Appendix B
History and Current Status of Medicare Graduate Medical Education Funding

FINANCING GRADUATE MEDICAL EDUCATION

Graduate medical education (GME) is financed by a variety of revenue sources that flow through a variety of related institutions. Academic health centers may consist of a medical school, an affiliated hospital and other patient care facilities, and a faculty practice plan. The medical school may receive funds from research grants, aid from state and local governments, and endowments and gifts; the hospital and other patient care facilities may also receive funds from these sources, as well as payments for patient care. Support for the expenses related to GME activities may, in turn, flow among these institutions in a number of ways (Reuter, 1996a).

Most of the revenue of teaching hospitals comes from payments for patient care. Private payers traditionally have paid a premium above the cost of treating the patients they cover. These higher payments are used to offset the costs of care provided to patients who cannot pay—which tend to be a greater share of total costs at teaching hospitals—as well as to support the expenses related to the training of residents. Medicare and Medicaid originally based their payments on the costs at each institution—which were higher for teaching hospitals—and later provided a system of direct and indirect subsidies for medical education and related costs; for Medicaid, the structure of these subsidies varies widely from state to state (Boex and Garg, 1993).

The environment in which teaching hospitals function has changed dramatically over the past few years. Surpluses from private payers, after growing steadily as a percentage of costs through the late 1980s, began to decline as managed care plