of qual- ity of care—remain the same as in 1977. The focus of the program, though, was expanded in 1980 from eval- uating possibly obsolete procedures to examining pro- cedures that may be overutilized or inappropriately utilized. In October 1982, the BC/BS national asso- ciations recommended new guidelines to member plans “intended to raise the level of cost consciousness” about the unnecessary use of certain respiratory care procedures, with expectations of potential savings of hundreds of millions of dollars annually (326). In arriving at its recommendations, the Medical Ne- cessity Program draws on the advice of national med- ical specialty organizations including ACP. Indeed, ACP’s involvement in evaluating the safety, efficacy, and effectiveness of clinical tests, procedures, and ther- apies began with its participation in the project (230). Commercial insurance companies make independ- ent coverage decisions regarding the coverage of new and emerging technologies. When a coverage question arises, a company may contact the Health Insurance Association of America (HIAA). HIAA is a trade asso- ciation of 338 commercial insurance companies. It will provide information to its member companies on the medical appropriateness of diagnostic and therapeu- tic procedures. HIAA does not conduct technology assessments, but reports opinions rendered by the Council of Medical Specialty Societies (CMSS) through its program for clinical procedure review (193). In late 1977, CMSS agreed to accept questions from insurers with HIAA acting as the intermediary between the many insurance companies and CMSS. HIAA will transmit the information to the requesting insurer and publish the CMSS opinion in its “Medical Relations” bulletin. Each company makes its own cov- erage policy decision using its discretion. Some prepaid group practices have also had experi- ence with medical technology assessment. The Depart- ment of Medical Methods Research of the Kaiser-Per- - manente Medical Care Program (KPMCP) of Northern California has conducted research on the utilization of modern technology for the development of im- proved methods of providing and delivering medical care within the KPMCP (72). The primary purpose of the department's Division of Technology Assessment is to aid in the selection of the most cost-effective tech- nology. The process of assessment the division uses is quite different from that employed elsewhere, pri- marily because of the unique financial structure of prepaid group practices. In a prepaid group practice, an increase in the use of a technology often increases expenses and does not generate revenues as it might in fee-for-service or cost-reimbursement programs. In addition, there are no savings from the purchases of equipment that could increase cash flow. Thus, the in- centives are for low-cost technology that maintains or improves the effectiveness of medical care. Assessments at the KPMCP begin with the identifi- cation of a technology that uses substantial resources. Next, the characteristics of the population utilizing the technology and the workloads for its utilization are determined. Alternative technologies used for the same specified objectives are evaluated as to important in- tended and unintended consequences. The technology assessments use epidemiological methods, controlled studies, medical record studies, literature reviews, con- sensus development, and sensitivity analyses. The results of the assessments are presented as important consequences of alternative technologies so that man- agement can make more rational decisions. A number of medical specialty societies are inter- ested in technology assessment. ACP is now conduct- App. C—Selected Activities in Medical Technology Assessment e 181 ing the Clinical Efficacy Assessment Project (CEAP), a major effort to study the efficacy, clinical effective- ness, and safety of tests, procedures, and therapeutic interventions. CEAP evaluated 50 procedures and tests from January 1981 through November 1983 and as of March 1984 had 11 other evaluations in process. Al- most all the technologies that have been evaluated to date were submitted by the Federal Government or in- surance companies who generally use the result in their coverage and reimbursement decisions (21). The steps in ACP’s evaluation process are similar to those used by OHTA. CEAP draws on its member- ship for information on the technology under evalua- tion, and after developing a draft statement, sends the statement to outside experts for review. The final state- ment is written by staff with the help of ACP’s Clini- cal Efficacy Subcommittee. In addition to ACP, other medical specialty societies and medical associations have either initiated or in- tensified existing activities involved with medical tech- nology assessment during the past few years. Among these are the American College of Cardiology (ACC), the American Medical Association (AMA), and the American Hospital Association (AHA).* Although the assessment activities of these organizations are not as directly linked with coverage and reimbursement deci- sions as ACP’s have been, their findings are often used by public and private insurers. ACC has three activities related to technology as- sessment. First, it responds to requests from Federal agencies as well as from the private sector (e.g., hos- pitals, clinics, third-party carriers) about the standards, criteria, and appropriateness of procedures usually per- formed in a hospital setting by physicians treating car- diovascular diseases. Second, it has a Joint Task Force in conjunction with the American Heart Association that undertakes in-depth technology assessments. The assessments include criteria of contribution unique- ness, sensitivity, specificity, indications and contra- indications, and cost effectiveness. The third activity of ACC is new and under devel- opment. A Cardiovascular Norms Committee is be- ing formed, to “establish and obtain consensus on dynamic norms (defined as factors essential for quality care decisionmaking) for the diagnoses and manage- ment of the most common cardiac diagnoses, including cost effectiveness of alternate plans or diagnostic tech- niques.” The first step of this ambitious undertaking is to investigate a mechanism for developing dynamic norms (a type of criteria setting) (196). *The American Academy of Pediatrics, the American Public Health Asso- ciation, the American College of Radiology are among the other professional associations used that have taken part in evaluating technologies. Their activ- ities are discussed in the OTA report Strategies for Medical Technology Assessment (359). 25-337 0 - 84 - 13 AMA has been involved in evaluating and providing information of technologies through its scientific pub- lications; through a series of reports published by its Council of Scientific Affairs dealing with diagnostic, therapeutic, and other medical technologies; and by responding to thousands of inquiries that involve in- formation on assessment of medical technologies, par- ticularly those that are well established. AMA just began a new project, Diagnostic and Therapeutic Technology Assessment (DATTA), whose purpose is to “expeditiously and effectively examine medical tech- nologies that are passing from experimental or investi- gational to accepted forms of treatment and define, where possible, indication for their use.” It is intended that the assessments will be limited to the evaluation of the safety and effectiveness of a technology (196). Other private sector parties with a longstanding in- volvement in medical technology assessment, espe- cially to determine safety and efficacy, are manufac- turers of drugs and devices. They initiate research and are required by FDA to conduct tests for premarket approval of their products. Large private clinics, e.g., the Cleveland and Mayo Clinics, also perform some assessments? (359). Recently, AHA and other members of the hospital community have become actively involved in evaluat- ing technologies, but mainly from the perspective of planning for new health care services and in review- ing existing services. AHA has played a catalytic role in this authority by issuing the'manual, Technology Evaluation and Acquisition Methods for Hospitals (TEAM), in 1979 (13). The program described in the manual not only provides a mechanism for hospitals to evaluate technologies for financial reasons but also emphasizes evaluation with respect to community needs and hospitals’ role in the community. AHA has continued to involve its members in re- cent developments in technology and technology as- sessment through its Hospital Technology Series Guidelines Reports, published as part of the Hospital Technology Series, which are individual reports de- voted mainly to specific technologies. They usually ex- amine the key factors a hospital should include when evaluating a particular technology for purchase, as well as product information on commercially avail- able models of the technology. In addition, hospitals and hospital chains, such as Humana and the Hospital Corporation of America, are examining technologies more carefully than before. However, when they evaluate technology, it is most often with respect to an “overall risk management pro- gram for the identification, evaluation, and treatment See Strategies for Medical Technology Assessment for a description of their activities (359). 182 e Medical Technology and Costs of the Medicare Program of risk or financial loss” (292). The Hospital Corpora- tion of America, for example, recommends the forma- tion of a Product Selection Committee by each of its member hospitals. The committees purpose is to eval- uate the degree of inherent risk to each patient and/or employee posed by material that may be introduced into or is already being used within the hospital (171). Prepaid group practices, such as Kaiser-Permanente of Northern California, conduct technology assess- ments in order to improve their methods of providing and delivering medical care within their organization. Finally, ECRI (formerly the Emergency Care Re- search Institute) is a nonprofit organization primarily involved in comparative product evaluations of diag- nostic and therapeutic devices and hospital equipment and supplies. ECRI provides a type of “consumer re- port” service for hospital administrators that gives ratings to comparable medical technologies based on performance safety, ease of use, and cost effectiveness. An emphasis on the larger economic, social and ethical issues surrounding health care technologies has recent- ly been added. Further, ECRI maintains a computer- ized health devices data base on over 6,000 categories of devices and hospital equipment (28). Public/Private Sector Activities There are some indications of cooperation between the public and private sector and among members of the private sector in evaluating medical technologies. Massachusetts Blue Shield, for example, has been send- ing its Interspecialty Medical Advisory Committee's monthly agenda to HCFA's Office of Coverage Pol- icy for 2 years (436). This mechanism informs the Of- fice of Coverage Policy of current issues pertaining to technology assessments considered in Massachusetts by Massachusetts Blue Shield's Interspecialty Medicare Advisory Committee. The monthly agendas are also exchanged with the Blue Shield of New Jersey's Medi- cal Advisory Committee and with the National Blue Cross and Blue Shield Medical Advisory Committee and their staff. Furthermore, the national BC/BS Asso- ciation has begun a comprehensive medical policy manual that will be the basis for a uniform medical policy for all of the Blues nationwide (436). The activities just mentioned are but a few of the numerous efforts underway to evaluate medical tech- nologies. The proliferation of medical technologies and the absence of an organization to coordinate and com- plement existing technology assessment activities pro- mpted the Institute of Medicine (IOM) to appoint a committee to develop a plan for a technology assess- ment organization that would be based in the private sector and supported by both governmental and non- governmental funds (234). The IOM report recom- mended the establishment of a medical technology consortium that would function under the auspices of IOM during an initial period of development, and then, after approximately 5 years, function as an in- dependent entity in the private sector. The functions of the medical technology consortium would be as follows: ® to serve as a clearinghouse of information on medical technologies and medical technology assessment; e to assemble and evaluate information and make recommendations concerning individual medical technologies; ® to act when necessary and appropriate to stimu- late, coordinate, undertake, or commission med- ical technology assessments, including activities that would complement those of others; e to identify needs in the assessment of specific med- ical technologies; e to develop and evaluate assessment criteria and methods; and e to provide education, training, and technical as- sistance in the use of medical technology assess- ment methods and results. The IOM report noted that the “consortium is not intended as a competitor or replacement for an existing entity involved in assessing medical technologies.” Rather, it is to be complementary and facilitative of the efforts of others involved in responsible assessments of medical technologies. The report recommended that when the consortium first starts to function, initial emphasis should be placed on the clearinghouse func- tion because of expected financial constraints. A pro- posal for the creation of the consortium awaits fund- ing (169). Appendix D.—Selected Alternatives to Traditional Health Care Delivery Introduction Strategies for containing Medicare costs by chang- ing the incentives for the adoption and use of medical technology were identified in chapters 5 through 8. Chapter 8 identified examples of alternatives to tradi- tional health care delivery sites and organizations! that might stimulate competitive behavior by health care providers. The purpose of this appendix is to provide additional information on those alternatives. Avail- able data on utilization by Medicare patients and costs to the program of these alternatives have been included in this appendix to illustrate potential data and evalua- tion needs. Assuming that a current goal of the Medi- care program is to contain or reduce costs without compromising quality of care, it is reasonable to ex- amine alternative sites and organizations of health care delivery for their potential in helping attain that goal. Development of Alternative Sites and Organizations for Health Care Delivery The precise reasons for the creation of alternative sites and organizations for health care delivery are dif- ficult, if not impossible, to specify, although the causes of the proliferation of alternative sites and organiza- tions for health care delivery clearly include economic and social forces. As the prices of medical care have soared, coverage of services by public programs such as Medicare and Medicaid and expanded coverage of services by private insurance companies have enabled a greater number of persons to obtain needed health services (135). Entrepreneurs responsive to financial incentives throughout the health care system have joined physicians and other health care providers in developing alternative sites and organizations of health care delivery. Many of the alternatives have developed as for-profit operations (325), and they have fostered competition in local markets. Freestanding ambulatory surgery centers and health maintenance organizations (HMOs), for example, developed under the influence of specific market conditions (e.g., overscheduled hos- pital operating rooms in the case of the centers and increased supply of physicians and Federal encourage- ment in the case of HMOs). Other alternatives have been developed to provide services at more convenient times and in more convenient locations for patients. Prominent examples of “traditional” sites or organizations are the independent community hospital, the solo, fee-for-service phy- sician, and use of only hospitals for surgery. Emergency care centers are an example. Medicare itself has had varying impacts on the es- tablishment of alternatives. Home health care agen- cies have proliferated since Medicare began covering home visits in 1966. Ambulatory surgery centers, on the other hand, developed without Medicare cover- age, and although freestanding centers have received specific coverage since September 1982, the impact of the new coverage is unknown. Emergency care centers generally do not encourage Medicare patients to use them, and some even exclude Medicare patients ex- cept in life-threatening situations. The Medicare pro- gram was used to directly encourage the development of HMOs. A major question with respect to alternative sites and organizations of health care delivery is whether the alternatives are less costly to patients, to the Medi- care program, or to both. Some alternative sites (e.g., ambulatory surgery centers) may substitute for tradi- tional hospital and physician office sites for some pa- tients, while others (e.g., emergency care centers) may complement them. Covered sites of care may or may not save money, depending on the extent and appro- priateness of use. As long as lower cost health serv- ices are substituted for more expensive ones, Medicare will save. If instead, the alternatives are used in addi- tion to more expensive, traditional care, Medicare costs will rise. Potential cost savings for the Medicare program depend on the number of beneficiaries using any service, on the prices paid for medical technol- ogies in various sites, and on the quality of the care produced. Quality is important in its own right. It is also important in economic terms, though, because lack of it may cost the beneficiaries and the program more in the long run. Measurements of quality of care for alternative sites are controversial. For those sites of care and services specifically covered by Medicare, there are conditions of participation that providers must meet in order to receive payment. The conditions for freestanding am- bulatory surgery centers, for example, specify that such centers must meet State licensing requirements and obtain accreditation by an appropriate associa- tion. Medicare conditions of participation have been criticized for concentrating on the structure and proc- ess aspects of quality measures (176) instead of on pa- tient outcomes. In home health care, the Medicare con- ditions of participation have been important in developing quality assurance programs (176). Costs of alternative sites for health care delivery are not easily compared. There are no published studies 183 184 e Medical Technology and Costs of the Medicare Program that account for differences in case mix and services rendered by freestanding emergency care centers, hos- pital emergency rooms or outpatient departments, or physicians’ offices (2,325). One study on ambulatory care found slight differences in case mix between hos- pital outpatient departments and private physician of- fice practices (200). Measuring differences in costs of various sites is also difficult, because hospitals, for ex- ample, do not use consistent methods of reporting costs, and, thus, do not accurately measure the true cost of ambulatory care (2). Measuring utilization of the alternative sites also presents problems. The cus- tomary measure is the visit, but differences in case mix, use of tests and procedures, and standby equipment and staff are not adequately accounted for by the visit measure (2). The incentives for provider behavior will be changed by Medicare's new prospective per case payment sys- tem for inpatient hospital services (see ch. 6). Under the payment system based on Diagnosis Related Groups (DRGs), hospitals have financial incentives to decrease lengths of stay and substitute outpatient serv- ices for inpatient services. Increased use of outpatient visits might help prevent use of inpatient care, but the research evidence is mixed (33). Early discharges of hospitalized patients will probably increase the need for care in skilled nursing facilities (SNFs) and for home health care. Hospice care may also be affected by the DRG payment system, although in the absence of pre- vious Medicare experience with hospices, interpreta- tion of the evidence will be difficult. The discussion below provides information on selected alternatives to traditional health care deliv- ery. The first section discusses alternative sites of care, i.e., alternatives to inpatient care in hospitals and to primary care in physicians’ offices. The second section describes two alternative organizations for health care delivery, HMOs and preferred provider organizations (PPOs), both of which may increase competition in the health field. Alternative Sites of Health Care Delivery Patients can obtain different types of medical care in a variety of locations. The sites described here are alternatives to inpatient care in hospitals and to pri- mary care in physicians’ offices. The discussion of each alternative site includes patterns of use, evidence on cost and quality of care, and effects of Medicare pol- icy. Unfortunately, data vary in availability and qual- ity, so comparisons are not always possible. Alternatives to Inpatient Hospital Care Entrepreneurs have reacted to the availability of money in the health care system by providing alter- native sites for hospital care. In some cases, medical technologies are being moved from their traditional sites in hospitals to other locations. Certain surgical procedures have been moved from traditional hospi- tal sites to ambulatory surgery centers. Hospital in- patient care for acute illness is being complemented, and, in some cases, replaced by home health and nurs- ing home care. Palliative care for terminally ill patients has also been moved out of hospitals to hospices. These alternatives are described below. Ambulatory Surgery Centers.—Units to accommo- date ambulatory surgery were developed in the early 1970's in response to overcrowded operating room schedules and inconvenience to patients and physicians (125). Ambulatory surgery centers could not have been established without the technological improvement of fast-acting anesthesia and the practice of making pa- tients walk soon after surgery (125). Some units are affiliated with hospitals and are located either in the hospitals or at other sites. Other units are not associ- ated with hospitals. These units, known as freestand- ing ambulatory surgery centers, are often physician- owned. Surgical procedures that are appropriate to ambulatory surgery centers are those using general anesthesia but requiring only a few hours of postop- erative monitoring of the patient. Patients are carefully screened. In recent years, third-party payers have ac- cepted claims for surgery performed in these centers, and some now require that certain procedures be done on an ambulatory basis for coverage. On September 7, 1982, Medicare changed its cover- age of ambulatory surgery to encourage more patients and surgeons to use the less expensive freestanding am- bulatory surgery centers (108). The purpose was to in- crease substitution of ambulatory surgery for inpatient surgery. The utilization of ambulatory surgery centers since that change is unknown. Yet there was sufficient concern about the possibility of adding to the surgi- cal rate for Medicare beneficiaries that the General Accounting Office was requested to study utilization patterns in the first year of the new coverage policy (177). Prior to the policy change, there was diversity ir the age distribution of patients among freestanding am- bulatory surgery centers, but there is no single source of reliable aggregate numbers of ambulatory surgica. procedures by age group (325). From 1973 to 1980, surgical rates rose even more rapidly for the elderly App. D—Selected Alternatives to Traditional Health Care Delivery ¢ 185 than for the general population in the United States (5). In 1967, 85.4 per 1,000 Medicare beneficiaries had surgery during their hospitalizations; but in 1972, the rate was 93.4; and in 1976, it was 104.9 surgeries per 1,000 beneficiaries (162). Again, improvements in anesthesia and in some procedures have reduced the risks of some types of surgery. More surgery is re- quired, also, as people age and suffer from cataracts, cancer, and cardiovascular diseases. Freestanding ambulatory surgery centers generally charge less than hospital inpatient or hospital out- patient surgery for most procedures. Clearly, inpatient surgery is more expensive, because patients must pay for a hospital stay of at least 1 day. In addition, freestanding ambulatory centers can charge lower prices because their construction costs are lower, they do not need some of the most expensive technologies available in hospital operating rooms, and their over- head is lower since many of the hospital ancillary serv- ices (e.g., food services) are not needed (124,428). It is likely that these fixed costs of the hospital setting will be spread among fewer surgical procedures, which will result in higher costs for hospital surgery. Under Medicare, freestanding ambulatory surgery facility costs are covered for 100 specific procedures (an additional list of 50 has been proposed) in four groups to encourage use of the centers. Facility fees are prospectively set by the Health Care Financing Administration (HCFA), and the schedule is based on complexity of procedure with no beneficiary deducti- ble or copayment. The least complex and least costly procedures (e.g., gastroscopy) are in Group 1, for which Medicare pays the center $231; the most com- plex and costly ambulatory surgical procedures (e.g., laparoscopy) are in Group 4, for which facilities are paid $336 (108). Physicians who accept assignment are paid 100 percent of reasonable charges for covered services. By statute, Medicare expenditures for am- bulatory surgery must be less than they would be if the procedures were performed on an inpatient basis. HCFA estimated that this benefit would save $2 mil- lion for Medicare in its first year of operation. Since hospital outpatient surgery is usually reim- bursed 80 percent of costs with a 20-percent copay- ment, some hospitals with ambulatory surgery units might be at a disadvantage with regard to Medicare payment unless they elect to participate as though their units are freestanding ambulatory surgery centers, i.e., accept the prospective fee schedule rather than cost- based reimbursement. To participate as freestanding units, they must establish centers that are physically, administratively, and financially independent from the rest of the hospital (108). Quality of care in ambulatory surgery centers ap- pears to be good. The centers have reported no deaths and lower complication rates than inpatient surgery (29,82,248). There is an accreditation association to advance their credibility to the public. In addition, centers that want to participate in Medicare must meet Medicare conditions of participation requiring that cer- tain staff and equipment be available. Home Health Care.— Continued growth in the home health industry is expected in response to the incen- tives under Medicare's DRG hospital payment system for shorter inpatient stays. The number of agencies providing home health care services has greatly in- creased since 1966 when Medicare began covering skilled nursing care and physical and speech therapy to homebound elderly people. The purposes of provid- ing those services was to lower the hospital length of stay for acutely ill patients, thus cutting costs to the program. The specific aspects of home health care have changed over time. Currently, the basic services are part-time or intermittent nursing care by or under the supervision of a registered nurse; physical, occupa- tional, or speech therapy; medical social services; and part-time or intermittent services from a home health aide. Certain medical technologies that used to be ad- ministered only on an inpatient basis (e.g., intravenous antibiotic therapy) are now part of home health care (248). Home health agencies may be licensed by the States, although the licensing requirement for Medicare par- ticipation was eliminated in 1980 with the passage of the Omnibus Reconciliation Act (Public Law 96-499) (207). Visiting nurses associations were among the first home health agencies (223). Other home health care providers are public health departments, hospitals, and independent agencies, both for-profit and not-for- profit. The hospital-affiliated home health agencies alone almost doubled between 1979 and 1982, when they numbered 450 and more than 720 respectively (207). The number of home health care visits paid for by Medicare almost tripled between 1969 and 1980 (from 8.5 million to 22.4 million) (392). Studies of home health care in the 1970's seemed to indicate that home care made early discharges from hospitals possible. Re- cent studies looked at overall hospital use, but not readmission rates or length of stay (333), so the long- term effect of the early discharges and substitution of home care is not evident. Home health care as a substitute for an extended hospital stay may be underutilized, although it appears to be used more often than SNFs by Medicare patients 186 Medical Technology and Costs of the Medicare Program (391). In 1980, home health agency services were used by 890,400 elderly and 67,000 disabled Medicare ben- eficiaries. Medicare reimbursed the agencies provid- ing these services a total of $662.1 million (392). A General Accounting Office study of home health care demonstration projects showed mixed effects of ex- panded home health care on Medicare costs (333). Sev- eral demonstration projects have studied the effects of expanded home health care on patient outcomes: their results are also mixed (333). The Omnibus Reconciliation Act of 1980 (Public Law 96-499) changed Medicare's home health care cov- erage by eliminating the remaining copayments, the limit on number of visits, and the hospitalization re- quirement. Patients must be homebound, under the care of a physician, and in need of skilled nursing or physical or speech therapy. Medicare Part A covers all home health visits unless a beneficiary has Part B coverage only. In the latter case, Part B covers the home health services (391). Nursing Homes.—SNF care usually consists of skilled nursing care and rehabilitation services. Medi- care covers 100 days of care in an SNF following an acute episode of illness. There is a required daily co- payment of one-eighth of the Part A deductible ($44.50 in 1984) for days 20 through 100 in an SNF. Not all SNFs are participants in Medicare. This sit- uation exists in part because there is some financial risk posed by submitting claims to Medicare interme- diaries that may deny payment, and in part because Medicare patients may require more intensive nurs- ing care than the longer term chronically ill Medicaid and private-pay patients (106). Because there is limited access to SNF beds, Medicare patients have often re- mained in the hospital for extra days. For these pa- tients, Medicare has paid as much as four times the necessary cost of patient care, possibly totaling $100 million to $900 million extra annually (106). By en- couraging earlier discharges from hospitals, the DRG payment system will probably decrease these backup hospital days, estimated at from 1 million to 9.2 mil- lion annually (104). To alleviate the SNF bed short- age, some hospitals with extra beds are converting them into nursing home beds for long-term care (103, 320). Hospital reporting requirements for skilled nurs- ing beds are different from the reporting requirements for separate SNFs, so this bed conversion may not re- duce costs to Medicare (278). In 1980, 269,500 elderly and 9,300 disabled Medi- care enrollees used SNF days (161). Yet Medicare pays only 2 percent of the total SNF industry revenues. In 1980, Medicare paid $339.3 million to SNFs for elderly beneficiaries and $13.5 million for disabled benefici- aries who were admitted. Table D-1 shows that dis- charges from and days of care in SNFs by Medicare enrollees have declined since 1969 (391), when a rig- orous claims review policy resulted in retroactive denial of many claims and substantial loss of revenues for some SNFs (106). Quality of care in nursing homes is variable. The Joint Commission on Accreditation of Hospitals ac- credits nursing homes. Medicare conditions of partici- pation for SNFs are complex. State licensure require- ments of nursing homes often have different definitions of “skilled” and “intermediate” care facilities than does HCFA. In many cases, nursing homes choose not to Table D-1.—Use of and Reimbursement for Skilled Nursing Facilities (SNFs) Under Medicare by Type of Enrollee, 1969-79 Discharges (thousands) Days of care (thousands) Reimbursements (millions) Year Elderly Disabled? Elderly Disabled? Elderly Disabled? 1989 ..0inaniaminmsnne 367.9 NAC 14,467.0 NA $277.3 NA 1970 sani niinisnmsnms 277.4 NA 9,901.0 NA 200.3 NA 1971 237.8 NA 7,230.0 NA 164.7 NA 1972 218.0 NA 5,837.0 NA 140.8 NA 1973 o.oo 237.2 2.2 6,372.0 51.0 167.7 $15 1974 ..uvs on cnnssmans 262.8 8.0 7,185.0 216.0 197.4 6.7 i IEE IT Ty 252.1 8.2 6,618.0 222.0 195.6 7.6 O78 2:00: aismeimanae 255.6 8.7 6,808.0 232.0 219.4 8.6 1977 «oo 260.8 8.9 7,145.0 249.0 243.8 9.8 1979 269.69 9.6 7,821.0 297.0 306.1 12.9 Annual compound rate of growth (%) ....... —4.2 3.6° —-8.4 4.9¢ —-1.6 13.5% 3ncludes SNF discharges with at least 1 day of covered care under Medicare. Excludes stays for enrollees who had exhausted their SNF benefits and for whom no discharge record was received. bpreliminary estimates. CNA—Not applicable. dRepresents Medicare beneficiaries admitted rather than those discharged. ©Annual compound rate of growth computed only for 1974-77 because the 1973 figure does not cover a full year due to a change in eligibility policy, and the 1979 data represent slightly different information and thus are not included in the calculation. SOURCE: C. Helbing, “Medicare: Use of Skilled Nursing Facilities, 1976-77," Health Care Financing Notes, U.S. DHHS publication No. 03021 (Baltimore, Md.: Health Care Financing Administration, 1980). App. D—Selected Alternatives to Traditional Health Care Delivery ¢ 187 participate in Medicare because of the extra adminis- trative burden of multiple regulations and requirements. Hospices. —The hospice is a relatively new concept in care for patients with terminal illnesses. Until hos- pice care became available in this country, beginning in 1971, most terminally ill patients had been kept in hospitals and nursing homes. Terminally ill patients have usually undergone highly sophisticated treat- ments (e.g., radiation therapy, chemotherapy, and ex- tensive surgery) without success. In hospices, such patients receive palliation of their symptoms and psy- chosocial care from a multidisciplinary team that in- cludes physicians, nurses, social workers, clergy, psychologists and psychiatrists, dieticians, lawyers, and specially trained volunteers. Home care is one of the desirable aspects of hospice, for economic as well as psychological reasons, although it is not always possible. Furthermore, the use of many volunteers by hospices helps keep costs low. There are about 500 full- service and 500 part-service operating hospices and 80 in the planning stages (362). In 1982, Congress man- dated in the Tax Equity and Fiscal Responsibility Act (Public Law 97-248) that Medicare cover 6 months of hospice care for terminally ill Medicare eligibles under Part A. There is no experience to report on hospice use by Medicare beneficiaries, because hospice care is a new benefit. However, during congressional deliberations, the Congressional Budget Office estimated that poten- tial users would number 268,000 in fiscal year 1984 but only about 12,000 would use the benefit that year (257). Prices for hospice care have been based on the type of care and on costs from demonstration project hos- pices. The type of care is determined by how much nursing time a patient requires in a day. Costs differ between freestanding hospice units and hospital-based units. Medicare will pay a maximum of $6,500 per pa- tient per year, on an aggregate basis (112). Alternatives to Physicians’ Offices Primary medical care traditionally has been pro- vided in physicians’ offices. In recent years, however, alternative sites of primary care have been established in response to economic, social, and health factors. The supply of physicians has increased in many re- gions, so there is competition for patients. In addition, Medicare and Medicaid, among other Government programs, have provided funds for medical care for the elderly and the poor, thereby increasing the po- tential patient population. Patients’ expectations have increased, and the U.S. population itself is growing older and needs more medical care. Two examples of alternative sites of primary care, hospital outpatient clinics and emergency care centers, are described below. Emergency care centers are also alternative sites for outpatient care. Hospital Outpatient Departments. —Hospitals, par- ticularly teaching hospitals, have long had outpatient departments. Yet in recent years, one of the ways hos- pitals have responded to financial pressures has been to expand services, including primary care in outpa- tient departments (136). The increased use of hospital ambulatory care in recent years is also due to limited access to private physicians for inner-city residents, increasing prevalence of chronic diseases, and greater expectations of patients for medical care from hospi- tals (2). Advances in medical technology have also resulted in the movement of some treatments from in- patient to outpatient settings (136,157,248). Visits to hospital outpatient departments by the elderly have been increasing, while the rest of the population has been using fewer outpatient visits (210). Some hospitals are also establishing satellite clinics, i.e., decentralized sites where ambulatory care is avail- able (142). Primary care clinics have been set up by many community hospitals throughout the country (40). Specialized clinics for the outpatient treatment of cancer patients have also been opened in some ur- ban areas (263). Over 10 percent of the physician visits made by per- sons 65 and older in 1978 were visits to hospital out- patient departments or emergency rooms (210). This was a 22.2-percent growth in outpatient department visits since 1973 for that age group. About one-third of these outpatient visits were to the emergency room, one-third to a physician in a clinic, and one-third to ancillary service referrals (210). The growth, especially in the ancillary service referrals and clinic visits, will probably continue after the implementation of Medi- care’s DRG payment system. Hospital outpatient department and emergency room visits are combined in Medicare Part B statis- tics. In 1980, Medicare reimbursed $1.9 billion for out- patient services for about 7.5 million beneficiaries (392). Medicare Part B pays 80 percent of reasonable charges for visits to hospital outpatient departments and emergency rooms. Part B beneficiaries must pay the initial deductible ($75 since 1983) and 20 percent coinsurance, just as they do for a physician office visit. Reasonable charges differ between physician offices and hospitals, and for outpatient visits, hospital charges are generally higher. Because hospitals must accept Medicare assignment, however, some patients may have an incentive to use hospital outpatient serv- ices instead of physician office visits. 188 ¢ Medical Technology and Costs of the Medicare Program Emergency Care Centers. —Emergency care centers are alternatives to hospital outpatient departments and to some emergency room care and to primary care in physician offices. Such centers, though generally equipped with some emergency technologies, do not treat life- or limb-threatening situations, so the name “emergency” may be misleading (325). They usually have more diagnostic technologies on location than a physician's office. Emergency care specialists and some family practitioners have opened emergency care cen- ters to make medical care more accessible to patients who have no primary care physician or who cannot find a physician after hours. The centers have extended hours during evenings and weekends when physicians’ offices are closed, and some are open 24 hours a day, 7 days a week. No appointments are necessary, so care is more convenient for some patients, although pa- tients may not experience desired continuity of care. Use of emergency care centers by the elderly has not been well documented. While one estimate of total visits to these centers was 12 million patient visits for 1982 (238), the elderly population is probably under- represented. To keep costs down, most freestanding emergency care centers accept cash or credit cards only (325), so patients must be reimbursed by their insur- ance companies. This requirement may deter elderly and disabled patients with limited incomes from using the centers. Furthermore, most freestanding centers do not accept Medicare assignment, and this means more money out-of-pocket for elderly and disabled benefi- ciaries. Emergency care centers usually compare their charges with those of hospital emergency rooms rather than with fees for physician office visits. This com- parison is not necessarily a good one, because the care provided is often more like office than hospital care. Nonetheless, the National Association of Freestanding Emergency Centers estimates that center charges are 30 to 50 percent lower than hospital emergency rooms for comparable services (238). If an emergency care center is affiliated with a hospital, Medicare will reim- burse for visits as though the center were a hospital department. If the center is freestanding, Medicare will pay as though the visit were a physician office visit (55). Hospitals must accept assignment to participate in Medicare, but physicians and freestanding centers need not. Thus, if elderly patients were informed about which centers were hospital affiliated or accepted assignment, they would be more likely to choose those centers over the other centers if total prices were com- parable. A 1979 study showed that most of the emer- gency care centers’ revenues came from private in- surers or patients who paid directly, with only a small fraction coming from Medicaid and even less from Medicare beneficiaries (55). The 1983 followup, al- though limited in sample size, showed more centers accepting Medicare funds but some centers specifically excluding Medicare cases (250). The National Association of Freestanding Emer- gency Centers has a policy of not judging the quality of care delivered in individual facilities but leaving that judgment to the patients (238). The centers are char- acterized as physician offices for quality review pur- poses by the physicians practicing in them. There are no licensing laws for freestanding emergency care centers in most States, although some States are try- ing to regulate them (239). If Medicare were to develop a special benefit program for emergency care centers like that for ambulatory surgery centers, more evi- dence of quality would probably be required in the conditions of participation. Alternative Organizations for Health Care Delivery Organizational differences among health care pro- viders allow patients choices and increase competition in the health care market. This section describes two examples of alternative organizations that may in- crease competition among providers. Patterns of use, evidence on cost and quality of care, and effects of Medicare policy are presented for HMOs. PPOs are described only briefly because of a lack of information. Health Maintenance Organizations A defined set of physicians who provide services for a voluntarily enrolled population paying a prospec- tive per capita amount is known as a group practice prepayment plan (GPPP). Some GPPPs have accepted financial risk for hospitalization of their patients and have become part of the competitive health care mar- ket known as HMOs. These organizations, thus, are both providers and insurers of comprehensive but specified medical services. In a series of laws, the Federal Government has pro- vided financial support for HMO development, in- cluding construction loans and mandatory access for HMOs (i.e., employers must offer an HMO health plan as an option for health insurance coverage). Not all HMOs have participated in these Federal programs because of the regulations imposed on participating HMOs (227). From 5.3 million enrollees and 142 plans in 1974, HMO enrollment doubled to 11.6 million enrollees in 269 plans in 1982 (85). Still, HMOs cover only about 5 percent of the U.S. population. The growth of HMOs and other GPPPs since the early 1970's has not been accompanied by a similar App. D—Selected Alternatives to Traditional Health Care Delivery * 189 growth in the number of Medicare enrollees joining these groups. By March 1982, 45 HMOs had cost con- tracts, 2 had normal risk contracts, and 8 had special experimental demonstration risk contracts with Medi- care (392). Another 33 GPPPs were also participating under Medicare in less restrictive contracts (392). Just under 116,000 HMO members were Medicare enrollees as of March 1982, and about 515,000 Medicare en- rollees were GPPP members (392). Thus, around 2 per- cent of the Medicare population is participating in an HMO or GPPP. Under the Social Security Amendments of 1972 (Public Law 92-603), payment to HMOs for Medicare beneficiaries could be made under the usual cost-based method or under a risk-sharing contract. Most HMOs with Medicare enrollees are under a cost-contracting arrangement with Medicare in which they receive monthly interim payments based on their estimated allowed costs with a year-end adjustment to allowed costs (392). As of March 1982, only two HMOs were under a risk contract in which its adjustment was compared with the adjusted average per capita cost (AAPCC) for its services. The AAPCC is the average per capita cost of providing services to the enrolled group of ben- eficiaries if they had been receiving fee-for-service care in their area (there are more specific actuarial and demographic factors that are used in calculating the actual AAPCC). If the HMO's costs are lower, it retains half of the savings above 80 percent of the AAPCC, for a maximum of 10 percent. Higher costs must be absorbed or carried over into the next budget year (392). In September 1982, HCFA began funding five dem- onstration projects on a risk contract basis to see if significant numbers of Medicare beneficiaries could be enrolled in HMOs through aggressive marketing tech- niques and attractive benefit packages (97). These demonstrations have added significance in view of pro- spective payment provisions in the Tax Equity and Fiscal Responsibility Act of 1982 (Public Law 97-248), and the establishment of Medicare's prospective hos- pital payment system using DRGs in the Social Secu- rity Amendments of 1983 (Public Law 98-21). One rea- son why HMOs have not actively sought to enroll Medicare beneficiaries may be that HMOs operate on a per capita payment basis, while their participation in Medicare necessitated significant administrative costs in order to conform to Medicare requirements for cost-based data. Medicare prospective payment would be in tune with HMO operations, and HMOs can be expected to be more aggressive in seeking Medi- care beneficiaries. Another consideration in increas- ing HMO participation in Medicare is that benefici- aries need incentives to join HMOs, because many would have to change physicians and hospitals and would have less freedom of choice of physicians and hospitals in HMOs (145). Analyses of Medicare beneficiaries’ use patterns prior to their enrollment in three of the HCFA HMO demonstrations addressed the issue of whether bene- ficiaries sicker than the average Medicare enrollee would be attracted to HMOs or whether HMOs would recruit Medicare beneficiaries who had fewer medical problems on average (i.e., whether adverse or favor- able selection from the standpoint of the HMO would occur) (97). In two of the HMOs, which were closed-panel HMOs, pre-enrollment use rates of Medicare-reim- bursed services were lower than the use rates of com- parison group beneficiaries. In the third HMO, which operated much like an independent practice associa- tion (IPA) and which was the only significant provider of care in the area, there was no difference between the enrolled and comparison groups. The analysts studying the HMOs concluded that the circumstances under which enrollment occurred probably ruled out deliberate selection of healthier beneficiaries by the closed-panel HMOs and that the selection bias was probably on the part of the enrollees. As beneficiaries enrolling in the closed-panel HMOs probably had to give up their previous doctors and hospitals, these two HMOs may have enrolled Medicare beneficiaries who did not have close ties to their physicians. In the IPA- type HMO, most enrollees were probably receiving care from a physician who belonged to the IPA and therefore did not have to change physicians (97). Interim results a study of one of the closed-panel HMOs (Oregon Region Kaiser-Permanente Medical Care Program) have shown that Medicare beneficiaries can be motivated to join HMOs through premium sav- ings or increased benefits over those available from the fee-for-service sector and that there is a high level of acceptance of continued HMO participation (145). In regards to utilization, recruited Medicare members used hospital beds at a rate slightly higher than in- dividuals over 65 previously enrolled in the plan. But the recruited Medicare members’ rate of use was still much lower than the rate of all individuals over 65 in the same area—1,677 days per 1,000 members per year versus 3,142 days per 1,000 people per year. New members also used about 20 percent more office visits than old members of the same age group. The suc- cessful recruitment of additional members and in- creased utilization of services through the enhanced Medicare coverage provided in the demonstration projects led the authors to hypothesize the following: 1) prior Medicare coverage did not meet a significant 190 » Medical Technology and Costs of the Medicare Program amount of need, and/or 2) those selecting HMOs were more likely to use services. These interim results also support the conclusion reached earlier in the pre- enrollment study (97) that the HMO did not recruit healthier beneficiaries. The evidence on cost savings by HMOs centers around lower hospital utilization, although there is some evidence that physicians in HMOs use fewer tests and procedures than physicians paid on a fee-for- service basis (203). One of the reasons HMOs have not enrolled more Medicare patients may be that the aged and disabled require more costly care, including more physician visits and more hospitalizations, than a younger, healthier population. Numerous studies have examined the structure, process, and outcome factors of quality for HMOs (205). In addition to the difficulty of defining and measuring quality, comparisons of HMO practices to fee-for-service practices are complicated by the insur- ance aspects of the former which change the financial incentives. The evidence on quality in HMOs does not support the contention that HMOs save money by pro- viding lower quality care (204). Neither does the evidence support the suggestion of substantially bet- ter care in all HMOs (205). Thus, care in HMOs ap- pears to be about equal in quality to that received through fee-for-service practices under conventional insurance coverage (429). Preferred Provider Organizations PPOs include a variety of organizational designs. Basically, PPOs are contract agreements between an insurer (or employer, if self-insured) and providers (physicians or hospitals or both) that give services at a reduced rate to the insured group. Patients are given a choice of seeing a physician from the PPO list at lit- tle or no out-of-pocket cost or seeing someone else and having to pay the difference in fees. Incentives for the insurers to enter into PPO agree- ments include reduced cost because of the reduced rate and some control over utilization of medical technol- ogies. This control comes from an agreement by the physicians to participate in utilization review and quality assurance programs. Physicians who overuti- lize tests and procedures potentially will be dropped from the PPO list. There are also incentives for phy- sicians to agree to the discounts of 5 to 20 percent (191). First, insurance claims on a fee-for-service basis will be paid quickly and in full. Second, billing is easier for these patients. Finally, patient volume is guaran- teed to be increased. Indeed, most PPOs have been initiated in areas where there are a very large number of physicians. It is difficult to determine the number of PPOs be- cause PPOs are agreements among entities, not entities themselves. Concentrations of PPOs are in Denver, Los Angeles, and San Francisco (100). MediCal (Cali- fornia’s Medicaid program) is contracting with Califor- nia hospitals, and a new law in that State allows con- tracting between hospitals and other third-party payers. Blue Cross and Blue Shield of Virginia is try- ing to set up a PPO in the Richmond area (147). Blue Cross and Blue Shield of Michigan is developing a PPO for Medicare recipients in Detroit under a HCFA grant (59). Data are not available on patient participation rates in PPOs. The agreements are too new to have generated much publishable data, and they are too di- verse to use their data in comparisons with other phy- sician organizations. It is still too early to draw conclusions regarding cost and quality of PPOs. If patients choose the physicians offering reduced fees, the third-party payers may save. At the same time, these physicians have agreed to par- ticipate in utilization review and quality assurance programs. The design and implementation of these programs will be important to their acceptance by phy- sicians and to their effectiveness. Discussion As noted earlier in this report, the original purpose of the Medicare program was to increase the access to medical care for the Nation's elderly population. Currently, the primary focus of policymakers is on cost containment. This appendix has described ex- amples of alternative sites and organizations for health care delivery. These alternatives may represent future directions for Medicare cost-containment efforts. Rational encouragement of the best alternatives would benefit the Medicare program and its enrollees. In order to decide which alternatives would provide the best quality of care at a low cost, comparisons of evidence on costs and quality are needed. This appen- dix has presented available evidence on patterns of Medicare beneficiaries’ use of several alternatives and the available evidence on cost and quality. Clearly, more research and better data collection are needed. Definitional problems regarding quality, cost, and what constitutes a particular type of care exacerbate the paucity of comparable data (2,205,429). Appendix E.—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies Prepared under contract to OTA by: Linda K. Demlo, Glenn T. Hammons, John M. Kuder, Sally Rodgers, Dianne Lynch The University of lowa College of Medicine and Graduate College and the University of lowa Hospitals and Clinics Introduction This study of decisionmaking by Medicare contrac- tors was conducted for OTA to determine whether there is variation in decisionmaking by Medicare con- tractors (intermediaries and carriers) with respect to the coverage of particular medical technologies. The findings are based primarily on the results of a 1983 telephone survey of Medicare contractors. Background and Objectives Conventional wisdom holds that there is wide varia- tion in the decisions made by Medicare contractors regarding the coverage of particular medical technol- ogies. This variation stems from the absence of precise national policy about which medical care technologies are “reasonable and necessary” and, hence, eligible for reimbursement; the decentralized process by which Medicare coverage policy is promulgated and imple- mented; and the wide range of discretion allowed to individual contractors in making coverage and reim- bursement decisions (54,143,353). Variation in Medi- care contractors’ interpretation of rules governing cov- erage of skilled nursing care has been documented (314). However, comparable information is not avail- able to document the variation in contractors’ cover- age decisions about medical technologies. Any attempt to change the economic incentives in the Medicare program by refining coverage policy in order to control or modify the adoption and use of medical technology, thereby constraining the growth of Medicare costs, ideally would be grounded on a bet- ter understanding of the way in which coverage deci- sions are currently made. This study was intended to assist in developing that information base by address- ing the following specific objectives: 1. To determine the manner in which Medicare in- termediaries and carriers identify new technol- ogies and new uses of established technologies; 2. To determine the manner in which intermediaries and carriers monitor and implement national cov- erage decisions; and 3. To determine whether there is variation among intermediaries and carriers in Medicare coverage of specific technologies: a. by any type of technology (drugs, devices, and medical and surgical procedures); and b. by stage of development (experimental, new, and established). This appendix presents the study findings. It includes a discussion of study methods, a description of find- ings with respect to reported coverage policies and fac- tors associated with those policies, and a summary of conclusions. Methods The study requirements included drawing a sample of Medicare carriers and intermediaries, selecting the technologies for which specific coverage questions were formulated, developing and pretesting the ques- tionnaire, and gathering and analyzing the survey data. The various aspects of the study methodology are presented in the following sections, along with a discussion of study limitations. Sampling Plan The sampling plan was developed to reflect charac- teristics of Medicare contractors that were hypothe- sized to influence their coverage decisionmaking. Ade- quate representation of both commercial insurance companies and Blue Cross/Blue Shield (BC/ BS) plans was indicated because of the possibility that variation in claims processing between these groups might influ- ence coverage decisions. Discussions with staff of OTA, the Health Care Financing Administration (HCFA), local BC/BS plans, and nearby hospitals sug- gested that intermediaries are more limited than car- riers in their ability to identify new technologies or new uses of established technologies because of the re- stricted information provided by the hospital claim form. Accordingly, both carriers and intermediaries were represented. The potential variation in physician practice patterns among geographic regions required 191 192 e Medical Technology and Costs of the Medicare Program a geographically balanced sample. To simplify the analysis, we preferred to avoid disproportionate sampling from fractions among these groupings of con- tractors. In the absence of firm information about the likely variation in decisionmaking among contractors, the sample size was governed primarily by the need to balance the contractors’ characteristics specified above and the time and funds available for conduct- ing the study. A total sample size of 60 contractors was deemed reasonable. After determining the target sample size, the most recent HCFA intermediary and carrier directory was used as the sampling frame (385). Contractors in Ha- waii and Puerto Rico were eliminated because of the time and expense involved in communicating with them. The HCFA Offices of Direct Reimbursement and Group Health Operations were also eliminated. (These offices serve as intermediaries and carriers for pro- viders who bill HCFA directly, and, as such, have a unique national perspective. Because of pending or- ganizational changes, however, information about their operations would provide little useful data on which to make policy recommendations.) Random sampling resulted in an initial sample of 21 intermediaries and 39 carriers, whose distribution among Blue Cross/Blue Shield and commercial insur- ance contractors and geographic region is shown in table E-1. After telephoning to verify the names and addresses of persons in charge of government pro- grams, a letter of introduction was sent to each sampled contractor, which described the purpose of the study and indicated that the program administrator would be contacted to schedule an interview. Almost immediately, however, changes were re- quired in the initial sample. Both the verification phone calls prior to mailing the introductory letter revealed that in many cases the HCFA directory was outdated. Some contractors no longer held HCFA contracts but were able to tell us who the current contractor was. In these instances, we replaced the prior contractor with the current contractor in order to maintain the geographic representativeness of the sample. Two other categories of sampled contractors performed no claims review functions: Railroad Retirement Boards, who contracted for review with a commercial insur- ance company; and home offices of large commercial companies, whose review functions were performed by field offices. Both were listed in the HCFA direc- tory and given intermediary or carrier numbers, so there was no way to identify them in advance. When such contractors were identified, they were declared ineligible for inclusion in the sample because they were not engaged in claims review. They were subsequently replaced by another randomly selected contractor. The final sample of contractors is shown in table E-2, which reflects the changes discussed above. Four contractors (all Blue Cross or Blue Shield plans) refused to participate. Their reasons for nonparticipation re- flected extreme staff shortages and time pressures in- volved in the need to implement system changes de- manded by the Tax Equity and Fiscal Responsibility Act of 1982 (Public Law 97-248) and Social Security Amendments of 1983 (Public Law 98-21). Despite these changes, the representation of carriers and intermedi- aries remained about the same in both the original and final samples. Similarly, the distribution of contrac- tors among HCFA regions remained about the same except that the minimum representation decreased in the final sample (i.e., only three contractors in the Denver region). Questionnaire Development A telephone-administered questionnaire was devel- oped to elicit information about intermediaries’ and Table E-1.—Distribution of Initially Sampled Medicare Contractors Number of carriers Number of intermediaries HCFA region BC/BS Commercial BC/BS Commercial Total Seattle.................. 1 2 2 1 6 San Francisco ........... 1 2 1 2 6 Denver.................. 2 2 0 0 4 Dallas .................. 2 2 1 0 5 Kansas City ............. 3 2 1 1 7 Chicago................. 2 3 3 0 8 Atlanta. ................. 3 2 1 1 7 Philadelphia ............. 2 0 2 0 4 New York ............... 2 3 1 1 7 Boston.................. 1 2 1 2 6 Total ................. 19 20 13 8 60 SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies * 193 Table E-2.— Distribution of Actual Study Participants Number of carriers Number of intermediaries HCFA region BC/BS Commercial BC/BS Commercial Total Seattle. ..o nis ninverves 1 2 1 1 5 San Francisco ........... 1 2 1 2 6 Demwer..... siawisminmes 2 1 0 0 3 Dallas .:zvsmes mms sms wo ss 2 2 1 0 5 Kansas City ............. 3 2 1 1 7 Chicago. ..........o.onnn 2 2 3 0 8 Atanta. ..... ov snesners 3 4 1 0 8 Philadelphia ............. 3 0 2 0 5 New York .........cov.n. 1 1 1 2 5 BOSON i csuss vn roms ims ws 2 1 2 0 5 TORBAY «isos ima vmin ssn e woe 20 17 13 6 56 SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. carriers’ claims review processes, their uses and impres- sions of HCFA transmittals that update the Medicare coverage issues appendix, their methods for identify- ing technologies for which coverage may be question- able, and their policies with respect to covering certain technologies specified in the questionnaire. Potential questionnaire items were discussed with regional and national reimbursement experts. Two pilot tests were conducted, and additional questions were eliminated or refined in order to accommodate the 40-minute time limit. The most complicated aspect of questionnaire de- velopment was identifying the technologies about which coverage questions were asked. This task was handled primarily by the physician member of the study team, in consultation with colleagues at the University of Iowa College of Medicine, medical con- sultants in insurance companies, and staff of OTA, HCFA, and the Office of Health Technology Assess- ment (OHTA) of the Public Health Service (PHS). In addition, recent HCFA updates to the coverage issues appendix, the Commerce Clearinghouse version of the appendix, and lists of studies completed or underway by OHTA were carefully reviewed. As indicated by the study objectives, the technol- ogies about which coverage questions were asked were chosen to represent a variety of types (drugs, devices, and medical and surgical procedures) at various stages of development (experimental, new, and established). This diversity was intended to raise different cover- age issues reflecting questions about the extent to which these technologies are safe, effective, and gen- erally accepted within the medical community. Devel- oping a matrix of technologies for potential inclusion in the study proved to be difficult, however, since assigning individual technologies to these categories frequently required some arbitrary decisions. The di- viding points between experimental, new, and estab- lished technologies are not clear-cut. Determining whether a procedure is medical or surgical may reflect personal biases about which medical specialty should be permitted to perform it, as well as its contribution to diagnostic v. therapeutic decisions, and its relative degree of invasiveness. The questionnaire was designed to include some technologies for which there is an ex- plicit HCFA policy and some for which there is not, in which case Medicare contractors are expected to make their own determinations. Finally, Medicare cov- erage policy is constantly evolving. Even during this limited study period, there were changes in HCFA pol- icy and its interpretation. The technologies included in the questionnaire are shown in table E-3. However, the sometimes arbitrary and fluid nature of the man- ner in which they are categorized should be noted. Data Gathering The interviews were conducted by a nonphysician senior member of the study team and two research assistants who were second-year graduate students in hospital and health administration. All interviewers consulted with the physician member of the study team and reviewed medical literature and the coverage issues appendix in order to become thoroughly familiar with the nature and uses of the medical technologies in- cluded in the study. Our intent was to interview a sen- ior official from each Medicare contractor, who would be familiar with the overall processes by which claims were reviewed, as well as the contractor's specific cov- erage policies. In most cases, the respondent was a nonphysician administrator responsible for the Part A or Part B Medicare contract. Sometimes, physician consultants and nurse reviewers responded; occasion- ally, a conference call was held so that all three 194 e Medical Technology and Costs of the Medicare Program Table E-3.—Medical Technologies Included in Questionnaire Type of Developmental Technology technology HCFA policy stage UseP Intraocular lens implant following cataract Surgical procedure Covered Experimental/new Therapeutic removal Chelation therapy with EDTA in the treatment of atherosclerosis Chelation therapy with EDTA in the treatment of rheumatic arthritis Pacemaker device and its implantation for a patient with chronic second degree AV block of Mobitz type Il, with symptoms attributable to intermittent complete heart block Pacemaker device and its implantation for a patient who has sinus bradycardia without symptoms Implantable chemotherapy infusion device, such as the “Infusaid,” and its implantation for a patient with primary hepatic malignancy, such as hepatoma Implantable chemotherapy infusion device for a patient with metastatic cancer in the liver External insulin infusion pump for a diabetic patient Purchase of a home blood glucose monitor such as the Dextrometer® Continuous 24-hour monitoring of blood pressure using an automatic device (i.e., preset intervals not under control of patient for readings) Continuous 24-hour monitoring of blood pressure using a semiautomatic or patient-activated device Percutaneous transluminal coronary angioplasty (PTCA) for a single vessel procedure PTCA for two or more coronary arteries Streptokinase administration at cardiac catheterization into a coronary artery to dissolve a clot in a patient with acute myocardial infarction (thrombolytic therapy) External osteogenic stimulator for use in the treatment of a long bone fracture Chemonucleolysis (i.e., injection of the enzyme chymopapain or “Discase”) in the treatment of a herniated disc Electroencephalographic (EEG) monitoring during carotid endarterectomy EEG monitoring during open heart surgery Apheresis (therapeutic pheresis) or plasma exchange in the treatment of hyperglobulinemias such as a multiple myeloma Drug/medical procedure Drug/medical procedure Device/surgical procedure Devicelsurgical procedure Device/surgical procedure Device/surgical procedure Device Device Device/medical procedure (interpretation) Device/medical procedure (interpretation) Medical procedure Medical procedure Drug/medical procedure Device/medical procedure Drug/medical or surgical procedure Medical procedure Medical procedure Medical procedure Not covered Local option Covered? Covered? Covered? Covered? Covered? Covered? Not covered (as of July 1983) Not covered Covered? Not covered? Local option Covered? Local option Covered? Not covered Covered? Experimental Experimental Established Established New New Experimental/new Established New New New/established New Experimental New Experimental/new New use of established New use of established New or new use of established Therapeutic Therapeutic Therapeutic Therapeutic Therapeutic Therapeutic Therapeutic Diagnostic Diagnostic Diagnostic Therapeutic Therapeutic Therapeutic Therapeutic Therapeutic Diagnostic Diagnostic Therapeutic App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies ¢ 195 Table E-3.—Categorization of Medical Technologies Included in Questionnaire—Continued Type of Developmental Technology technology HCFA policy stage UseP Apheresis (therapeutic pheresis) or plasma Medical procedure Not covered Experimental use Therapeutic exchange in the treatment of systemic of established lupus erythymatosis (SLE) Topical oxygen therapy for decubitus Device/medical ulcers (e.g., Topox device) procedure Not covered Experimental Therapeutic Biofeedback therapy for intractable pain Device/medical Not covered New use of Therapeutic procedure established PUVA therapy for psoriasis Device/medical Covered? New Therapeutic procedure ajndicates that Medicare coverage is contingent on compliance with specified criteria or guidelines. bupjagnostic” includes use to monitor or guide therapy. SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. perspectives could be represented. The latter occurred at the suggestion of the respondent. The variation in respondents introduces the potential for bias; however, when the data were analyzed according to the posi- tion and discipline of the respondent, no systematic differences were detected. Another source of potential bias was introduced by variation in the method by which answers were ob- tained. As noted earlier, the questionnaire was de- signed to be administered by telephone. In some in- stances, however, the respondent insisted on reviewing the questionnaire in advance and then responding by telephone or submitting the answers by mail. This occurred when the respondents were short of staff or when the contractor had a policy of only responding to self-administered mail questionnaires. Rather than lose the respondent, we reluctantly agreed to send the questionnaire in advance. Finally, in some cases, re- spondents were unable to respond to all the questions about technologies and preferred to check on cover- age policy and provide the answers later, either in a followup telephone call or in writing. Obviously, the respondents who reviewed the ques- tionnaire in advance or who provided answers in followup contacts had the opportunity to consult the coverage issues appendix and provide the “correct” answer. We attempted to minimize this possibility in several ways. In administering the questionnaire, we were careful to emphasize that we expected variation to occur, that there was no “correct” answer, and that we were simply interested in ascertaining what the con- tractor’s customary policy would be. In some in- stances, there was a general HCFA coverage policy, but in others, there was not. Ultimately, the data were analyzed according to the method by which answers were obtained; entirely by telephone interview (61 per- cent of the responding contractors); entirely by self- administration (14 percent); or by a combination of telephone interview and followup telephone or writ- ten response to questions that were initially unan- swered (25 percent). When the reported coverage policies for individual technologies were categorized according to these three general methods of response, a statistically significant difference was found for only one technology (the im- plantation of a pacemaker for a patient with sinus bradycardia without symptoms). For this technology, the responses derived entirely through a telephone in- terview indicated lowest adherence to HCFA policy, while those based on mixed methods showed highest adherence. The differences in reported coverage for a 24-hour, semiautomatic or patient-activated blood pressure monitor were almost statistically significant (p = 0.079). In this case, the responses to questions that were totally self-administered showed lowest adherence to HCFA policy, while those based on mixed methods again showed highest adherence. For all other technologies for which HCFA has established a coverage policy, the differences were not statistically significant at the 0.05 level and were about evenly divided among general methods of response. However, there remained a slight tendency (not statistically sig- nificant) for responses from contractors who utilized the combination method or self-administered responses to show greater adherence to HCFA policy. If one as- sumes that the responses derived entirely from tele- phone interviews are most likely to reflect actual con- tractor behavior (e.g., the reviewer simply made a decision based on the claim without bothering to re- fer to the coverage issues appendix), then the findings from this study are biased in the direction of under- estimating deviation from HCFA coverage policy. Study Limitations The potential for bias stemming from the sample, variation in the position and discipline of the respond- ent, and the methods by which responses were ob- 196 ¢ Medical Technology and Costs of the Medicare Program tained was noted above. In addition, responses reflect the opinion of (usually) one respondent at one point in time about a limited set of medical technologies. Thus, the data would not reflect potentially different judgments about coverage policy made by different in- dividuals within a single contractor organization or by a single individual at different points in time. Time and resource limitations precluded an indepth survey of all contractors in a manner that would illuminate both inter- and intracontractor variability in coverage deci- sions, categorized by a wide range of technologies and administrative and policy variables. Perhaps a more definitive study would be based on the submission of actual, identical claims to a sample of Medicare con- tractors and the analysis of variation in claims proc- essing and decisionmaking; however, that was not our charge. Despite the limitations of the current study, we believe that the findings are generally reliable and that the resulting information on coverage decisions of a sample of Medicare contractors should provide a useful framework for considering alternative Medicare coverage policies. Findings Although there was variation in reported coverage decisions by the Medicare carriers and intermediaries in our study, the processes by which claims are re- viewed were quite predictable. The characteristics of the participating contractors are described below. Characteristics of Participating Contractors The 56 Medicare contractors (37 carriers and 19 in- termediaries) included in the study reported fairly sim- ilar methods for processing Medicare claims. After an initial review of completeness, most claims pass through (potentially) three levels of review that reflect differing degrees of comprehensiveness, specificity, and clinical judgment and involve clerical employees, registered nurses, and physicians or some other health professional (pharmacist, podiatrist, occupational therapist, etc.) in that order. All but nine (16 percent) of the respondents utilized some automated screening procedures that vary in sophistication. The nine re- spondents without automated screening procedures were all intermediaries (contractors that process Part A claims), rather than carriers (contractors that proc- ess Part B claims). The most common computerized screens flag cases that exceed certain utilization parameters (specified numbers of hospital days, office visits, lab services, nursing home visits, etc.), claims requesting payment for noncovered services, incom- patible diagnostic and procedure codes, and claims in- volving diagnoses, procedures, or providers which are automatically submitted to medical review by phy- sicians. The referral of claims to registered nurses is gener- ally based on rather clear-cut screening and referral guidelines, recorded in policy manuals and employed either by clerical workers or automated review proc- esses. All but two contractors (4 percent) also have established criteria to assist nurse reviewers in deter- mining when to refer a coverage question to a medi- cal consultant. Referral guidelines often reflect cases that have been troublesome in the past, as well as claims for technologies for which there is no prior claims experience, so there is no precedent for mak- ing a coverage determination. Some guidelines specify particular conditions or technologies for which review by a physician consultant is always required, such as bypass surgery, pacemaker implantations, computed tomography (CT) scans, and cases involving enteral “and parenteral feeding. Physician consultants utilize their own knowledge of medical practice and the scientific literature to re- solve many coverage questions. Institutional sources of information utilized by consultants in making cov- erage determinations are shown in table E-4. The HCFA regional office, colleagues in other insurance companies or BC/BS Plans, and State and national medical and specialty associations are most frequently consulted. Professional Standards Review Organiza- tions (PSROs) (now utilization and quality control peer review organizations (PROs)) are least frequently used. Other resources include “inhouse” peer review panels, publications of the Food and Drug Adminis- tration (FDA), drug manuals, informal medical con- sultation, and policies established for private programs whose claims are also processed by the Medicare con- tractor. The responses were very similar for both com- mercial insurance companies and BC/BS plans, except that the “Blues” were much more likely to turn to na- tional insurance associations (in this case, the Blue Cross and Blue Shield Association), while commercial insurers were more likely to rely on State or national medical or specialty associations. The responses for intermediaries and carriers were also quite similar, ex- cept that intermediaries were more likely than carriers to rely on information from PSROs, HCFA regional offices, national insurance associations, and manufac- turers. These differences were not statistically signifi- cant, however. Most contractors report that they learn about new technologies for which coverage questions might be raised prior to the actual submission of claims through inquiries from providers and manufacturers, drug and App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies ® 197 Table E-4.—Sources of Information Used by Medical Consultants in Making Coverage Decisions Percent of consultants using source Sources Yes No Colleagues in another insurance company or BC/BS plan .............. 78.6 21.4 HCFA regional office. .......... ieee 87.5 12.5 University medical C8nlar ...... ais umismsamssminmesms sms ones ams meson 44.6 55.4 PORE (PROV vs vva snus mes mes mes wns assassins oss ass aus sass sms mes wa 26.9 73.1 National insurance association ................ coin... 37.5 62.5 State or national medical or specialty association .................... 75.0 25.0 Drug Or device manufacturBr . ..... cs ssssamsmvanssymsams sw ssnss oes us 53.6 46.4 [14 = CP 51.0 49.0 SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. device approval lists from the FDA, and HCFA an- nouncements. Once claims are submitted, new tech- nologies are identified because of the absence of code numbers or the presence of codes that are not recog- nized by the claims reviewers. A few contractors con- duct extensive research to assist in identifying and making coverage decisions for such claims, utilizing in-house review panels, specialists in nearby medical colleges, and medical journals. The majority, however, assume a more passive role. Similarly, the majority are reasonably well satisfied with existing methods for identifying new technologies and report that this is not a big problem for them. Some mentioned a need for greater cooperation between national medical and in- surance associations and governmental agencies, as well as faster turnaround on coverage decisions, once questions are raised. There were no noticeable dif- ferences between intermediaries and contractors in these responses. The volume of claims processed by the study con- tractors varies widely. During the March to January quarter of 1983, the total number of Medicare claims processed ranged from 18,000 to 4,155,000 claims per contractor, with a mean of 854,741. Reported Coverage Decisions The basic data reflecting reported coverage decisions for specified medical technologies are shown in table E-5. The columns headed “covered” and “not covered” include clear-cut responses with no associated quali- fications or criteria. To estimate the frequency with which contractors cover individual technologies, the reader should combine the responses from the first and third columns. The latter include responses indicating that coverage would be dependent on compliance with specified criteria or guidelines, discussed in more detail below. Whenever possible, respondents were limited 25-337 0 - 84 - 14 to these first three categories, even if that required a followup phone call to determine actual coverage pol- icy. The fourth, fifth, and sixth columns (referral to the coverage issues appendix, to a physician consul- tant, or to the HCFA regional office) may be viewed as “last resort” responses. They were used in instances where the respondent had never heard of the technol- ogy, or had never seen a relevant claim, and had no precedent for making a decision other than to refer to one of the three sources of assistance. The number of responses is sometimes less than the total number of study participants. Since the same questions were asked of both carriers and intermediaries, there were some instances when the question was not applicable. For intermediaries, this was especially true for ques- tions about durable medical equipment for home use. When a question was not applicable, the response was coded as “missing data” and excluded from the reported responses. Table E-5 reveals some instances of near unanimity or high levels of agreement in coverage decisions— particularly for intraocular lens implant following cat- aract removal (98.2 percent covered), the use of chela- tion therapy in the treatment of atherosclerosis (87.0 percent not covered), and the use of topical oxygen therapy for decubitus ulcers (86.3 percent not covered). However, there are also examples of considerable va- riation in coverage policies—particularly for a pace- maker device and its implantation for a patient who has sinus bradycardia without symptoms, an implant- able chemotherapy infusion device for a patient with metastatic liver cancer, an external insulin infusion pump for a diabetic patient, and streptokinase admin- istration at cardiac catheterization to dissolve a clot in a patient with acute myocardial infarction. These responses and their implications are perhaps better pre- sented in table E-6, which categorizes them according to HCFA coverage policy. Table E-5.—Reported Coverage Decisions for Specified Medical Technologies Referral to Referral to Referral to Covered with coverage issues physician HCFA regional Number of Technology Covered Not covered qualifications appendix consultant office responses Intraocular lens implant ................... 98.2% _ — 1.8% — — 55 Chelation therapy: atherosclerosis .......... — 87.0% 3.7% 7.4 1.9% _— 54 Chelation therapy: rheumatoid arthritis ...... 3.8 81.1 1.9 7.5 1.9 3.8% 53 Pacemaker: chronic second degree AV block 70.9 —_ 20.0 1.8 7.3 —- 55 Pacemaker: sinus bradycardia without BYMPLOMS nics snssmes sms vms Smes mss wns 13.0 44.4 29.6 1.9 11.3 — 54 Implantable chemotherapy infusion device: primary hepatic malignancy .............. 51.9 21.2 11.5 3.8 11.5 — 52 Implantable chemotherapy infusion device: cancer metastatic to liver................ 38.5 26.9 15.4 3.8 13.5 1.9 52 External insulin infusion pump ............. 10.6 46.8 27.7 4.3 8.5 21 47 Home blood glucose monitor .............. 11.8 7.8 72.5 7.8 - 51 24-hour blood pressure monitoring: automatic 8.3 52.1 27.1 4.2 6.3 21 48 24-hour blood pressure monitoring: semi- automatic or patient-activated ............ 4.2 77.1 8.3 6.3 4.2 — 48 PTCA: single vessel procedure ............. 61.1 3.7 25.9 9.3 — — 54 PTCA: 2 or more coronary arteries .......... 19.2 51.9 19.2 3.8 3.8 1.9 52 Streptokinase at cardiac catheterization: .... 30.2 45.3 15.1 1.9 5.7 1.9 53 External osteogenic stimulator: long bone fracture ........... 275 3.9 58.8 9.8 — — 51 Chemonucleolysis: herniated disc .......... 64.0 10.0 14.0 2.0 8.0 2.0 50 EEG monitoring: carotid endarterectomy .... 68.0 14.0 10.0 6.0 2.0 — 50 EEG monitoring: open heart surgery ........ 15.4 71.2 9.6 1.9 1.9 ion 52 Apheresis: hyperglobulinemias (multiple myeloma) ............. i 81.1 5.7 7.5 5.7 — -_ 53 Apheresis: systemic lupus erythymatosis. . .. 18.0 60.0 14.0 4.0 4.0 - 50 Topical oxygen therapy: decubitus ulcers. ... 7.8 86.3 5.9 — — — 51 Biofeedback: intractable pain .............. 9.3 55.6 31.5 1.9 1.9 — 54 PUVA: pSONasis .. .uc:usiiusansnmvsnsabe ss 38.5 5.8 50.0 - 1.9 3.8 52 SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assess- ment, U.S. Congress, Oct. 28, 1983. weiboid eJedIpey 8y} JO S}SoD pue ABojouyos] [BOIPON eo 861 App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies * 199 Technologies Covered by HCFA The technologies for which HCFA explicitly provide coverage! have a higher percent of responses indicating unqualified coverage than those in any subsequent coverage category (see table E-6). The implantation of an intraocular lens provides an exception to HCFA's general policy of not covering experimental or in- vestigational items or services. Although FDA still con- siders intraocular lenses to be investigational, Congress directed FDA to study them without interfering with One may argue that even for these technologies, HCFA specifies some cri- teria or guidelines for approving coverage. For example, the services must still be reasonable and necessary for the individual patient. With respect to pacemakers, HCFA has specified three groups of medical indications that: 1) may be covered without further claims development, 2) require more spe- cific claims information, or 3) would generally result in denial. Furthermore, final coverage determination of a claim for pacemaker implantation must “take account of the circumstances of the particular claim, as well as factors such as the medical history of the individual patient.” However, our phras- ing of the questions in the first grouping was such that their eligibility for Medicare coverage was intended to be obvious. their availability to patients. Lens implantation is the only technology included in this study for which almost all respondents (about 98 percent) indicated a policy of unqualified coverage. For the remaining tech- nologies in this category, most qualified responses re- flect the contractor's policy of assuring that HCFA's coverage criteria are met, even though the questions were phrased so as to leave little doubt about their eligibility for coverage. Because of the publicity sur- rounding pacemakers, many contractors automatically submit all such claims to medical consultants for in- depth investigation prior to payment. If the qualified and unqualified coverages are combined, close to 90 percent of the respondents indicated that they approve coverage for the technologies that are explicitly cov- ered according to HCFA policy. The one exception was electroencephalographic monitoring during carotid endarterectomy, which reportedly would not be cov- ered by about 14 percent of the contractors. Table E-6.— Reported Coverage Decisions by Medicare Contractors Decisions/policy of contractors Covered with Technology/HCFA coverage policy Covered Not covered qualifications Refer for advice HCFA explicitly covers: Intraocular lens implant . ........... iin 98.2% — — 1.8% Pacemaker: chronic second degree AV block ....... w... 109 — 20.0% 9.1 PTCA: single vessel procedure ........................ 61.1 3.7% 25.9 9.3 EEG monitoring: carotid endarterectomy ............... 68.0 14.0 10.0 8.0 Apheresis: hyperglobulinemias (multiple myeloma). ...... 81.1 5.7 75 5.7 HFCA covers with qualifications: Home blood glucose monitor ......................... 11.8 7.8 725 7.8 External osteogenic stimulator: long bone fracture ...... 27.5 3.9 58.8 9.8 PUVA: PSOM@SIS . . . «vo ovvvviiin inne iin nana annnns 38.5 5.8 50.0 5.8 Implantable chemotherapy infusion device: primary hepatic malignancy ...............ooiiiniinnnnn. 51.9 21.2 11.5 15.4 Implantable chemotherapy infusion device: cancer Metastatic 10 JIVOF «. onions musmms sms pw ssims wus nans 38.5 26.9 15.4 19.2 External insulin infusion pump ............. a 10.6 46.8 27.7 14.9 No explicit HCFA policy—contractor decides (local option): Chelation therapy: rheumatoid arthritis ................. . 81.1 1.9 13.2 Streptokinase at cardiac catheterization: AMI ........... 30.2 45.3 15.1 9.4 Chemonucleolysis: herniated disc ..................... 64.0 10.0 14.0 12.0 HCFA denies, but not explicitly: Biofeedback: intractable pain ......................... 9.3 55.6 31.5 3.7 PTCA: two or more coronary arteries... ................ 19.2 51.9 19.2 9.6 Apheresis: systemic lupus erythymatosis ............... 18.0 60.0 14.0 8.0 HCFA explicitly denies: Chelation therapy: atherosclerosis ..................... 87.0 3.7 9.3 Pacemaker: sinus bradycardia without symptoms ........ 13.0 44.0 29.6 13.0 24-hour blood pressure monitoring: automatic policy effective 7/83)... 8.3 52.1 271 12.5 24-hour blood pressure monitoring: semiautomatic or patient activated ............. iii 4.2 771 8.3 10.4 EEG monitoring: open heart surgery ................... 15.4 71.2 9.6 3.8 Topical oxygen therapy: decubitus ulcers............... 7.8 86.3 5.9 —- SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. 200 e Medical Technology and Costs of the Medicare Program The second category of technologies in table E-6 in- cludes those for which HCFA has delineated criteria that must be met in order for coverage to be approved and for which our descriptions of the procedure were less specific than in the first category. The qualifica- tions associated with approving a claim for the pur- chase of a home blood glucose monitor, the use of an external osteogenic stimulator in the treatment of a long bone fracture, and the use of PUVA for psoriasis consistently reflect the application of criteria specified in the coverage issues appendix. If the qualified and unqualified coverages are combined, from 84 to 88 per- cent of the respondents approve coverage for these three procedures, although some approvals may be granted without ascertaining whether the specified cri- teria are met. The responses pertaining to the coverage of infusion devices require somewhat more interpretation. Cur- rently, the only pertinent national coverage instruc- tions are those in section 60-9 of the coverage issues appendix and in section 2100.5 of the Medicare car- rier’s manual. The coverage issues appendix provides that infusion pumps are covered if the contractor's medical staff verifies the appropriateness of the ther- apy and of the prescribed pump for home use. The Medicare carriers manual provides that reimbursement may be made for supplies necessary for the effective use of durable medical equipment, citing as an exam- ple tumor chemotherapy agents used with an infusion pump, as long as the drug or biological used with a pump is reasonable and necessary for the patient's treatment. The state of the art of infusion pump tech- nology has changed considerably since these policies were developed. They have been interpreted by HCFA to include an implantable pump such as that manu- factured by the Infusaid Corp. (referred to in the ques- tionnaire), even though the policies were not devel- oped with that technology in mind. PHS is examining issues related to the coverage of infusion pumps, in- cluding chemotherapy and insulin devices. HCFA is postponing the issuance of revised chemotherapy guidelines until those studies are completed. The re- sponses to our questionnaire items reflect this uncer- tainty. The respondents were most likely to cover an im- plantable chemotherapy infusion device, such as the “Infusaid,” and its implantation for patients with pri- mary hepatic malignancy, such as hepatoma. They were less likely to cover the same device for a patient with cancer metastatic to the liver. They were least likely to cover an external insulin infusion pump for a diabetic patient; of the three technologies, the re- sponses cover the total range of response options— perhaps more so than for any other technology con- sidered to this point. The circumstances of coverage for the chemotherapy infusion device reflected FDA approval status of the drug to be administered, as well as the condition of the patient. For example, one re- spondent indicated that the device would be covered for a patient with metastasized liver cancer only if the liver were the key to survival and at least a 6-month survival period were likely. Criteria were also speci- fied for coverage of an external insulin infusion pump, including patient condition (e.g., a “brittle diabetic”), case-by-case review to assure medical necessity rather than simply convenience, and a multidisciplinary as- sessment of diet and exercise programs and patient motivation. One respondent indicated that the same criteria would be applied as for a home blood glucose monitor. “Local Option” Technologies Technologies are included in the coverage issues ap- pendix only if they have presented difficult coverage questions for Medicare contractors, who generally would have submitted them to a HCFA regional of- fice, and, eventually, to the central office in Baltimore for a national coverage determination. HCFA policy holds that individual Medicare contractors should make coverage decisions for all technologies that are not mentioned in the coverage issues appendix. In other words, their coverage status is determined by “local option.”? The technologies included in this study that fall into the “local option” category all involve the use of drugs, which in turn, requires that contrac- tors refer to FDA policy in making Medicare cover- age decisions. These technologies are included in the third category in table E-6. Although the drug mentioned in our questions about chelation therapy (endrate or disodium EDTA) has been approved by FDA, the labeling indications do not mention its use for rheumatoid arthritis. About 81 per- cent of the respondents approve coverage of strep- tokinase administered at cardiac catheterization to dissolve a clot in a patient with acute myocardial in- farction, as well as chemonucleolysis injections in the *This distinction between technologies for which coverage is specified in the coverage issues appendix and those for which coverage is expected to be a matter of local option is overly simplistic and glosses over an important source of ambiguity and confusion in HCFA policy. Not all technologies that are not mentioned in the coverage issues appendix are expected to be a mat- ter of local option. As discussed in the next section, some technologies are not explicitly denied in the coverage issues appendix, even though, accord- ing to HCFA staff, the contractors should know that they are not covered. How this knowledge is transmitted is unclear. Furthermore, when contrac- tors encounter claims for technologies that are not included in the coverage issues appendix and which raise issues of national importance because they are experimental or potentially costly, the contractors are expected to ask HCFA for a national coverage determination. However, the criteria for iden- tifying such technologies are not well specified. App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies ¢ 201 treatment of a herniated disc. In both instances, cov- erage is generally contingent on FDA approval. The timing of the streptokinase is important: respondents require that it be administered from 4 to 6 hours after onset of symptoms; one respondent required that the patient be placed in a specialized care unit postopera- tively. In approving chemonucleolysis claims, two respondents require that its use be limited to the lum- bar region; two others require that the provider have special training. Technologies Not Covered by HCFA For some technologies, HCFA's Office of Coverage Policy expects that claims will be denied, even though there is no HCFA policy explicitly denying coverage. For example, some sections of the coverage issues ap- pendix state that technologies will be covered only in specified circumstances, meaning by implication, that all other uses of the technology will not be covered. The use of biofeedback for intractable pain is the only study technology to fall into this category. Section 35- 27 of the coverage issues appendix states: Biofeedback therapy is covered under Medicare only when it is reasonable and necessary for the individual patient for muscle reeducation of specific muscle groups or for treating pathologic muscle abnormalities of spasticity. . .(etc.). . . This therapy is not covered for treatment of ordinary muscle tension states or for psychosomatic conditions. Fifty-five percent of the contractors follow HCFA's intent and reported that they would not cover biofeed- back therapy for intractable pain. Reasons given for approving coverage with qualifications include: docu- mentation that all other methods have failed, when used in conjunction with physical therapy, when part of a pain rehabilitation center program, and when used for specific muscle reeducation. (The final qualifica- tion listed may indicate that the respondent misunder- stood the question. Although specific muscle reeduca- tion would be a reimbursable use of biofeedback, that is not what the question addressed.) The questionnaire included two other technologies which are not explicitly denied coverage, but for which HCFA believes the contractors should understand that coverage should be denied. In both instances, other uses of the same technology are specifically covered. According to HCFA, percutaneous transluminal cor- onary angioplasty (PTCA) to eliminate obstruction in two or more coronary arteries should not be covered, even though a single vessel procedure is covered. Almost 52 percent of the respondents would not ap- prove a claim for PTCA involving two or more ar- teries, while about 38 percent would approve the claim, either with or without qualifications. The 25-337 0 - 84 - 15 reported qualifications suggest that the contractor would review each case individually to determine med- ical necessity, that the criteria specified for a single- vessel procedure would be applied, that the claim would be approved only if the requested charges were reasonable, and that the claim would be approved, but the level of reimbursement would be the same as for a single-vessel procedure. Presumably, the HCFA re- striction on multiple-vessel procedures was based on a concern for increased patient risk; however, it does provide an opportunity for “creative” billing. The other instance in which coverage should be un- derstood to be denied is the use of apheresis (thera- peutic apheresis) or plasma exchange in the treatment of systemic lupus erythymatosis (SLE). Therapeutic apheresis is covered for several indications listed in the coverage issues appendix, but not for systemic lupus. In May 1983, PHS tentatively recommended that apheresis for SLE should not be covered; however, that recommendation was withdrawn and the issue will be reconsidered. This uncertainty is reflected in our re- sponses. Sixty percent of the contractors do not cover apheresis for patients with SLE, while 18 percent would cover without restriction, and another 14 percent would cover with qualifications. The final category of technologies includes those for which HCFA policy explicitly denies coverage. The highest adherence to that policy is seen in claims for chelation therapy with EDTA in the treatment of ath- erosclerosis and for topical oxygen therapy (e.g., using a Topox device) in the treatment of decubitus ulcers. The lower rate of claims denial for a pacemaker de- vice for a patient with sinus bradycardia without symptoms may reflect the policy of at least some con- tractors to submit all pacemaker claims to indepth medical review, prior to making coverage determina- tions. The primary reason for covering elec- troencephalographic (EEG) monitoring during open- heart surgery was documentation by a physician that it was needed. Some contractors noted the difficulty of determining whether a monitor was used, indicating that they do not usually receive a separate bill for this, but that the bill for the surgery may be somewhat inflated. The responses for continuous 24-hour monitoring of blood pressure reflect, in part, changing HCFA pol- icy. Use of a semiautomatic or patient-activated por- table monitor had been specifically not covered since October 1982. However, there was no explicit policy regarding coverage of the automatic, continuous monitoring device until July 1983, when coverage was explicitly denied. Some respondents may not have been aware of this recent policy issuance. Furthermore, some respondents appear to have confused the auto- matic device in the question (which includes a sensing 202 ¢ Medical Technology and Costs of the: Medicare Program apparatus and provides continuous monitoring and recording) with an automatic blood pressure monitor that can be covered, if prescribed by a physician for use as part of a home dialysis delivery system. Reim- bursement for the latter is limited to the amount which would be payable for a sphygmomanometer with cuff and stethoscope unless there is documentation that the patient's condition is such that conventional methods of monitoring are not successful. Both potential sources of confusion may have resulted in a greater tendency to approve coverage of automatic, continu- ous, 24-hour blood pressure monitors than would have been expected. In summary, it appears that with the exception of intraocular lens implantation, there is variation in cov- erage decisions made by Medicare contractors, regard- less of HCFA policy. Interpretation of the variation is complicated by the dependence of coverage policy on consensus within the medical community that a given technology is safe and effective. If that consensus is lacking or just emerging, it might be expected that a coverage policy, if one existed, would be lacking in specificity and inconsistently adhered to by contrac- tors. Nevertheless, differences in coverage decisions by Medicare contractors appear to be related to the clarity or specificity of HCFA policy. The variation is least in some instances in which HCFA has explicitly approved coverage. Variation is much greater for tech- nologies that HCFA intends to be denied, but for which there is no explicitly stated denial policy, as well as technologies for which coverage policy is changing. In most instances, however, the majority of contrac- tors adhere to HCFA's intentions. Contractor's Views and Interpretation of HCFA Coverage Policy Several questions explored the contractor's percep- tions of HCFA coverage policy and the processes by which it is promulgated. For example, the respondents were asked how frequently HCFA coverage transmit- tals can be implemented as written, rather than requir- ing further interpretation. As shown in table E-7, over half the respondents indicated that HCFA transmittals can always or almost always be implemented as writ- ten. When further clarification is needed, the sources of assistance tend to be the same as those utilized in making any coverage determination (see table E-4). It was noted earlier that about 88 percent of the respondents indicated that they utilize HCFA's regional office in their area for assistance in making coverage decisions. To ascertain whether there is any predicta- bility in the types of coverage questions referred to the regional office and, perhaps, to HCFA's central office, the contractors were asked whether they used any cri- teria to determine when to refer a coverage question to HCFA. The responses generally fell into two cate- gories. Some contractors appear to use the regional office rather routinely if the coverage issues appendix does not address the technology in question, HCFA policy is ambiguous, the contractors are not familiar with the technology, or “we don't know what to do.” Other contractors attempt to establish their own pol- icies and use the regional office as a last resort, i.e., only if their own decisions have been challenged by providers or if a technology is experimental and has a potential for abuse. Some contractors are guided by Table E-7.—Medicare Contractors’ Attitudes and Actions Regarding HCFA Coverage Transmittals Contractors’ response Almost Number of Never Rarely Sometimes always Always responses HCFA coverage transmittals can be implemented as written rather than requiring further interpretation......... 71% 71% 30.4% 35.7% 19.6% 56 Contractor informs HCFA regional office of its interpretations of HCFA policy ... 10.6 14.9 — 48.9 54 Number of More Less Not applicable responses Specificity of HCFA policy compared with policy for other claims reviewed by CONLIACKON oan avasmssmssvusminnes 35.7% 26.8% 3.6% 33.9% 56 SOURCE: L. K. Demlo, G. T. Hammons, J. M. Kuder, et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Technology Assessment, U.S. Congress, Oct. 28, 1983. App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies ¢ 203 the advice of their onsite HCFA representatives about when to refer coverage questions to the regional office. When contractors take the initiative to further in- terpret HCFA coverage policies themselves, most of them inform HCFA's regional office of their interpreta- tion (see table E-7). Of those respondents, most have onsite HCFA representatives to whom they automat- ically send copies of their policy interpretations. When comparing HCFA coverage policy with pol- icies governing other health insurance claims reviewed by the contractor, most contractors report that HCFA policy is as specific or more specific than policies for other insurance claims (see table E-7). About one-third of the contractors were unable to answer the question, because they review only Medicare claims and had no basis for comparison. Interrelationships Between Coverage Decisions and Contractor Characteristics To better understand the extent to which character- istics of Medicare contractors and their claims review processes may influence coverage decisionmaking, sev- eral additional analyses were performed. The re- sponses about coverage of individual technologies were categorized according to whether the respond- ent organization was an intermediary or carrier, whether the contractor was a commercial insurance company or a Blue Cross or Blue Shield plan, and according to the contractor's geographical location, the discipline and position of the individual respondent, claims volume, and the contractor’s reasons for and willingness to refer coverage questions to the regional office. Occasionally, statistically significant differences emerged; however, no meaningful trends were ob- served for any of these analyses. Contractors’ Suggestions for Improving Coverage Transmittals Given the importance of HCFA transmittals in shap- ing coverage decisions, the respondents were asked to suggest specific changes that might improve either the process by which HCFA coverage policies are trans- mitted to Medicare contractors or the content of those coverage policies. The timing of HCFA coverage trans- mittals is the primary concern of respondents. A ma- jority of them cite difficulties in implementing trans- mittals simply because they lack adequate lead time to prepare their organizations; some transmittals ar- rive after the date on which they were to become ef- fective. Suggested improvements include increasing re- liance on interim memos and guidelines transmitted prior to the final directive, as well as simply length- ening the lead time for implementation and eliminat- ing retroactive policies. Many contractors note a need for greater national consistency in interpreting coverage policies. They mention differences in policy interpretations among HCFA regional offices, which may increase contrac- tors’ vulnerability to litigation and are particularly troublesome for large commercial insurance companies that serve as carriers in several States. They also note instances in which a single contractor receives a pol- icy determination from HCFA's central office in re- sponse to an inquiry; that is not shared nationwide, the result being inconsistent interpretation of cover- age policy. These problems, as well as inadequate local resources for contractors to pursue the status of new technologies themselves, lead many contractors to rec- ommend the creation of a national coverage clearing- house. Such a clearinghouse, perhaps supported by the Government, insurance companies, and medical and specialty associations, could disseminate information on new technologies, the status of their evaluation, and implications for coverage policies. It should also be noted, however, that many contractors are quite con- tent with the status quo and prefer to utilize local resources in making their own coverage decisions. Another area of concern is the specificity of HCFA coverage transmittals. Contractors report that the pol- icies are too technical, too ambiguous, and lacking in specificity. The need for more specific criteria for covering durable medical equipment was frequently mentioned, as well as the need to update all references to a specific technology in order to eliminate incon- sistencies within the coverage issues appendix and manual. Several respondents note that the content of coverage transmittals is improving, citing the policy covering pacemakers as a good example. On the other hand, one respondent cited the pacemaker issuance as indicative of the failure of HCFA's central office to un- derstand the claims review process. In this respond- ent’s opinion, the policy governing pacemakers does not take into consideration the process by which claims are submitted and the information included on the claim, making it impossible to process the claim with- out extensive and expensive investigation. Some con- tractors thought that these issues were not worth debating, because Medicare's move to prospective pay- ment will render any changes outdated or useless. Still others felt the coverage issues appendix is clear and sufficient and that the coverage system as a whole works “remarkably well.” Summary and Conclusions This study has confirmed the conventional wisdom that there is variation in the decisions made by Medi- 204 ¢ Medical Technology and Costs of the Medicare Program care contractors regarding the coverage of particular medical technologies. There are high levels of agree- ment and near unanimity in coverage decisions for some study technologies—particularly for intraocular lens implant following cataract removal (98.2 percent covered), the use of chelation therapy in the treatment of atherosclerosis (87.0 not covered), and the use of topical oxygen therapy for decubitus ulcers (86.3 per- cent not covered). However, there are also examples of wide variations in coverage policies—particularly for a pacemaker device and its implantation for a pa- tient who has sinus bradycardia without symptoms, an implantable chemotherapy infusion device for a pa- tient with metastatic to the liver cancer, and external insulin infusion pump for a diabetic patient, and strep- tokinase administration at cardiac catheterization to dissolve a clot in a patient with acute myocardial in- farction. Despite these variations in coverage policies, there is considerable uniformity among contractors in the methods by which claims are reviewed and national coverage decisions are implemented. After an initial review for completeness, most claims pass through (potentially) three levels of review that reflect differ- ing degrees of comprehensiveness, specificity, and clin- ical judgment. The levels of review generally involve clerical employees, registered nurses, physicians, or some other health professionals (pharmacist, podia- trist, occupational therapist, etc.) in that order. Al- though the specifics may vary, general considerations at each level of review are similar, as are the criteria for referring a claim from one level of review to the next. Eighty-four percent of the respondents use some type of automated screening procedure. When revi- sions or additions to the coverage issues appendix are received from HCFA, most contractors review them internally and modify their claims review procedures accordingly. Coverage issues requiring further clarifi- cation are pursued either locally or nationally; HCFA's regional offices are the most frequently utilized sources of clarification and interpretation. Similarly, the methods by which contractors iden- tify new technologies or new uses of established tech- nologies are quite predictable. Most contractors report that they learn about new technologies for which cov- erage questions might be raised through inquiries from providers and manufacturers prior to the actual sub- missions of claims, drug and device approval lists from FDA, and HCFA announcements. Once claims are submitted, new technologies are identified because of the absence of code numbers or the presence of codes that are not recognized by the claims reviewers. Al- though a few contractors conduct extensive research to assist in identifying and making coverage decisions for such claims, the majority assume a more passive role. The majority are reasonably well satisfied with existing methods for identifying new technologies and report that this is not a big problem for them. Mak- ing coverage determinations for new technologies is a problem, however. Some characteristics of the contractors included in the study did vary. The respondents include a mix of intermediaries and carriers; some are commercial in- surance companies and others are BC/BS plans; they come from different geographic locations, handle dif- ferent volumes of claims, and show different tenden- cies and reasons for referring coverage questions to HCFA's regional office, rather than attempting to re- solve the issue themselves. Nevertheless, these char- acteristics were not systematically related to variations in coverage decisions in any meaningful way. The most illuminating approach to examining varia- tions in coverage decisions was based on coverage pol- icy, according to the following categories: technologies that HCFA explicitly covers; those that HCFA covers with qualifications; technologies that HCFA explicitly denies; those that HCFA denies, but not explicitly; and those for which HCFA has no explicit policy, but rather, the local contractor is expected to determine coverage policy. There is variation in every category of HCFA coverage policy, with the possible exception of claims for intraocular lens implantation. Interpreta- tion of that variation is complicated by the dependence of coverage policy on consensus within the medical community that a given technology is safe and effec- tive. If that consensus is lacking or just emerging, it might be expected that a coverage policy, if one ex- isted, would be lacking in specificity and inconsistently adhered to by contractors. Nevertheless, the differ- ences in coverage decisions by Medicare contractors appear to be related, at least in part, to the clarity or specificity of HCFA policy. The variation is least in instances in which HCFA has explicitly approved cov- erage. Variation is much greater for technologies that HCFA intends to be denied, but for which there is no explicitly stated denial policy, as well as technologies for which coverage policy has recently changed or is expected to change. Despite the variation, however, when the responses for individual technologies are analyzed, the preponderance of respondents tend to adhere to HCFA intentions. If less variation in coverage decisions is desired, the study findings suggest some ways in which this might be accomplished. Greater specificity and uniform in- terpretation of wording would help to lessen the uncer- tainty about the technologies that HCFA intends to be denied, but for which denial is not explicitly stated in written policy. Similarly, uniform and more timely App. E—Decisionmaking by Medicare Contractors for Coverage of Medical Technologies ¢ 205 communication from HCFA might lessen the variation in coverage for technologies that are currently being reviewed by PHS or for which policy is expected to change. The respondents’ suggestions about a national coverage clearinghouse should also be considered. Reducing variation in coverage decisionmaking should help to eliminate any existing inequities in the availability of benefits to persons eligible for Medicare. To the extent that a technology is not intended to be covered, greater adherence to HCFA policy would also save money. If changes in coverage policy are intended to influence the proliferation and use of medical tech- nology so as to result in significant cost savings, how- ever, more major revisions are needed in the way in which coverage determinations are made. The cost ef- fectiveness of technologies would have to assume a more prominent role in coverage deliberations, as well as the possibility of limiting the health care settings and providers who would be eligible to claim reim- bursement. The potential for increased equity of benefits and financial savings that might accrue from more uniform coverage decisions must be balanced against the po- tential negative effects. It is unlikely that any na- tionally determined, uniform coverage policy can ever take into consideration the uniquely personalized needs of all patients. In some small number of cases, in- creased uniformity of coverage probably would occur at the expense of quality of care. Carried to the ex- treme, increased explicitness of coverage policy may have a negative effect on access to care and the po- tential for innovation in medicine. Furthermore, the costs of implementing and enforcing a uniform sys- tem might outweigh the advantages. In the final analysis, the decision to reduce variation in coverage policy and increase the explicitness and uniformity of Medicare benefits requires careful judg- ment and balance. Some lessening in that variation ap- pears to be desirable and achievable, provided it is carefully coordinated with forthcoming changes in the overall reimbursement system. However, it is unlikely that HCFA policy can ever be so precise as to achieve a totally uniform interpretation and implementation of Medicare coverage policy throughout the Nation. prreeem— Semmes me ————p References References 10. 11. 12. 13. 14. . Adams, K., American Medical Association, Cen- ter for Health Policy Research, personal commu- nication, June 1983. . Altman, S. H., and Socholitzky, E., “The Cost of Ambulatory Care in Alternative Settings: A Review of Major Research Findings,” Annual Re- view of Public Health 2:117-143, 1981. . Altman, S. H., and Wallack, S. S., “Technology on Trial—Is It the Culprit Behind Rising Health Costs? The Case For and Against,” Medical Tech- nology: The Culprit Behind Health Care Costs? S. H. Altman and R. Blendon (eds.), proceedings of the 1977 Sun Valley Forum on National Health, U.S. Department of Health, Education, and Welfare, publication No. (PHS) 79-3216 (Washington, D.C.: U.S. Government Printing Office, August 1979). . American College of Surgeons, Surgical Practice Department, Second Surgical Opinion Programs: A Review and Progress Report (Washington, D.C.: ACS, March 1982). . American College of Surgeons, Surgical Practice Department, Socio-Economic Factbook for Sur- gery, 1982, J. N. Haug and R. Seeger (eds.) (Chi- cago: ACS, 1982). . American Enterprise for Public Policy Research, Conference on “Strategies for Hospital Cost Con- tainment: A Look at the All-Payer Model,” Washington, D.C., Nov. 21, 1983. . American Hospital Association, Division of Fi- nancial Management, “Report on Budget/Rate Review Programs,” update report, January 1979. . American Hospital Association, Division of Health Financing Policy, Office of Public Policy Analysis, “Hospital-Blue Cross Contract Pro- visions, July 1, 1981,” AHA No. 073200, February 1982. . American Hospital Association, Division of Health Financing Policy, Office of Public Policy Analysis, “Incorporating Capital Costs Into Medicare Prospective Prices,” unpublished draft, July 1983. American Hospital Association, Hospital Statis- tics, 1983 (Chicago: AHA, 1983). American Hospital Association, Hospital Statis- tics, 1982 (Chicago: AHA, 1982). American Hospital Association, Hospital Statis- tics, 1978 (Chicago: AHA, 1978). American Hospital Association, Technology Evaluation and Acquisition Methods for Hospi- tals (Chicago: AHA, 1979). American Society for Parenteral and Enteral Nu- trition, Public and Private Insurance Coverage 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. of Parenteral and Enteral Nutrition Services, background paper prepared by Meml, Jacobs, Pierne & Gersh, Washington, D.C., March 1983. Andersen, G., and Ginsberg, P. B., “Prospective Capital Payments to Hospitals,” Health Affairs 2(3):52-63, fall 1983. Anderson, J. G., and Jay, S. ]J., “Computer Ap- plications in Medicine: Adoption and Utiliza- tion,” D. A. B. Lindberg, E. E. van Brunt, and M. A. Jenkin (eds.) AAMSI Congress 83, pro- ceedings of the Congress on Medical Informatics (Bethesda, Md.: American Association for Med- ical Systems and Informatics, 1983). Ashby, J. L., Jr., “An Analysis of Hospital Costs by Cost Center, 1971 Through 1978,” Health Care Finan. Rev. 4:37, 1982. Austrian, R., “Some Observations on the Pneu- mococcus and on the Current Status of Pneumo- coccal Disease and Its Prevention,” Reviews of Infectious Diseases 3(Supplement):S1-17, March- April 1981. Baker, D., Health Care Financing Administra- tion, U.S. Department of Health and Human Services, personal communication, July 1983. Banta, H. D., “Computed Tomography: Cost Containment Misdirected” (editorial), Am. J. Public Health 70:215, 1980. Banta, H. D., “Notes of a Peripatetic Correspond- ent,” The Sorcerer's Apprentice 3(2):2, April 1983. Banta, H. D., statement before the Subcommit- tee on Investigation and Oversight, Committee on Science and Technology, U.S. House of Rep- resentatives, hearings on Organ Transplantation, Medical Technology Assessment and Resource Allocation, Apr. 27, 1983. Banta, H. D., and Behney, C. ]., “Policy For- mulation and Technology Assessment,” prepared for the Experiment and Social Policy Issue of Health and Society, Milbank Mem. Fund Q. 59(3):445, 1981. Banta, H. D., Behney, C. J., and Willems, J. S., Toward Rational Technology in Medicine, Vol- ume 5 in Springer Series on Health Care and Soci- ety (New York: Springer Publishing Co., 1981). Banta, H. D., and Kemp, K. B., The Manage- ment of Health Care Technology in Nine Coun- tries (New York: Springer Publishing Co., 1982). Banta, H. D., Ruby, G., and Burns, A. K., “Using Coverage Policy to Contain Medicare Costs,” paper prepared for the Proceedings of the Conference on the Future of Medicare, Sub- committee on Health, Committee on Ways and 209 210 ¢ Medical Technology and Costs of the Medicare Program 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 33. 39. 40. Means, U.S. House of Representatives, Feb. 1, 1984. Barer, M. L., Evans, R. G., and Stoddart, G. L., Controlling Health Care Costs by Direct Charges to Patients: Snare or Delusion? Occasional Paper 10 (Toronto: Ontario Economic Council, 1979). Bauer, R. A., and Wortzel, L. H., “Doctors Choice: The Physician and His Sources of Infor- mation About Drugs,” J. Mkt. Research 3:40, 1966. Beazley, W. S., Ill, “The Trend to Outpatient Care: I. ‘In-and-Out’ Surgery,” Virginia Medical 107:620-621, September 1980. Beck, R. H., “The Effects of Copayment on the Poor,” J. Hum. Resour. 11(1):129, 1974. Bentley, D. W., “Pneumococcal Vaccine in the Institutionalized Elderly: Review of Past and Re- cent Studies,” Reviews of Infectious Diseases 3(Supplement):S61-70, March-April 1981. Berenson, R., Washington, D.C., personal com- munication, April 1983. Berk, A. A., and Chalmers, T. C., “Cost and Ef- ficacy of the Substitution of Ambulatory for In- patient Care,” N. Eng. J. Med. 304(7):393-97, Feb. 12, 1981. Berk, M. L., and Wilensky, G. R., “Health Care of the Poor Elderly: Supplementing Medicare,” NCHSR National Health Care Expenditures Study, U.S. Department of Health and Human Services, Public Health Service, National Center for Health Services Research, 1983. Berry, R., “A Review of the Medicare Economic Index: An Analysis of Variation in Practice Costs Among Specialties,” in Issues in Physician Reim- bursement, N. T. Greenspan (ed.), HCFA pub- lication No. 03121, U.S. Department of Health and Human Services, Health Care Financing Ad- ministration, August 1981. Biles, B., Schramm, C. J., and Atkinson, J. R., “Hospital Cost Inflation Under State Rate-Setting Programs,” N. Eng. J]. Med. 303(12):644-67, Sept. 18, 1980. Biomedical Business International 6(10):94, June 7, 1983. Blagg, C. R., and Scribner, B., “Medicare End- Stage Renal Disease Program: More Than a Bil- lion Dollar Question,” Ann. Int. Med. 93:501, 1980. Block, J. A., Brideau, L. C., Burns, A. K., et al., “Hospital-Sponsored Primary Care: The Com- munity Hospital Program,” Journal of Am- bulatory Care Management, February 1980. Block, J. A., Regenstreif, D. I., and Shute, L. J., “Hospital Experimental Payments Program: It’s 41. 42. 43 46. 47. 48. 49. 50. 51. 52. 53. 54. Working in Rochester,” Hospital Financial Man- agement, September 1981. Blue Cross and Blue Shield Association, “Back- ground on the Blue Cross and Blue Shield Asso- ciation’s Medical Necessity Program,” Blue Cross/Blue Shield press briefing, Washington, D.C., Oct. 21, 1982. Blue Cross of Massachusetts, Inc., meeting of the Central Professional Service Committee of the Blue Shield Board of Directors, July 21, 1976. . Blue Sheet, June 23, 1983. 44. 45. Blue Sheet, Mar. 30, 1983. Blumenthal, D., “Federal Policy Toward Health Care Technology: The Case of the National Cen- ter,” Milbank Mem. Fund Q./Health & Society 61(4):584, 1983. Bombardier, C., Fuchs, V. R., Lillard, L. A., et al., “Socioeconomic Factors Affecting the Utili- zation of Surgical Operations,” N. Eng. J. Med. 297(13):699, Sept. 29, 1977. Brady, J. F., St. Joseph's Hospital and Medical Center, panelist for session on “Living Under an All-Payer System — Views From the Hospital Sec- tor,” at conference on “Strategies for Hospital Cost Containment: A Look at the All-Payer Model,” American Enterprise Institute for Pub- lic Policy Research, Washington, D.C., Nov. 21, 1983. Brook, R. H., and Lohr, K. N., “Second-Opinion Programs: Beyond Cost-Benefit Analysis,” Med- ical Care 20(1):1-2, January 1982. Brook, R. H., Ware, J. E., Rogers, W. H., et al., “Does Free Care Improve Adults’ Health? Results From a Randomized Controlled Trial,” N. Eng. J. Med. 309:1426-34, Dec. 8, 1983. Brook, R. H., Williams, K. N., and Rolph, J. E., “Controlling the Use and Cost of Medical Serv- ices: The New Mexico Experimental Medical Care Review Organization—A Four-Year Case Study,” Medical Care 16(9) Supplement, September 1978. Brown, M., and McCool, B. P., “Changes in the Ownership and Governance of America’s Hos- pitals,” paper prepared for the Office of Technol- ogy Assessment, U.S. Congress, contract No. 233-8460.0, unpublished draft, 1983. Bunker, J. P., Stanford, Calif., personal commu- nication, June 16, 1981. Bunker, J. P., Barnes, B. A., Mosteller, F. (eds.), Costs, Risks, and Benefits of Surgery (New York: Oxford University Press, 1977). Bunker J. P., Fowles, J., and Schaffarzick, R., “Evaluation of Medical Technology Strategies: Evaluation of Coverage and Reimbursement,” N. Eng. J. Med. 306(10):620-24, Mar. 11, 1982. References ¢ 211 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. Burns, L. A., and Ferber, M. S., “Freestanding Emergency Care Centers Create Public Policy Issues,” Hospitals 55(10):73-76, May 16, 1981. Burrier, B., Inspector General's Office of Health Financing Integrity, Health Care Financing Ad- ministration, U.S. Department of Health and Human Services, personal communication, Aug. 12, 1983. Campion, E. W., Mulley, A. G., Goldstein, R. L., et al., “Medical Intensive Care for the Elderly: A Study of Current Use, Costs, and Outcomes,” JJ A.M. A. 246(18):2052, Nov. 6, 1981. Carter, E., National Center for Health Services Research, U.S. Department of Health and Human Services, personal communication, Mar. 12, 1984. Cassidy, R., “Will the PPO Movement Freeze You Out?” Medical Economics 60:262-70,274, Apr. 18, 1983. Chassin, M., “Containment of Hospital Costs: A Strategic Assessment, Chapter V, Utilization Re- view,” Medical Care 16(10):27-35, October 1978. Chassin, M., personal communication, Feb. 17, 1984. Clegg, J., Health Care Financing Administration, U.S. Department of Health and Human Services, personal communication, Aug. 9 and 17, 1983. Cocco, A. E., and Cocco, D. V., “A Survey of Cimetidine Prescribing,” Medical Intelligence 304(21):1281, May 21, 1981. Coelen, C., and Sullivan, D., “An Analysis of the Effects of Prospective Reimbursement Pro- grams on Hospital Expenditures,” Health Care Finan. Rev. 2(3):1-39, winter 1981. Cohen, A. B., and Cohodes, D., “Certificate of Need and Low Capital-Cost Medical Technol- ogy,” Milbank Mem. Fund Q. 60(2):307-28, 1982. Cohen, D. I., Jones P., Littenberg, B., et al., “Does Cost Information Availability Reduce Phy- sician Test Usage?” Medical Care 20(3):286-91, March 1982. Cohen, D. I, Littenberg, B., Wetzel, C., et al., “Improving Physician Compliance With Preven- tive Medicine Guidelines,” Medical Care 20(10):140-45, October 1982. Cohen, H. A., Solnick, M., and Stephenson, A., “On the Quality of Life: Some Philosophical Re- flections,” Circulation 66(5) Part II, AHA Mono- graph No. 92, Technology Assessment Forum on Coronary Artery Bypass Surgery: Economic, Ethical and Social Issues, November 1982. Cohodes, D. R., “The State Experience With Cap- ital Management and Capital Expenditure Review Programs,” article cited in “Health Capital 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. Issues,” papers presented at the Health Capital - Conference sponsored by the Bureau of Health Facilities, Feb. 19-20, 1980. Cohodes, D. R., “Which Will Survive? The $150 Billion Capital Question,” Inquiry 20(1):5-11, spring 1983. Cohodes, D. R., and Kinkead, B. M., “Hospital Capital Formation in the 1980's: Is There a Cri- sis?” The Johns Hopkins Medical Institutions, prepared for the Robert Wood Johnson Founda- tion, Princeton, N.J., Aug. 1, 1982. Collen, M. F., “Technology Assessment in Pre- paid Group Practice,” prepared for the Commit- tee on Technology Assessment, Institute of Medi- cine, National Academy of Sciences, unpublished draft, revised May 1982. Collins, S., Boston Regional Office, Health Care Financing Administration, U.S. Department of Health and Human Services, personal commu- nication, December 1982. Commerce Clearing House, Medicare and Med- icaid Guide, pp. 390, 9011-3, 35-60, May 1983. Conrad, D. A., University of Washington, pan- elist for session on “Living Under an All-Payer System—Views From the Hospital Sector,” at conference on “Strategies for Hospital Cost- Containment: A Look at the All-Payer Model,” American Enterprise Institute for Public Policy Research, Washington, D.C., Nov. 21, 1983. Cromwell, J., and Burstein, P., “Physician Losses From Medicare/Medicaid Participation: How Real Are They?” paper prepared for the U.S. De- partment of Health and Human Services, Health Care Financing Administration, Office of Re- search, June 1982. Cromwell, J., and Kanak, J., “The Effects of Pro- spective Reimbursement Programs on Hospital Adoption and Service Sharing,” Health Care Finan. Rev. 4(2):67, December 1982. Cummings, K. M., Frishof, K. B., Long, M. ]., et al., “The Effects of Price Information on Phy- sicians’ Test-Ordering Behavior,” Medical Care 20(3):293-301, March 1982. Daniels, M., and Schroeder, S. A., “Variation Among Physicians in Use of Laboratory Tests. II. Relation to Clinical Productivity and Out- comes of Care,” Medical Care 15:482-87, 1977. Davis, C. K., testimony of the Administrator, Health Care Financing Administration, U.S. De- partment of Health and Human Services, before the 97th Congress of the United States, Second Session Joint Hearing of the House of Represent- atives Select Committee on Aging and the Sub- committee on Health and Long Term Care, Mar. 212 e Medical Technology and Costs of the Medicare Program 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 19, 1982, publication No. 97-346 (Washington, D.C.: U.S. Government Printing Office, 1982). Davis, K., “Medicare Reconsidered,” paper pre- sented for Duke University Medical Center Sev- enth Sector Conference on the Financial Support of Health Care of the Elderly and the Indigent, Durham, N.C., Mar. 14-16, 1982. Davis, K., “Rising Hospital Costs: Possible Causes and Cures,” Bull. N. Y. Acad. Med. 48(11):1354-71, December 1972. Davis, J. E., and Detmer, D. E., “The Ambula- tory Surgical Unit,” Ann. Surg. 175(6):856-62, June 1982. Davis, K., and Rowland, D., “Reforming Medi- care: A New Approach to Financing,” paper pre- pared for the Proceedings of the Conference on the Future of Medicare, Subcommittee on Health, Committee on Ways and Means, U.S. House of Representatives, Feb. 1, 1984. Demkovich, L. E., “Private Sector Moves in as Washington Ends Its Financial Assistance for HMOs,” National J. 15:1757-89, 1983. Demlo, L. K., Hammons, G. T., Kuder, J. M., et al., “Report of a Study on Decisionmaking by Medicare Contractors for Coverage of Medical Technologies,” prepared for the Office of Tech- nology Assessment, U.S. Congress, Oct. 28, 1983. Dennison, C., National Hospital Discharge Survey, National Center for Health Statistics, U.S. Department of Health and Human Services, personal communication, December 1983. Desmarais, H. R., Bureau of Eligibility, Reim- bursement and Coverage, Health Care Financing Administration, U.S. Department of Health and Human Services, statement before the Subcom- mittee on Health and Long-Term Care, Select Committee on Aging, U.S. House of Represent- atives, Mar. 15, 1984. Devitt, J. E., and Ironside, M. R., “Can Patient Care Audit Change Doctor Performance?” J. Med. Educ. 50:1122-23, 1975. Dixon, R. H., and Laszlo, J., “Utilization of Clin- ical Chemistry Services by Medical House Staff: An Analysis,” Arch. Int. Med. 134:1064-67, 1974. Donabedian, A., “Effects of Medicare and Med- icaid on Access to and Quality of Health Care,” Pub. Hith. Rpts., July-August, 1976. Drake, D. F., “Does Money Spent on Health Care Really Improve U.S. Health Status?” Hospitals 52(20), Oct. 16, 1978. Dresnick, S. J., Roth, W. I., Linn, B. S., et al., “The Physician's Role in the Cost-Containment Problem,” J. A.M.A. 241(15):1606-9, Apr. 13, 1979. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. Duke Law Journal, “The Medical Malpractice Threat: A Study of Defensive Medicine,” pp. 939- 993, 1971. Edelson, E., “Cleansing the Blood,” Science 82 (American Association for the Advancement of Science, Washington, D.C.) 3(8):72-77, October 1982. Eggers, P. W., Connerton, R., and McMullan, M., “The Medicare Experience With End-Stage Renal Disease: Trends in Incidence, Prevalence, and Survival,” working paper, Office of Research and Demonstrations, Health Care Financing Ad- ministration, U.S. Department of Health and Human Services, May 10, 1983. Eggers, P. W., and Prihoda, R., “Pre-Enrollment Reimbursement Patterns of Medicare Benefici- aries Enrolled in ‘At-Risk’ HMOs,” Health Care Finan. Rev. 4(1):55-73, September 1982. Eisenberg, J. M., “The Use of Ancillary Services: A Role for Utilization Review?” Medical Care 20(8):849-60, August 1982. Elliott, W. B., Strand, H. M., Meyers, F. H., et al., “The Hospital Capitation Payment Project: New Incentives and Tools for Cost Contain- ment,” Inquiry 20(2):114-20, summer 1983. Ellwein, L. K., and Gregg, D. D., “An Intro- duction to Preferred Provider Organizations (PPOs),” InterStudy, Excelsior, Minn., revised June 1982, mimeographed. Elstein, A. S., Rovner, D. R., and Rothert, M. L., “A Preclinical Course in Decision Making,” Medical Decision Making 2(2):209-216, 1982. Enthoven, A. C., Health Plan: The Only Prac- tical Solution to the Soaring Cost of Medical Care (Reading, Mass.: Addison-Wesley Publishing Co., 1980). Evashwick, C., “Long-Term Care: Long-Term Care Becomes Major New Role for Hospitals,” Hospitals 56::50-55, July 1, 1982. Feder, J. M., Medicare: The Politics of Federal Hospital Insurance (Lexington, Mass.: D.C. Heath & Co., 1977). Feder, J., Hadley, J., and Mullner, R., “Poor Peo- ple and Poor Hospitals: Implications for Public Policy,” J. Health Polit. Policy Law, forthcoming. Feder, J., and Scanlon, W., “The Underused Ben- efit: Medicare's Coverage of Nursing Home Care,” Milbank Mem. Fund Q. 60(4):604-632, fall 1982. Federal Register 45(153):52296-52297, Aug. 6, 1980. Federal Register 47(151), Aug. 5, 1982. Federal Register 47:36742-37844, Aug. 26, 1982. Federal Register 48(13):2444, Jan. 19, 1983. Federal Register 48:21254-21291, May 11, 1983. References ¢ 213 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 124a. 125. Federal Register 48(191), Oct. 1, 1983. Federal Register 48(243):56008, Dec. 16, 1983. Federal Register 49(1):282, Jan. 3, 1984. Feldstein, M. S., Hospital Costs and Health In- surance (Cambridge, Mass.: Harvard University Press, 1981). Feldstein, M. S., The Rising Cost of Hospital Care (Washington, D.C.: Information Resources Press, 1971). Feldstein, M. S., and Taylor, A., The Rapid Rise of Hospital Costs, GPO stock No. 052-003- 003.03.1 (Washington, D.C.: U.S. Government Printing Office, January 1977). Ferry, T. P., Gornick, M., Newton, M., et al., “Physicians’ Charges Under Medicare: Assign- ment Rates and Beneficiary Liability,” Health Care Finan. Rev. 2:49-73, winter 1980. Fineberg, H. V., “Background Session on Ulcer Disease: Epidemiology, Diagnosis, and Treat- ment,” app. A in Cost-Benefit and Cost- Effectiveness Analysis in Policymaking: Cime- tidine as a Model, B. S. Bloom (ed.), proceedings of an International Symposium, Nov. 22-24, 1981, Tarpon Springs, Fla. (New York: Biomedi- cal Information Corp. Publications, 1982). Fineberg, H. V., “Clinical Chemistries: The High Cost of Low-Cost Diagnostic Tests,” Medical Technology: The Culprit Behind Health Care Costs? S. H. Altman and R. Blendon (eds.), pro- ceedings of the 1977 Sun Valley Forum on Na- tional Health, U.S. Department of Health, Edu- cation, and Welfare, publication No. (PHS) 79-3216 (Washington, D.C.: U.S. Government Printing Office, 1979). Fineberg, H. V., and Pearlman, L. A., Benefit and Cost Analysis of Medical Interventions: The Case of Cimetidine and Peptic Ulcer Disease, OTA- BP-H-9(11) (Washington, D.C.: U.S. Govern- ment Printing Office, September 1981). Fineberg, H. V., and Pearlman, L. A., “Surgical Treatment in Peptic Ulcer in the United States,” Lancet 1(8252):1305-7, June 13, 1981. Finkel, M. L., McCarthy, E. G., and Miller, D., “Podiatric Surgery: The Need for a Second Opin- ion,” Medical Care 20(8):862-70, August 1982. Fisher, C. R., “Differences by Age Groups in Health Care Spending,” Health Care Finan. Rev. 1(4):65-90, spring 1980. Flood, A. B., Scott, W. R., and Ewy, W., “Does Practice Make Perfect? Part I: The Relation Be- tween Hospital Volume and Outcomes for Selected Diagnostic Categories,” Medical Care 22(2):98-114, February 1984. Ford, J. L., “Outpatient Surgery: Present Status 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. and Future Projections,” South. Med. J. 71(3): 311-15, March 1978. Freeland, M. S., and Schendler, C. E., “National Health Expenditure Growth in the 1980's: An Aging Population, New Technologies, and In- creasing Competition,” Health Care Finan. Rev. 5(3):1-58, March 1983. Friedlob, A. S., “Medicare Second Surgical Opin- ion Programs: The Effect of Waiving Cost-Shar- ing,” Health Care Finan. Rev. 4(1):99-106, Sep- tember 1982. Fries, J. F., “The Compression of Morbidity,” Milbank Mem. Fund Q./Health and Society 61(3):397, 1983. Fuchs, V. R., “Foreword” in Cost-Benefit and Cost-Effectiveness Analysis in Health Care: Prin- ciples, Practice, and Potential, K. E. Warner and B. R. Luce (eds.) (Ann Arbor, Mich.: Health Administration Press, 1982). Fuchs, V. R. (ed.), “The Growing Demand for Medical Care,” Essays in the Economics of Health and Medical Care (New York: Columbia Univer- sity Press, 1972). Garnick, D. W., “The Impacts of Rate Regula- tion and Unionization on the Demand for Hos- pital Labor,” prepared as part of the Hospital Cost and Utilization Project for the National Cen- ter for Health Services Research, U.S. Depart- ment of Health and Human Services, unpublished draft, January 1983. Gaus, C. R., and Cooper, B. S., “Technology and Medicare: Alternatives for Change,” background paper prepared for the Conference on Health Care Technology and the Quality of Care, Bos- ton University, Nov. 19-20, 1976 (Boston, Mass. : Boston University, Program on Public Policy for Quality of Health Care, 1976). Gertman, P. M., Stackpole, D. A., Levenson, D. K., et al., “Second Opinions for Elective Surgery: The Mandatory Medicaid Program in Massachu- setts,” N. Eng. J. Med. 302(21):1169-74, May 22, 1980. Gibson, R. M., “National Health Expenditures, 1978,” Health Care Finan. Rev. 1:1-36, summer 1979. Gibson, R. M., Waldo, D. R., and Levit, K. R., “National Health Expenditures, 1982,” Health Care Finan. Rev. 5(1):1-32, fall 1983. Gilbert, R. N., “Competition Spurs Ambulatory Choice,” Hospitals 57(10):67-68, May 16, 1983. Gillick, M. R., “Is the Care of the Chronically Ill a Medical Prerogative?” N. Eng. J. Med. 310(3):190, Jan. 19, 1984. Ginsburg, P., and Mannheim, L., “Insurance, 214 ® Medical Technology and Costs of the Medicare Program 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. Copayment and Health Services Utilization, A Critical Review,” Journal of Economics and Busi- ness 25(2):142-53, spring-summer 1973. Goodgame, J. T., Jr., “A Critical Assessment of the Indications for Total Parenteral Nutrition,” Surg. Gynecol. Obstet. 151(3):433-441, Septem- ber 1980. Goran, M. ]., “The Evolution of the PSRO Hos- pital Review System,” Medical Care 18(5) Sup- plement, 1979. Graham, R., Director, Health Systems, National Blue Cross/Blue Shield, personal communication, December 1983. Greater Detroit Area Hospital Council, “Greater Detroit Area Hospital Council Pinpointing Sat- ellite Clinic Growth Trends,” Michigan Medicine 81(8):70-71, February 1982. Greenberg, B., and Derzon, R. A., “Determin- ing Health Insurance Coverage of Technology: Problems and Options,” Medical Care 19(10): 967-78, October 1981. Greenland, P., Mushlin, A. E., and Griner, P. F., “Discrepancies Between Knowledge and Use of Diagnostic Studies in Asymptomatic Patients,” J. Med. Educ. 54:863-69, November 1979. Greenlick, M. R., Lamb, S.]., Carpenter, T. M., Jr., et al., “Kaiser-Permanente’s Medicare Plus Project: A Successful Medicare Prospective Pay- ment Demonstration,” Health Care Finan. Rev. 4(4):85-97, summer 1983. Greer, D. S., Mor, V., Birnbaum, H., et al., “Na- tional Hospice Study, Preliminary Final Report, Extended Executive Summary,” Brown Univer- sity, prepared for the Health Care Financing Ad- ministration, U.S. Department of Health and Human Services, grant No. 99-P-97793/1-0, No- vember 1983. Grenell, B., “The Richmond Proposed Preferred Provider Organization,” Blue Cross and Blue Shield of Virginia, paper presented at American Public Health Conference, Montreal, Canada, Nov. 16, 1982. Grindstaff, G., Hirsch, B., and Silverman, H., “Medicare—Use of Prescription Drugs by Aged Persons Enrolled for Supplementary Medical In- surance, 1967-1977,” Health Care Financing Pro- gram Statistics, Health Care Financing Adminis- tration, U.S. Department of Health and Human Services, publication No. 03080 (Baltimore, Md.: HCFA, March 1981). Griner, P. F., and Glaser, R. J., “Misuse of Lab- oratory Tests and Diagnostic Procedures,” N. Eng. J. Med. 307(21):1336-39, Nov. 18, 1982. Griner, P. F., and Liptzin, B., “Use of the Lab- 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. oratory in a Teaching Hospital: Implications for Patient Care, Education, and Hospital Costs,” Ann. Int. Med. 75:157-63, 1971. Hadley, D., Statistical Information Services, Health Care Financing Administration, U.S. De- partment of Health and Human Services, per- sonal communication, Nov. 25, 1983. Hadley, J., “Background Information on the Reimbursement of Hospitals’ Teaching Expenses,” working paper No. 1302-01 (Washington, D.C.: The Urban Institute, December 1979). Hadley, J., Holahan, J., and Scanlon, W., “Can Fee-For-Service Reimbursement Coexist With De- mand Creation?” Inquiry 16:247-58, fall 1979. Hamilton, D. C., and Kamens, G., “Case Study of Prospective Reimbursement in New York,” Health Care Financing Grants and Contracts Reports, Health Care Financing Administration, U.S. Department of Health, Education and Wel- fare, contract No. 500-78-0036, April 1980. Hamilton, D., Walter, R., Cromwell, J., et al., “First Annual Report of the National Hospital Rate-Setting Study: A Comparative Review of Nine Prospective Rate-Setting Programs,” Health Care Financing Grants and Contracts Reports prepared under contract No. 500-78-0036 by ABT Associates, Inc., August 1980). Hanft, R. S., Washington, D.C., personal com- munication, Dec. 28, 1983. Harris, D. L., and de Lashmutti-Robbins, M., “The Ambulatory Response,” Hospitals 57(10): 69-71, May 16, 1983. Heimberger, H. C., “Home Parenteral Nutrition,” Ala. J. Med. Sci. 19(4):377-80, October 1982. Helbing, C., “Medicare: Use of Skilled Nursing Facilities, 1976-1977,” Health Care Financing Notes, Health Care Financing Administration, U.S. Department of Health and Human Services, publication No. 03021 (Baltimore, Md.: HCFA, 1980). Helbing, C., Office of Research, Division of Ben- eficiary Studies, Health Care Financing Admin- istration, U.S. Department of Health and Human Services, 1979 MedPar data xerox, undated. Helbing, C., Supervisory Statistician, Office of Research, Division of Beneficiary Studies, Health Care Financing Administration, U.S. Department of Health and Human Services, personal commu- nication, June 6, 1983. Helbing, C., “Ten Years of Short-Stay Hospital Utilization and Costs Under Medicare: 1967- 1976,” Health Care Financing Research Report, Health Care Financing Administration, U.S. De- partment of Health and Human Services, publica- References ¢ 215 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. tion No. 03053 (Baltimore, Md.: DHHS, August 1980). Hellinger, F. J., “Controlling Costs by Adjusting Payment for Medical Technologies,” Inquiry 19:34-43, spring 1982. Hendricks, J., and Hendricks, C. D., Aging in Mass Society: Myths and Realities (Cambridge, Mass.: Winthrop Publishers, Inc., 1977). Hershey, N., “The Defensive Practice of Medi- cine: Myth or Reality,” Milbank Mem. Fund Q. 50(1):69-98, January 1972. Heyse, S. P., National Center for Health Care Technology, “Geriatric Herniorrhapies at the Shouldice Hospital, Ontario, Canada,” memo- randum to Deputy Director, Bureau of Program Policy, Health Care Financing Administration, U.S. Department of Health, Education, and Wel- fare, Oct. 23, 1980. Himler, G., “PSRO, Cost Versus Quality,” N. Y. St. J. Med. 79:1863-65, November 1979. Hirschmann, J. V., and Lipsky, B. A., “Pneumo- coccal Vaccine in the United States—A Critical Analysis,” J. A.M.A. 246(13):1428-32, Sept. 25, 1981. Hodges, A., Study Director, Institute of Medi- cine, National Academy of Sciences, personal communication, December 1983. Holahan, J., Scanlon, W., Hadley, J., et al., “The Effect of Medicare/Medicaid Reimbursement on Physician Behavior,” in Physicians and Financial Incentives, J. R. Gabel, J. Taylor, N. T. Green- pan, and M. O. Blaxall (eds.), U.S. Department of Health and Human Services, Health Care Financing Administration, publication No. 03067, December 1980. Hospital Corporation of America and Parthenon Insurance Co., “Product Standardization Man- ual,” internal document, undated. Hospitals, (Headlines) “Guy Defends Medi-Cal Contracting,” 57(7):74-80, Apr. 1, 1983. Hoyer, T., Health Care Financing Administra- tion, U.S. Department of Health and Human Services, personal communication, January 1983. Hsiao, W. C., and Stason, W. B., “Toward De- veloping a Relative Value Scale for Medical and Surgical Services,” Health Care Finan. Rev. 1(3):23-38, 1979. Hudson, J. I., and Braslow, J. B., “Cost Contain- ment Education Efforts in United States Medical Schools,” J. Med. Educ. 54:835-40, November 1979. Hughes, S. L., “Ensuring Quality in Home Care Services,” Hospitals 56(21)74-80, Nov. 1, 1982. Hultgren, R., Government Accounting Office, U.S. Congress, personal communication, Jan. 23, 1984. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. IMS America, Ltd., “The Importance of Govern- ment Funded Social Welfare Programs to Private Medical Treatment, 1976-1981,” Bulletin #3, IMS Research at Work for Our Panel Participants, based on data collected for the National Disease and Therapeutic Index, American College of Phy- sicians, IMS America, Ltd.; Ambler, Pa., undated. Jensen, G., Kidder, D., Lee, J., et al., “Quan- tification of Factors Causing Variations in Hos- pital Staffing and Analysis of Monthly AHA Hos- pital Data,” prepared for ABT Associates, Inc. Cambridge, Mass., under contract No. HRA 232- 79-0076, Oct. 21, 1980. Jones, D., Health Insurance Association of Amer- ica, personal communication, Sept. 6, 1983. Journal of the American Medical Association, “Squaring Off Over Total Parenteral Nutrition” (Medical News section), 243(16):1610-16, Apr. 25, 1980. Joskow, P. L., Controlling Hospital Costs: The Role of Government Regulation (Cambridge, Mass.: MIT Press, 1981). Kambric, H., and Nose, Y., Plasmapheresis: His- torical Perspective, Therapeutic Applications, and New Frontiers (Cleveland, Ohio: Interna- tional Center for Artificial Organs and Trans- plantation, 1982). Kane, F., Health Care Financing Administration, U.S. Department of Health and Human Services, personal communication, Aug. 11 and 17, 1983. Keane, J., Maryland Health Service Cost Review Commission, personal communication, May 5, 1983. Kennedy, J. W., Killip, T., Fisher, L. D., et al., “The Clinical Spectrum of Coronary Artery Dis- ease and Its Surgical and Medical Management, 1974-1979. The Coronary Artery Surgery Study,” Circulation 66(5) Part II, November 1982, AHA Monograph No. 92, Technology Assessment Forum on Coronary Artery Bypass Surgery: Eco- nomic, Ethical and Social Issues. Kidder, D., and Sullivan, D., “Hospital Payroll Costs, Productivity, and Employment Under Pro- spective Reimbursement,” Health Care Finan. Rev. 4(2):89-99, December 1982. Kinney, E. D., and Lefkowitz, B., “Capital Costs: Reimbursement to Community Hospitals Under Federal Health Insurance,” Jnl. Hith. Polit. Pol- icy Law, fall 1982. . Kleinfield, N. R., “The Home Health Care Boom,” The New York Times, June 30, 1983, p. D1. Knaus, W. A., and Wagner D. P., “Intensive Treatment for the Elderly: Comment,” letter to the editor, J A.M.A. 247:3185, 1982. 216 » Medical Technology and Costs of the Medicare Program 191 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. . Kodner, K., “Competition: Getting a Fix on PPOs,” Hospitals 56(22):59-66, Nov. 16, 1982. Kolata, G., “Some Bypass Surgery Unnecessary,” Science 222(4624):605, Nov. 11, 1983. Korsak, A., Health Insurance Association of America, personal communication, December 1983. Kovar, M. G., “Elderly People: The Population 65 Years and Over,” Health— United States, 1976- 77, U.S. Department of Health, Education, and Welfare, publication No. (HRA) 77-1232 (Wash- ington, D.C.: DHEW, December 1977). Krakauer, H., Grauman, J. S., McMullan, M. R., et al., “The Recent U.S. Experience in the Treat- ment of End-Stage Renal Disease by Dialysis and Transplantation,” N. Eng. J. Med. 308(26):1558- 63, June 30, 1983. Langenbrunner, J. C., “Technology Assessment in the U.S. Private and Nonprofit Sectors: Part 1,” The Sorcerer's Apprentice 3(2):5-7, April 1983. Lawrence, R. S., “The Role of Physician Educa- tion in Cost Containment,” J. Med. Educ. 54:841- 47, November 1979. Lelah, T., and Brook, R. H., “Ischemic Heart Dis- ease: Assessing the Quality of Medical Care Using Short-Term Outcome Measures,” in Quality of Medical Care Using Short-Term Outcome Meas- ures: Eight Disease-Specific Applications, A. D. Avery, et al. (eds.) (Santa Monica, Calif.: Rand Corp., 1976). Link, C.R., Long, S. H., and Settle, R. F., “Cost Sharing, Supplementary Insurance and Health Services Utilization Among the Medicare Elder- ly,” Health Care Finan. Rev. 2:25-31, fall 1980. Lion, J., and Altman, S., “Case-Mix Differences Between Hospital Outpatient Departments and Private Practice,” Health Care Finan. Rev. 4(1):89-98, September 1982. Lipscomb, J., Chapel Hill, N.C., personal com- munication, Feb. 16, 1984. Lohr, K. N., Winkler, J. D., and Brook, R. H., Peer Review and Technology Assessment in Med- icine (Santa Monica, Calif.: Rand Corp., August 1981). Long, S. H., Settle, R. F., and Link, C. R., “Who Bears the Burden of Medicare Cost Sharing,” In- quiry 19:222-34, fall 1982. Luft, H. S., “Assessing the Evidence on HMO Performance,” Milbank Mem. Fund Q. 58(4):501- 36, 1980. Luft, H. S., Health Maintenance Organizations: Dimensions of Performance (New York: John Wiley & Sons, 1981). Luft, H. S., “How Do Health Maintenance Or- 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. ganizations Achieve Their Savings,” N. Eng. J. Med. 298(24):1336-43, June 15, 1978. Lundberg, C. J., “New Tax Bill Mandates Changes for HHAs,” Hospitals 56(21):81-84, Nov. 1, 1982. Lyman, R., Health Care Financing Administra- tion, U.S. Department of Health and Human Services, personal communication, June 20, 1983. MacEwan, D. W., Kavanagh, S., Chow, P., et al., “Manitoba Barium Enema Efficacy Study,” Radiology 126:39-44, 1978. Mannheim, L. M., and Friedman, B., “Factors Contributing to the Phenomenal Growth of Medicare Expense for Hospital-Based Ambu- latory Care,” Inquiry 19:295-307, winter 1982. Margulies, H., Office of Health Technology Assessment, National Center for Health Services Research, U.S. Department of Health and Human Services, personal communication, June 24, 1983. Martin, A. R., Wolf, M. A., Thibodeau, L. A., et al., “A Trial of Two Strategies to Modify the Test-Ordering Behavior of Medical Residents,” N. Eng. J. Med. 303(23):1330-36, Dec. 4, 1980. Martin, S. G., Shwartz, M., Whalen, B. ]., et al., “Impact of a Mandatory Second-Opinion Pro- gram on Medicaid Surgery Rates,” Medical Care 20(1):21-39, January 1982. Martin, S. P., Donaldson, M. D., London, C. D., et al., “Inputs Into Coronary Care During 30 Years: A Cost Effectiveness Study,” Ann. Int. Med. 81:289-93, 1974. McCall, N., and Wai, H. S., “An Analysis of the Use of Medicare Services by the Continuously En- rolled Aged,” Medical Care 21:567-85, 1983. McCall, N., and Wai, H. S., “Medicare Benefi- ciaries’ Choice of Supplemental Insurance Plans,” prepared for SRI International, Health Policy Re- search Series, Discussion Paper No. 81-2, Septem- ber 1981. McCall, N., and Wai, H. S., “Medicare Benefi- ciaries’ Choice of Supplemental Insurance Plans,” prepared for SRI International, unpublished draft, 1981. McCarthy, E. G., and Finkel, M. L., “Second Consultant Opinion for Elective Gynecologic Sur- gery,” Obstet. Gyn. 56(4):403-10, October 1980. McCarthy, E. G., and Finkel, M. L., “Second Opinion Elective Surgery Program: A Concept Whose Time Has Come?” Employee Benefits Journal 4:34-83, summer 1979. McCarthy, E. G., and Finkel, M. L., “Second Opinion Elective Surgery Programs: Outcome Status Over Time,” Medical Care 16(12):984-94, December 1978. McCarthy, E. G., Finkel, M. L., and Ruchlin, H. References © 217 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. S., “Second Opinions on Elective Surgery. The Cornell/New York Hospital Study,” Lancet 1(8234):1352-56, June 20, 1981. McMillan, A., Pine, P. L., and Newton, M., “Medicare: Use of Physicians’ Services Under the Supplementary Medical Insurance Program, 1975-1978,” Health Care Financing Program Sta- tistics, Office of Research and Demonstrations, Health Care Financing Administration, U.S. De- partment of Health and Human Services, March 1983. McNamara, E., “Hospitals Rediscover Compre- hensive Home Care,” Hospitals 56(2):60-66, Nov. 1, 1982. Medical Device and Diagnostics Report, “Health Incentive Reform Package Would Refashion Medicare Along the Competitive Health Model; Schweicker Predicts 1984 Savings of $2.1 Bil.,” Feb. 7, 1983, p. S-2. Medical World News, “Second Opinions Called Cost-Saver, But Few Voluntarily Seek Them,” 23(16):8-9, Aug. 2, 1982. Melia, E. P., Aucoin, L. M., Duhl, L. J., et al., “Competition in the Health-Care Marketplace: A Beginning in California,” N. Eng. J. Med. 308(13): 788-92, Mar. 31, 1983. Meyers, S. M., “Growth in Health Maintenance Organizations,” Health— United States, 1982, National Center for Health Statistics, U.S. De- partment of Health and Human Services, pub- lication No. (PHS) 82-1232 (Hyattsville, Md.: DHHS, 1982). Mitchell, J. B., and Cromwell, J., “Alternative Methods for Describing Physician Services Per- formed and Billed, Executive Summary,” report prepared for the Office of Research, Health Care Financing Administration, U.S. Department of Health and Human Services, October 1982. Monsma, G., “Marginal Revenue and the De- mand for Physicians’ Services,” in Empirical Studies in Health Economics, H. E. Klarman (ed.) (Baltimore, Md.: Johns Hopkins Press, 1970). Moser, R. H., Executive Vice President, Amer- ican College of Physicians, Washington, D.C, statement on “Respiratory Therapy Guidelines,” released Oct. 21, 1982. Mushkin, S. J., Paringer, L. C., and Chen, M. M., “Returns to Biomedical Research, 1900-1975: An Initial Assessment of Impacts on Health Ex- penditures,” mimeograph (Washington, D.C.: Georgetown University, Public Services Labora- tory, December 1977). Myers, L. P., and Schroeder, S. A., “Physician Use of Services for the Hospitalized Patient: A 25-337 0 - 84 - 16 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. Review, With Implications for Cost Contain- ment,” Milbank Mem. Fund Q./Health and Society 59:481-507, 1981. National Academy of Sciences, Institute of Medi- cine and National Research Council, Committee on Technology and Health Care, Medical Tech- nology and the Health Care System: A Study of the Diffusion of Equipment-Embodied Technol- ogy (Washington, D.C.: NAS, 1979). National Academy of Sciences, Institute of Medi- cine, A Consortium for Assessing Medical Tech- nology: A Planning Study Report (Washington, D.C.: NAS, November 1983). National Academy of Sciences, Institute of Medi- cine, Division of Health Manpower and Re- sources Development, A Manpower Policy for Primary Health Care (Washington, D.C.: NAS, May 1978). National Academy of Sciences, Institute of Medi- cine, Reliability of Hospital Discharge Abstracts, publication No. IOM-77-01 (Washington, D.C.: NAS, 1977). National Academy of Sciences, Institute of Medi- cine, Reliability of National Hospital Discharge Survey Data, publication No. IOM-80-82 (Wash- ington, D.C.: NAS, 1980). National Association of Freestanding Emergency Centers, “Perspectives on Health Care Costs and Freestanding Emergency Center Issues” (Dallas, Tex.: NAFEC, January 1983). National Association of Insurance Commissions, Malpractice Claims Final Compilation (Medical Malpractice Closed Claims 1975-1978) 2(2), Sep- tember 1980. National Kidney Foundation, Washington, D.C., personal communication, Feb. 9, 1984. New York Academy of Medicine, “Registry of Patients on Home Total Parenteral Nutrition (TPN),” unpublished data, 1979-81. New York Academy of Medicine, “Registry of Patients on Home Total Parenteral Nutrition (TPN),” unpublished data, 1982. Newhouse, J. P., Manning, W. G., Morris, C. N., et al., “Some Interim Results From a Con- trolled Trial of Cost Sharing in Health Insur- ance,” N. Eng. J. Med. 305:1501, Dec. 17, 1981. Newhouse, J. P., and Phelps, C. E., “New Esti- mates of Price and Income Elasticities of Medi- cal Care Services,” in The Role of Health Insur- ance in the Health Services Sector, R. N. Rosett (ed.) (New York: National Bureau of Economic Research, 1976). Newhouse, J. P., and Phelps, C. E., On Having Your Cake and Eating It Too: An Analysis of Es- 218 e Medical Technology and Costs of the Medicare Program 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. timated Effects of Insurance on Demand for Med- ical Care (Santa Monica, Calif.: The Rand Corp., April 1974). Newhouse, J. P., and Phelps, C. E., “Price and Income Elasticities for Medical Care Services,” in The Economics of Health and Medical Care, M. Perlman (ed.) (London: MacMillan & Co., 1974). Newhouse, J. P., Manning, W. G., Morris, C. N., et al., “Some Interim Results From a Con- trolled Clinical Trial of Cost Sharing in Health Insurance,” N. Eng. J. Med. 305:1501, Dec. 17, 1981. Olseth, D. R., Chairman-Elect of Health Indus- try Manufacturers Association, statement on health care costs before the Senate Committee on Labor and Human Resources, May 19, 1983. Orkand Corp., Comparative Evaluation of Costs, Quality and System Effects of Ambulatory Surgery Performed in Alternative Settings: Final Report, prepared for the Health Care Financing Administration, U.S. Department of Health, Edu- cation and Welfare, 1977. Orkand Corp., Study to Provide Updated Infor- mation on Emergicenters, prepared for the Of- fice of Research and Demonstrations, Health Care Financing Administration, U.S. Department of Health and Human Services, Feb. 22, 1983. Policy Analysis, Inc., “Evaluation of the Effects of Certificate of Need Programs,” Volume III, Analytic Reports, contract report prepared for the Bureau of Health Planning and Resources De- velopment, Health Resources Administration, Department of Health and Human Services, con- tract No. 231-77-0114, Brookline, Mass., August 1980. Palmer, R. H., and Nesson, H. R., “A Review of Methods for Ambulatory Medical Care Eval- uations,” Medical Care 20(8):758-81, August 1982. Pauly, M. V., and Langwell, K. M., Research on Competition in the Financing and Delivery of Health Services: Future Research Needs, National Center for Health Services Research, U.S. Depart- ment of Health and Human Services, DHHS pub- lication No. (PHS) 83-3328-2, 1982. Paxton, H. T., Carlova, J. E., Eisenberg, H., et al., “Making Your Practice More Malpractice Proof,” Medical Economics, Sept. 30, 1974. Peel, E., and Scharff, J., “Impact of Cost Shar- ing on Ambulatory Services Under Medicare, 1969,” Soc. Sec. Bull. 36(10):3-24, October 1973. Perspectives, “Health Planning Looks to the Future,” Dec. 5, 1983. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. Perspectives, “Hospice: Medicare's Newest Pro- vider,” Nov. 29, 1982. Perry, S., Washington, D.C., personal commu- nication, Dec. 27, 1983. Phelps, C. E., and Newhouse, J. P., Coinsurance and the Demand for Physician Services, R-964- 1-OEO/NC (Santa Monica, Calif.: The Rand Corp., October 1974). Phelps, C. E., and Newhouse, J. P., “Coinsur- ance, the Effect of Time, and the Demand for Medical Services,” Research Economics and Sta- tistics 56(3):334-42, August 1974. Phelps, C. E., and Newhouse, J. P., The Effects of Insurance on Demand for Physician Services R-976-OEO (Santa Monica, Calif.: The Rand Corp., June 1973). Plough, A. L., and Berry, R. E., “Regulating Medical Uncertainty: Policy Issues in End-Stage Renal Disease,” in Federal Health Programs: Problems and Prospects, S. H. Altman and H. M. Sapolsky (eds.) (Lexington, Mass. : Lexington Books, 1981). Potenza, R., “Community Cancer Management,” Hospitals 57(1):80-85, Jan. 1, 1983. Praise, I., and Gjerde, C., “Cost Containment Through Medical Education,” J.A.M.A. 244(1): 53-55, July 4, 1980. Primus, W. E., “Financing Medicare Through 1995,” National J., May 1, 1982. Rapaport, E., “An Overview of Issues,” Circula- tion 66(5) Part II, AHA Monograph No. 92, Technology Assessment Forum on Coronary Artery Bypass Surgery: Economic, Ethical and Social Issues, November 1982. Redisch, M. A., “Hospital Inflationary Mecha- nisms,” paper presented at the Western Econom- ics Association Meetings, Las Vegas, Nev., June 10-12, 1974. Reiss, J. B., “Thoughts on Technology Assess- ment and Paying for Services,” paper submitted to the Office of Technology Assessment, U.S. Congress, November 1981. Relman, A. S., “The New Medical-Industrial Complex,” N. Eng. J. Med. 303(17):963, Oct. 23, 1980. Relman, A. S., “The Rand Health Insurance Study: Is Cost Sharing Dangerous to Your Health?” N. Eng. J. Med. 309(23):1453, Dec. 8, 1983. Rettig, R., “End-Stage Renal Disease and the ‘Cost’ of Medical Technology,” in Medical Tech- nology: The Culprit Behind Health Care Costs? S. Altman and R. Blendon (eds.), proceedings of References 219 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. the 1977 Sun Valley Forum on National Health, U.S. Department of Health, Education, and Wel- fare, publication No. (PHS) 79-3216 (Washing- ton, D.C.: U.S. Government Printing Office, August 1979). Rhode Island Health Services Research, Inc., “The Effect of Cimetidine on Peptic Ulcer in Rhode Island,” Dec. 24, 1980. Rice, D. P., and Feldman, J. J., “Living Longer in the United States: Demographic Changes and Health Needs of the Elderly,” Milbank Mem. Fund Q./Health and Society 61(3):362-97, 1983. Rice, T. H., “Determinants of Physician Assign- ment Rates: An Analysis by Type of Service,” Health Care Finan. Rev., forthcoming, 1984. Rice, T. H., “The Impact of Changing Medicare Reimbursement Rates on Physician-Induced De- mand,” Medical Care 21(8):803, August 1983. Rice, T., and McCall, N., “Changes in Medicare Reimbursement in Colorado: Impact on Physi- cians’ Economic Behavior,” Health Care Finan. Rev. 3:67-85, June 1982. Richardson, W. C., “Developing Primary Care Networks: Lessons Learned From Safeco,” pres- entation to National Health Policy Forum, Wash- ington, D.C., Oct. 21, 1982. Ricker-Smith, K. L., “A Challenge for Public Pol- icy: The Chronically Ill Elderly and Nursing Homes,” Medical Care 20(11):1071-79, Novem- ber 1982. Roberts, S., et al., “Cost-Effective Care of End- Stage Renal Disease,” N. Eng. J. Med. 30:243, 1980. Rochester Area Hospitals’ Corp., 1980 Annual Report: Rx for Hospitals (Rochester, N.Y.: RAHC, 1981). Roenick, D., and Bartlett, L., Controlling Med- icaid Costs: Second Surgical Opinion Programs (Washington, D.C.: National Governors’ Asso- ciation, November 1982). Romeo, A. J., Wagner, J. L., and Lee, R. H., “Prospective Hospital Reimbursement and the Diffusion of New Hospital Technologies,” pre- pared for The Urban Institute, 1982. Rosenthal, G., “Anticipating the Costs and Benefits of New Technology,” in Medical Tech- nology: The Culprit Behind Health Care Costs? S. H. Altman and R. Blendon (eds.), proceedings of the 1977 Sun Valley Forum on National Health, U.S. Department of Health, Education, and Welfare, publication No. (PHS) 79-3216 (Washington, D.C.: U.S. Government Printing Office, August 1979). 284. 285. 286. 287. 288. 289. 290. 291. 292. 293. 294. 295. 296. Rosenwaike, I., Yaffe, N., and Sagi, P. C., “The Recent Decline in Mortality of the Extreme Aged: An Analysis of Statistical Data,” Am. J. Public Health 70(10):1074-80, October 1980. Rossiter, L. F., and Hogan, C. M., “Unequal Fi- nancial Incentives for Diagnostic and Preventive Health Care,” paper prepared for the 109th An- nual Meeting of the American Public Health Association, Medical Care Section, Los Angeles, Calif., November 1981. Rossiter, L. F., and Wilensky, G. R., “A Reex- amination of the Use of Physician Services: The Role of Physician-Initiated Demand,” Inquiry 20:162-72, summer 1983. Rothwell, R. S., Davis, P., Gordon, P. A., etal., “A Controlled Study of Plasma Exchange in the Treatment of Severe Rheumatoid Arthritis,” Arth. & Rheumat. 23:785-90, 1980. Ruchlin, H. S., Finkel, M. L., and McCarthy, E. G., “The Efficacy of Second-Opinion Consulta- tion Programs: A Cost-Benefit Perspective,” Medical Care 20(1):3-20, January 1982. Russell, L. B., Technology in Hospitals: Medical Advances and Their Diffusion (Washington, D.C.: The Brookings Institution, 1979). Schade, R. R., and Donaldson, H., Jr., “How Physicians Use Cimetidine—A Survey of Hospi- talized Patients and Published Cases,” Medical Intelligence 304(21):1281-84, May 21, 1981. Schaeffer, L., “Role of the HCFA in the Regula- tion of New Medical Technologies,” in Critical Issues in Medical Technology, B. J. McNeil and E. G. Cravalho (eds.) (Boston, Mass.: Auburn House, 1982). Schaeffer, M. J., and Gordon, M. R., “Clinical Engineering Standards, Obligations, and Ac- countability,” Medical Instrumentation 13(4):209- 15, July-August 1979. Schiffman, G., “Chemistry and Inmunochemis- try of the Pneumococcal Polysaccharide Vaccine With Special Reference to Cross-Reactions and Immunologic Factors,” Reviews of Infectious Dis- eases 3(Supplement):S18-30, March-April 1981. Schneeweiss, R., et al., “Diagnostic Clusters: A New Tool for Analyzing the Content of Ambu- latory Medical Care,” Medical Care 21:105-22, January 1983. Schroeder, S. A., and Showstack, J. A., “Finan- cial Incentives to Perform Medical Procedures and Laboratory Tests: Illustrative Models of Of- fice Practice,” Medical Care 16:289-98, April 1978. Schroeder, S., et al., “Use of Laboratory Tests 220 * Medical Technology and Costs of the Medicare Program 297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. and Pharmaceuticals: Variation Among Physi- cians and Effect of Cost Audit on Subsequent Use,” J.LA.M.A. 225:969, 1973. Schwartz, W. B., and Aaron, H. J., “Rationing Hospital Care: Lessons From Britain,” Special Re- port in N. Eng. J. Med. 301(1):52-56, Jan. 5, 1984. Scitovsky, A. A., and McCall, N., Changes in the Costs of Treatment of Selected Illnesses, 1951- 1964-1971, U.S. Department of Health, Educa- tion, and Welfare, publication No. (HRA) 77- 3161 (Washington, D.C.: U.S. Government Printing Office, 1976). Scitovsky, A. A., and McCall, N., “Coinsurance and the Demand for Physician Services: Four Years Later,” Soc. Sec. Bull. 40:19-27, 1977. Sheeley, D. R., and Sherman, H., “Conservation in Hospital Resource Use: Treatment of Pneumo- nias: An Investigation of Care in Four Hospitals Over the Past Decade,” Ann. Int. Med. 85:648- 52, 1976. Shils, M. E., “Registry of Patients on HPN,” in Home Parenteral Nutrition, symposium pro- ceedings, American Society for Parenteral and Enteral Nutrition, pp. 42-44, 1980. Showstack, J. A., Schroeder, S. A., and Mat- sumoto, M. F., “Changes in the Use of Medical Technologies, 1972-1977,” N. Eng. J. Med. 306(12):706-12, Mar. 25, 1982. Sigmond, R. M., “Capitation Reimbursement,” paper presented at the 1980 Conference on Health Care in the American Economy, New York, June 8-11, 1980. Silverman, H. A., Medicare Program Statistics Branch, Health Care Financing Administration, U.S. Department of Health and Human Services, personal communication, August 1983. Sisk, J. E., Office of Technology Assessment, U.S. Congress, personal communication, Sept. 15, 1983. Skipper, J. K., Jr., Smith, G., Mulligan, J. L., et al., “Physicians’ Knowledge of Cost: The Case of Diagnostic Tests,” Inquiry 13(2):194-98, June 1976. Sloan, F. A., “Effects of Health Insurance on Phy- sicians’ Fees,” paper presented at the Annual Meeting of the Southern Economic Association, Washington, D.C., Nov. 6, 1980. Sloan, F. A., “Physician Fee Inflation: Evidence From the Late 1960's,” in The Role of Health In- surance in the Health Services Sector, R. N. Rosel (ed.) (New York: National Bureau of Economic Research, 1976). Sloan, F. A., “Rate Regulation as a Strategy for Hospital Cost Control: Evidence From the Last 310. 311. 312. 313. 314. 315. 316. 317. 318. 319. 320. 321. 322. Decade,” Milbank Mem. Fund Q. 61(2), spring 1983. Sloan, F. A., “Regulation and the Rising Costs of Hospital Care,” Review of Economics & Sta- tistics 63:479-87, 1981. Sloan, F. A., and Steinwald, B., “Effects of Reg- ulation on Hospital Costs and Input Use,” Van- derbilt University for presentation at the Annual Meeting of the American Economic Association, Chicago, Ill., Aug. 29, 1978. Smith Kline & French, unpublished data (as quoted in Wagner, J. L., “Economic Evaluations of Medicines: A Review of the Literature,” pre- pared for the Pharmaceutical Manufacturers As- sociation, October 1982. Smits, H., “The PSRO in Perspective,” N. Eng. J. Med. 305(5):253-59, July 30, 1981. Smits, H., Feder, J., and Scanlon, W., “Medi- care’s Nursing Home Benefit: Variations in Inter- pretation,” N. Eng. J. Med. 307(14):855-62, Sept. 30, 1982. Soboski, W. S., “Effects of the 1969 California Relative Value Studies on Costs of Physician Services Under SMI,” Health Insurance Statistics, 69, Oftice of Research and Statistics, Social Secu- rity Administration, U.S. Department of Health, Education, and Welfare, 1975. Somers, A. R., “And Who Shall Be the Gate- keeper?” paper presented to the Duke University Medical Center, Duke-Watts Family Medicine Program, Alumni Symposium: The First Ten Years, Durham, N.C., June 16, 1983, forthcom- ing in Inquiry, winter 1984 (in press). Somers, A. R., and Somers, H. M., Health and Health Care (Germantown, Md.: Aspen System, 1977). Somers, H. M., and Somers, A. R., Medicare and the Hospitals: Issues and Prospects (Washington, D.C.: The Brookings Institution, 1967). Sorenson, A. A., Saward, E. W., Steward, D. W., “Hospital Cost Containment in Rochester: From Maxicap to the Hospital Experimental Pay- ments Program,” Inquiry 19:327-35, winter 1982. Speltz, D. E., “Long-Term Care: Hospitals Must Change Missions and Marketing to Better Serve the Elderly,” Hospitals 56:56-57, 59-60, July 1, 1982. Strauss, A., Chronic Illness and the Quality of Life (St. Louis, Mo.: The CV Mosby Co., 1975). Sunshine, J. H., “Medicare: Past, Present and Future,” in Rethinking Medicare to Meet Future Needs, The Government Research Corp., Issues Book of the Seventh Annual Leadership Confer- ence on Health Policy, June 17-18, 1982. References ¢ 221 323. 324. 325. 326. 327. 328. 329. 330. 331. 332. 333. 334. Tancredi, L. R., and Barondess, J. A., “The Prob- lem of Defensive Medicine,” Science 200:879-82, May 26, 1978. Tanon, C. P., and Rogers, E., Diffusion of Med- ical Technology, G. Gordon and L. Fisher (eds.) (Cambridge, Mass.: Ballinger Publishing Co., 1975). Trauner, J. B., Luft, H. S., and Robinson, J. O., Entrepreneurial Trends in Health Care Delivery: The Development of Retail Dentistry and Free- standing Ambulatory Services (Washington, D.C.: U.S. Government Printing Office, Dec. 7, 1982). Tresnowski, B. R., President, Blue Cross/Blue Shield Association, “Remarks on the Medical Ne- cessity Program,” Blue Cross/Blue Shield press briefing, Washington, D.C., Oct. 21, 1982. U.S. Congress, Congressional Budget Office, Changing the Structure of Medicare Benefits: Issues and Options (Washington, D.C.: U.S. Government Printing Office, March 1983). U.S. Congress, Congressional Budget Office, The Effects of PSROs on Health Care Costs: Current Findings and Future Evaluations (Washington, D.C.: U.S. Government Printing Office, 1979). U.S. Congress, Congressional Budget Office, The Impact of PSROs on Health Care Costs: Update of CBO’s 1979 Evaluation (Washington, D.C.: U.S. Government Printing Office, January 1981). U.S. Congress, Congressional Budget Office, Re- ducing the Deficit: Spending and Revenue Op- tions, a report to the Senate and House Commit- tees on the Budget—Part III (Washington, D.C.: U.S. Government Printing Office, February 1983). U.S. Congress, Congressional Budget Office, “Re- sponse From Alice M. Rivlin to Howard M. Metzenbaum on the Impact of a Medicare Pneu- mococcal Vaccination Benefit,” Sept. 5, 1980. U.S. Congress, General Accounting Office, Hu- man Resources Division, “Waiver of Medicaid Freedom-of-Choice Reimbursement: Potential Savings and Practical Problems,” memorandum to Richard S. Schweiker, Secretary of Health and Human Services, No. GAO/HRD-82-90, July 20, 1982. U.S. Congress, General Accounting Office, The Elderly Should Benefit From Expanded Home Health Care But Increasing These Services Will Not Insure Cost Reductions, publication No. IPE- 83-1 (Washington, D.C.: U.S. Government Print- ing Office, Dec. 7, 1982). U.S. Congress, General Accounting Office, Prob- lems With Evaluating the Cost Effectiveness of 335. 336. 337. 338. 339. 340. 341. 342. 343. 344. 345. 346. Professional Standards Review Organizations (Washington, D.C.: U.S. Government Printing Office, July 19, 1979). U.S. Congress, House of Representatives, Com- mittee on Interstate and Foreign Commerce, Sub- committee on Health and the Environment, A Discursive Dictionary of Health Care (Washing- ton, D.C.: U.S. Government Printing Office, 1976). U.S. Congress, Library of Congress, Congres- sional Research Service, “Health Insurance: The Medicare Voucher Proposals,” by G. R. Markus, issue brief No. IB81179, updated Feb. 17, 1983. U.S. Congress, Library of Congress, Congres- sional Research Service, “Health Insurance: The Pro-Competition Proposals,” by J. P. Lundy and G. R. Markus, issue brief No. IB81046, updated Feb. 9, 1983. U.S. Congress, Library of Congress, Congres- sional Research Service, “How Medicare Pays Doctors,” by J. O'Sullivan and G. Markus, issue brief No. 1B82202, revised Nov. 10, 1982. U.S. Congress, Library of Congress, Congres- sional Research Service, “Medicare,” by G. R. Markus, issue brief No. IB82044, updated Jan- uary 1983. U.S. Congress, Library of Congress, Congres- sional Research Service, “Medicare,” by J. O’'Sul- livan and G. Markus, issue brief No. 1B82044, updated August 1983. U.S. Congress, Office of Technology Assessment, Assessing the Efficacy and Safety of Medical Technologies, OTA-H-75 (Washington, D.C.: U.S. Government Printing Office, September 1978). U.S. Congess, Office of Technology Assessment, Development of Medical Technology: Opportu- nities for Assessment, OTA-H-34 (Washington, D.C.: U.S. Government Printing Office, August 1976. U.S. Congress, Office of Technology Assessment, Diagnosis Related Groups (DRGs) and the Medi- care Program: Implications for Medical Technol- ogy, OTA-TM-H-17 (Washington, D.C.: U.S. Government Printing Office, July 1983). U.S. Congress, Office of Technology Assessment, “Effectiveness and Costs of Continuous Ambu- latory Peritoneal Dialysis,” unpublished draft prepared by W. B. Stason and B. A. Barnes, No- vember 1983. U.S. Congress, Office of Technology Assessment, “Federal Policies and the Medical Devices Indus- try,” status report, July 1983. U.S. Congress, Office of Technology Assessment, 222 e Medical Technology and Costs of the Medicare Program 347. 348. 349. 350. 351. 352. 353. 354. 355. 356. 357. Forecasts of Physician Supply and Requirements, OTA-H-113 (Washington, D.C.: U.S. Govern- ment Printing Office, 1980). U.S. Congress, Office of Technology Assessment, Health Technology Case Study #2: The Feasibility of Economic Evaluation of Diagnostic Proce- dures: The Case of CT Scanning, OTA-HCS-2 (Washington, D.C.: U.S. Government Printing Office, April 1981). U.S. Congress, Office of Technology Assessment, Health Technology Case Study #22: The Effec- tiveness and Costs of Alcoholism Treatment, OTA-HCS-22 (Washington, D.C.: U.S. Govern- ment Printing Office, June 1983). U.S. Congress, Office of Technology Assessment, Health Technology Case Study #23: The Safety, Efficacy, and Cost Effectiveness of Therapeutic Apheresis, OTA-HCS-23 (Washington, D.C.: U.S. Government Printing Office, July 1983). U.S. Congress, Office of Technology Assessment, Health Technology Case Study #24: Variations in Hospital Length of Stay: Their Relationship to Health Outcomes, OTA-HCS-24 (Washington, D.C.: U.S. Government Printing Office, August 1983). U.S. Congress, Office of Technology Assessment, Health Technology Case Study #26: Assistive Devices for Severe Speech Impairments, OTA- HCS-26 (Washington, D.C.: U.S. Government Printing Office, December 1983). U.S. Congress, Office of Technology Assessment, Impact of Randomized Clinical Trials on Health Policy and Medical Practice, OTA-BP-H-22 (Washington, D. C.: U.S. Government Printing Office, August 1983). U.S. Congress, Office of Technology Assessment, Implications of Cost-Effectiveness Analysis of Medical Technology, OTA-H-126 (Washington, D.C.: U.S. Government Printing Office, August 1980). U.S. Congress, Office of Technology Assessment, “Intensive Care Units (ICUs): Costs, Outcomes, and Decisionmaking,” unpublished draft pre- pared by R. A. Berenson, August 1983. U.S. Congress, Office of Technology Assessment, Medical Technology Under Proposals to Increase Competition in Health Care, OT A-H-190 (Wash- ington, D.C.: U.S. Government Printing Office, October 1982). U.S. Congress, Office of Technology Assessment, Policy Implications of the Computed Tomogra- phy (CT) Scanner: An Update, OTA-BP-H-8 (Washington, D.C.: U.S. Government Printing Office, January 1981). U.S. Congress, Office of Technology Assessment, 358. 359. 360. 361. 362. 363. 364. 365. 366. 367. 368. 369 “Report of a Study on Decisionmaking by Medi- care Contractors for Coverage of Medical Tech- nologies,” prepared by L. K. Demlo, G. T. Ham- mons, J. M. Kuder, et al., Oct. 28, 1983. U.S. Congress, Office of Technology Assessment, A Review of Selected Federal Vaccine and Immu- nization Policies Based on Case Studies of Pneu- mococcal Vaccine, OTA-H-96 (Washington, D.C.: U.S. Government Printing Office, Septem- ber 1979). U.S. Congress, Office of Technology Assessment, Strategies for Medical Technology Assessment, OTA-H-181 (Washington, D.C.: U.S. Govern- ment Printing Office, September 1982). U.S. Congress, Office of Technology Assessment, Technology and Aging in America, forthcoming, summer 1984. U.S. Congress, Office of Technology Assessment, Technology and Handicapped People, OT A-H- 179 (Washington, D.C.: U.S. Government Print- ing Office, May 1982). U.S. Congress, U.S. Senate, Committee on Fi- nance, Report No. 92-1230, Sept. 9, 1983. U.S. Congress, U.S. Senate, Special Committee on Aging, “The Hospice Alternative,” hearing held in Pittsburgh, Pa., 97th Cong., 2d sess., May 24, 1982. U.S. Department of Commerce, Bureau of Labor Statistics, personal communication, Sept. 1, 1983. U.S. Department of Health, Education, and Wel- fare, Health Care Financing Administration, 1978 PSRO Program Evaluation, DHEW publication No. (HCFA) 03000 (Baltimore, Md.: DHEW, Jan- uary 1979). U.S. Department of Health, Education, and Wel- fare, Health Care Financing Administration, memorandum of the Administrator from the Di- rector of the Office of Policy Analysis on “Cov- erage Under Medicare,” prepared by M. Chassin, Aug. 24, 1979. U.S. Department of Health, Education, and Wel- fare, National Center for Health Statistics, “De- tailed Diagnosis and Surgical Procedures for Pa- tients Discharged From Short-Stay Hospitals: United States, 1977,” DHEW publication No. (PHS) 79-1279 (Hyattsville, Md.: DHEW, Sep- tember 1979). U.S. Department of Health, Education, and Wel- fare, National Center for Health Statistics, “The Change in Mortality Trends in the United States,” Vital and Health Statistics, Series 3, No. 1, DHEW publication No. (PHS) 1000 (Hyattsville, Md.: DHEW, 1964). . U.S. Department of Health, Education, and Wel- References 223 370. 371. 372. 373. 374. 375. 376. 377. 378. 379. 380. 381. fare, National Center for Health Statistics, “Uti- lization of Short-Stay Hospitals, Annual Sum- mary for the United States, 1975,” Vital and Health Statistics, Series 13, No. 31, DHEW pub- lication No. (HRA) 77-1782 (Washington, D.C.: DHEW, April 1977). U.S. Department of Health, Education, and Wel- fare, Report of the Secretary’s Commission on Medical Malpractice (Washington, D.C.: DHEW, Jan. 16, 1973). U.S. Department of Health and Human Services, Office of the Inspector General, “Long-Term Care: Service Delivery Assessment, Report to the Secretary,” 1981. U.S. Department of Health and Human Services, Social Security Advisory Council to the Secre- tary, meeting of June 20, 1983, as reported in The Washington Post, June 21, 1983. U.S. Department of Health and Human Services, Health Care Financing Administration, 1966-1979 Data, Persons Enrolled for Medicare, 1979, HCFA publication No. 03079 (Baltimore, Md.: HCFA, January 1981). U.S. Department of Health and Human Services, Health Care Financing Administration, 1979 PSRO Program Evaluation (Washington, D.C.: DHHS, 1980). U.S. Department of Health and Human Services, Health Care Financing Administration, unpub- lished data, May 1982. U.S. Department of Health and Human Services, Health Care Financing Administration, “Annual Report to the Congress From the Secretary of Health and Human Services on the Professional Standards Review Organizations Program, Fiscal Year 1980,” memo, February 1981. U.S. Department of Health and Human Services, Health Care Financing Administration, Division of Public Information staff, personal communi- cation, Jan. 23, 1984. U.S. Department of Health and Human Services, Health Care Financing Administration, End-Stage Renal Disease Report to Congress, 1983, in press. U.S. Department of Health and Human Services, Health Care Financing Administration, End Stage Renal Disease Second Annual Report to Con- gress, Fiscal Year 1980, 1980. U.S. Department of Health and Human Services, Health Care Financing Administration, Health Care Financing Trends, DHHS publication No. 03091 (Baltimore, Md.: DHHS, winter 1981). U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Coverage Policy staff, personal communication, December 1982. 382. 383. 384. 385. 386. 387. 388. 389. 390. 391. 392. 393. U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Coverage Policy, Medicare Carriers Manual, Part 3, Claims Process, Coverage Index Appendix, sec. 35-50, Sept. 15, 1981. U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Coverage Policy, Medicare Carriers Manual, Part 3, Claims Process Transmittal, No. 984, sec. 35- 70, July 1983. U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Coverage Policy, “Operating Guidelines for De- termining Medical Necessity of Medicare Items and Services: Discussion Draft,” mimeo, Feb. 2, 1981. U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Program Administration, “Health Care Financ- ing Administration Intermediary and Carrier Directory,” Medicare Manual, October 1982. U.S. Department of Health and Human Services, Health Care Financing Administration, “Operat- ing Guidelines for the HCFA Physician Panel,” mimeo, 1980. U.S. Department of Health and Human Services, Health Care Financing Administration, “PSRO Evaluation,” Research and Demonstrations in Health Care Financing: 1980-1981, HCFA pub- lication No. 03144 (Rockville, Md.: HCFA, 1982). U.S. Department of Health and Human Services, Health Care Financing Administration, Region J, Boston Regional Intermediary Bulletin, No. 197, Subject: Coverage Issues, mimeo, Jan. 17, 1978. U.S. Department of Health and Human Services, Health Care Financing Administration, Report to Congress: Hospital Prospective Payment for Medicare (Rockville, Md.: HCFA, December 1982). U.S. Department of Health and Human Services, Health Care Financing Administration, Statisti- cal Information Services staff, personal commu- nication, Sept. 1, 1983. U.S. Department of Health and Human Services, Health Care Financing Administration, The Medicare and Medicaid Databook, 1981, HCFA publication No. 03128 (Baltimore, Md.: HCFA, April 1982). U.S. Department of Health and Human Services, Health Care Financing Administration, The Medicare and Medicaid Databook, 1983, HCFA publication No. 03156 (Baltimore, Md.: HCFA, December 1983). U.S. Department of Health and Human Services, 224 e Medical Technology and Costs of the Medicare Program 394. 395. 396. 397. 398. 399. 400. Health Care Financing Administration, “Your Medicare Handbook, January 1983 Edition,” Social Security Administration publication No. 05-10050 (Baltimore, Md.: SSA, January 1983). U.S. Department of Health and Human Services, National Center for Health Services Research, Eyeglasses and Contact Lenses: Purchases, Ex- penditures, and Sources of Payment, NCHSR National Health Care Expenditure Study, Data Preview 11, prepared by D. C. Walden, DHHS publication No. (PHS) 82-3322 (Hyattsville, Md.: DHHS, October 1982). U.S. Department of Health and Human Services, National Center for Health Services Research, In- patient Hospital Services: Use, Expenditures, and Sources of Payment, NCHSR National Health Care Expenditure Study, Data Preview 15, pre- pared by A. K. Taylor, DHHS publication No. (PHS) 83-3360 (Hyattsville, Md.: DHHS, Octo- ber 1982). U.S. Department of Health and Human Services, National Center for Health Services Research, Out-of-Pocket Expenditures for Personal Health Services, NCHSR National Health Care Expend- itures Study, Data Preview 13, prepared by L. F. Rossiter and G. R. Wilensky, DHHS publication No. (PHS) 82-3332 (Hyattsville, Md.: DHHS, October 1982). U.S. Department of Health and Human Services, National Center for Health Services Research, Medical Equipment and Supplies: Purchases and Rentals, Expenditures, and Sources of Payment, NCHSR National Health Care Expenditure Study, Data Preview 10, prepared by M. M. Hagan, DHHS publication No. (PHS) 82-3321 (Hyattsville, Md.: DHHS, October 1982). U.S. Department of Health and Human Services, National Center for Health Services Research, Prescribed Medicines: Use, Expenditures, and Sources of Payment, NCHSR National Health Care Expenditure Study, Data Preview 9, pre- pared by J. A. Kasper, DHHS publication No. (PHS) 82-3320 (Hyattsville, Md.: DHHS, Octo- ber 1982). U.S. Department of Health and Human Services, National Center for Health Services Research, Of- fice of Health Technology Assessment, “OHTA Guidelines for the Evaluation of the Safety and Clinical Effectiveness of Medical Technologies,” mimeo, September 1982. U.S. Department of Health and Human Services, National Center for Health Statistics, “1982 Sum- mary: National Hospital Discharge Survey,” No. 95, DHHS Publication No. (PHS) 84-1250 (Hy- attsville, Md.: DHHS, Dec. 27, 1983). 401 402. 403. 404. 405. 406. 407. 408. 409. 410. 411. . U.S. Department of Health and Human Services, National Center for Health Statistics, “Changes in Mortality Among the Elderly: United States, 1940-1978,” Vital and Health Statistics, Series 3, No. 22, DHHS publication No. (PHS) 82-146 (Hyattsville, Md.: DHHS, 1982). U.S. Department of Health and Human Services, National Center for Health Statistics, “Current Estimates From the National Health Interview Survey: U.S. 1981,” Vital and Health Statistics, Series 10, No. 141, DHHS publication No. (PHS) 82-1569 (Washington, D.C.: DHHS, December 1982). U.S. Department of Health and Human Services, National Center for Health Statistics, “Detailed Diagnosis and Surgical Procedures for Patients Discharged From Short-Stay Hospitals: United States, 1978,” DHHS publication No. 80-1274 (Hyattsville, Md.: DHHS, September 1980). U.S. Department of Health and Human Services, National Center for Health Statistics, “Detailed Diagnosis and Surgical Procedures for Patients Discharged From Short-Stay Hospitals: United States, 1979,” DHHS publication No. (PHS) 82- 1274-1 (Hyattsville, Md.: DHHS, January 1982). U.S. Department of Health and Human Services, National Center for Health Statistics, Health— United States, 1980, DHHS Publication No. (PHS) 81-1232 (Hyattsville, Md: DHHS, Decem- ber 1980). U.S. Department of Health and Human Services, National Center for Health Statistics, Health— United States, 1981, DHHS publication No. (PHS) 82-1232 (Hyattsville, Md.: DHHS, Decem- ber 1981). U.S. Department of Health and Human Services, National Center for Health Statistics, Health— United States, 1982, DHHS publication No. (PHS) 83-1232 (Hyattsville, Md.: DHHS, Decem- ber 1982). U.S. Department of Health and Human Services, National Center for Health Statistics, Health— United States, 1983, DHHS Publication No. (PHS) 84-1232 (Hyattsville, Md: DHHS, Decem- ber 1983). U.S. Department of Health and Human Services, National Institutes of Health Consensus Devel- opment Conference Statement, “Critical Care Medicine,” Mar. 7-9, 1983. U.S. Department of Health and Human Services, Public Health Service, “Memorandum of Under- standing Regarding Medicare Coverage Determi- nation,” mimeo, Dec. 16, 1980. U.S. Department of Health and Human Services, Public Health Service, PSRO: An Initial Evalua- References e 225 412. 413. 414. 415. 416. 417. 418. 419. 420. 421. 422. tion of the Professional Standards Review Orga- nization, Vol I: Executive Summary, February 1978. U.S. Government, Executive Office of the Presi- dent, Office of Management and Budget, Budget of the United States Government, Fiscal Year 1984, 1983. United Mine Workers of America, Health and Retirement Funds, “Effects of Cost Sharing on Beneficiaries of the United Mine Workers’ Health Plan,” prepared for U.S. Department of Health and Human Services, National Center for Health Services Research, contract No. 233-79-2034, De- cember 1982. : Urban Systems Research and Engineering, and Policy Analysis, Inc., Certificate of Need Pro- grams: A Review, Analysis, and Annotated Bib- liography of the Research Literature, prepared under contract No. HRA-231-77-0114 for the De- partment of Health, Education, and Welfare, publication No. (HRA) 79-14006 (Washington, D.C.: DHEW, November 1978). Vautrain, R. L., and Griner, P. F., “Physician's Orders, Use of Nursing Resources, and Subse- quent Clinical Events,” J. Med. Educ. 53:125-28, 1978. Vladeck, B. C., “Paying Hospitals,” New Ap- proaches to the Medicaid Crisis, R. J. Blendon and T. W. Moloney (eds.) (New York: F & S Press, 1982). Wagner, J. L., “Economic Evaluations of Medicines: A Review of the Literature,” prepared for the Pharmaceutical Manufacturers Associa- tion, mimeo, October 1982. Wagner, J. L., Krieger, M. ]., Lee, R. H., et al., “A Study of the Impact of Reimbursement Strat- egies on the Diffusion of Medical Technologies, Vol. I: Summary and Findings,” prepared for The Urban Institute, 1982. Waldman, S., “The Effect of Changing Technol- ogy on Hospital Costs,” Research and Statistics Note, U.S. Department of Health, Education, and Welfare, publication No. (OS) 76-502 (Washing- ton, D.C.: DHEW, Feb. 28, 1972). Warner, K. E., and Luce, B. R., Cost-Benefit and Cost-Effectiveness Analysis in Health Care: Prin- ciples, Practice, and Potential (Ann Arbor, Mich.: Health Administration Press, 1982). Wateska, L. P., Sattler, L. L., and Steiger, E., “Cost of Home Parenteral Nutrition Program,” J.A.M.A. 244(20):2302-04, Nov. 21, 1980. Weinstein, M. C., and Stason, W. B., “Cost- Effectiveness of Coronary Artery Bypass Sur- gery,” Circulation 66(5) Part II, AHA Monograph 423. 424. 425. 426. 427. 428. 429. 430. 431. 432. 433. 434. 435. No. 92, Technology Assessment Forum on Cor- onary Artery Bypass Surgery: Economic, Ethical and Social Issues, November 1982. Weinstein, M. D., Fineberg, H. V., Elstein, A. S., et al., Clinical Decision Analysis (Philadelphia, Pa.: W. B. Saunders Co., 1980). Wennberg, J. E., “Should the Cost of Insurance Reflect the Cost of Use in Local Hospital Mar- kets?” N. Eng. J. Med. 307(22):1374-81, Nov. 25, 1982. Westphal, M., Frazier, E., and Miller, M. C., “Changing in Average Length of Stay and Aver- age Charges Generated Following Institution of PSRO Review,” Health Serv. Res. 14(4):253-65, winter 1979. Wilensky, G. R., and Rossiter, L. F., “The Mag- nitude and Determinants of Physician Initiated Visits in the United States,” U.S. Department of Health and Human Services, National Center for Health Services Research, prepared for presen- tation at the World Congress on Health Econom- ics, Leiden University, The Netherlands, Sept. 8- 11, 1980. Williamson, J. W., Alexander, M., and Miller, G. E., “Continuing Education and Patient Care Research: Physician Response to Screening Test Results,” J, A.M.A. 201:118-22, 1967. Wolff, J., and Dunnihoo, D. R., “A Freestanding Ambulatory Surgical Unit: A Success or Failure?” Am. J. Obstet. Gynecol. 143(3):270-76, June 1, 1982. Wolinsky, F. D., “The Performance of Health Maintenance Organizations: An Analytic Re- view,” Milbank Mem. Fund Q. 58(4):537-87, fall 1980. Worthington, N. L., “Expenditures for Hospital Care and Physicians’ Services: Factors Affecting Annual Changes,” Soc. Security Bull. 38:3-15, 1975. Worthington, N., and Piro, P., “The Effects of Hospitals’ Rate Setting Programs on Volumes of Hospital Services,” Health Care Finan. Rev. 4(2):47, December 1982. Wylie, C., “The Complex Wane of Peptic Ulcer, II. Trends in Duodenal and Gastric Ulcer Admis- sions to 790 Hospitals—1974-79,” Journal of Gas- troenterology 3:333-39, 1981. Wyllie, J. H., et al., “Effect of Cimetidine on Sur- gery for Duodenal Ulcers,” Lancet 1(8252): 307- 8, June 13, 1981. Young, D. A., U.S. Department of Health and Human Services, Health Care Financing Admin- istration, personal communication, January 1983. Young, D. A., “List and Description of Govern- 226 » Medical Technology and Costs of the Medicare Program 436. ment Payers,” paper submitted to the Commit- tee on Evaluating Medical Technologies in Clin- ical Use, Institute of Medicine, National Academy of Sciences, unpublished, 1984. Young, J. M., Blue Cross/Blue Shield, personal communication, 1984. 437. Zook, C.]., and Moore, F. D., “High-Cost Users of Medical Care,” Special Article in N. Eng. J. Med. 302(18):996, May 1, 1980. Index Index ambulatory surgery centers, 137 American Hospital Association, 107 apheresis, therapeutic, 11, 56 approaches to changing incentives for medical technology adoption and use, 16 Association of American Medical Colleges, 119 beneficiary cost-sharing, 9, 36, 37, 108, 129 Blue Cross, 30, 74, 94, 95, 101, 122 California, 108, 116 capital payment methods, 102 history under Medicare, 102 under State protective payment programs, 103 under DRG payment, 104 certificate of need (CON), 103, 105, 107, 162 Cleveland Clinic, 59 Colorado, 116 Colorado Blue Cross, 101 competitive behavior, 136 computed tomography (CT) scanners, 32, 98 Congress: House Committee on Energy and Commerce, 3 Subcommittee on Health and Environment, 3 Senate Committee on Finance, 3 Subcommittee on Health, 3 Congressional Budget Office (CBO), 26, 57, 120, 128, 129 Consumer Price Index, 10, 48, 122, 123 coronary artery bypass surgery, 11, 53 cost-benefit analysis (CBA), 12, 85 cost consciousness, physicians’, 118 cost containment technologies, 11 cost-effectiveness analysis (CEA), 12, 85, 153 cost saving technology, 86 coverage, Medicare, 74 process, 76 policy under DRG, 82 Current Procedural Terminology (CPT), 124 defensive medicine, 117, 118 Department of Health and Human Services, 14, 33, 104, 127, 156 Diagnosis Related Groups (DRGs), 3, 5, 12, 13, 14, 30, 32, 33, 39, 67, 82-83, 89, 93, 94, 98, 99, 100, 104, 105, 121, 152, 157 dialysis, kidney, 11, 59 Economic Stabilization Program (ESP), 115, 116 emergency care centers, 139 end-stage renal disease (ESRD), 6, 7, 26, 59, 60 payment for treatment, 34 Food and Drug Administration (FDA), 5, 23, 55, 67, 77, 84 General Accounting Office (GAO), 120 Great Britain, 140 Health Care Financing Administration (HCFA), 11, 12, 56, 59, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 120, 124, 127, 138, 152 Boston Regional Office, 78 Common Procedure Coding System, 124 Office of Coverage Policy, 78 Health Industry Manufacturers Association (HIMA), 80 health maintenance organizations (HMOs), 139, 140 home health care, 137 Home Health Care of America, 58 hospices, 138 Hospital Insurance program, 8, 27, 37 hospital outpatient departments, 138 impact of medical technology on Medicare costs, 9, 33, 51..82 Indiana, 95, 98 intensive care units (ICUs), 11, 44, 45, 57, 58 “intensity of care” approach, 47-50 interactions between medical technology and Medicare, 104 John A. Hartford Foundation, 101 legislation: End-Stage Renal Disease Program Amendments of 1978, 36 Omnibus Budget Reconciliation Act of 1981, 36, 37, 95, 108, 128 Social Security Act, 5, 11, 25, 27, 29, 34, 57, 73, 88 Social Security Amendments of 1972, 26, 31, 34, 60, 95 Social Security Amendments of 1983, 4, 13, 25, 30, 32, 37, 93, 120, 156 Tax Equity and Fiscal Responsibility Act (TEFRA), 13, 32, 33, 93, 94, 120, 156 Maryland, 95, 98, 103 Massachusetts, 95, 101, 103 Medicaid, 93-95, 108, 121 medical technology: alternative approaches to changing incentives, 16 development and diffusion of, 23-25 impact on Medicare costs, 4, 9, 43-61 conclusions from aggregate studies, 52 contribution to health care costs, 51 intensity of care, 47 intensive care units (ICUs), 44, 45, 52, 57, 58 Medicare beneficiaries, 43 selected technologies, 53 cimetidine, 54 coronary artery bypass graft (CABG) surgery, 53, 54 hemodialysis for chronic renal failure, 59 pneumococcal vaccine, 56 therapeutic apheresis, 56 total parenteral nutrition (TPN), 58, 59 Medicare policies affecting, 7 229 230 » Medical Technology and Costs of the Medicare Program Medicare hospital payment, 13 rational and appropriate adoption and use, 5 Medicare: beneficiaries, 3 beneficiary cost-sharing, 9, 108, 129 “Medigap,” 9 costs, 3, 4, 11, 45, 60 increase, 10 saving, 86 coverage, 74-83 coverage decisions, technologies for, 84-88 DRG hospital payment system, 82 expenditures, 3 interactions with health care system, 5 payment policies, 9, 27 physician payment, 15 policies, 3, 4 policies affecting medical technology, 23-39 benefits, 27, 28 development and diffusion of medical technologies, 23 discussion, 38 eligibility, 26 payment, 30 beneficiary cost-sharing, 36 hospital services, 30 physicians’ services, 32 reimbursement policy, 5 social and political climate, 5 spillover effects, 5 Medicare Economic Index, 34, 123 “Medigap,” 9, 165 Minnesota, 103 National Center for Health Care Technology, 56 National Institutes of Health (NIH), 5, 23, 57, 67 National Medical Care Expenditures Survey, 45, 117 New Jersey, 95, 103 New York, 95, 98 New York Academy of Medicine, 59 North Dakota, 101 nursing homes, 138 Office of Health Technology Assessment (OHTA), 79, 84, 85, 152 parenteral nutrition, total, 11, 58 payment: hospital, 13, 30, 94, 95, 97, 98, 100-102, 156 physician, 15, 32, 34, 122, 164 Pennsylvania, 103 per capita prospective payment, 100 pneumococcal vaccine, 11,.56 policy options, 17, 151 preferred provider organizations (PROs), 135, 141 primary care gatekeepers, 140 Professional Standards Review Organizations (PSROs), 119, 120, 168 Prospective Payment Assessment Commission (ProPAC), 14, 33, 83 Public Health Service (PHS), 59, 84-86 Rand National Health Insurance Study, 36, 56, 109 Rochester Area Hospitals’ Corp., 101, 102 SAFECO Insurance Co., 140 second surgical opinion, 121 Supplementary Medical Insurance program, 8, 27, 37, 60 U.S. death rate, decline in, 29 Washington, 103 787 WNIT C(0291282kLY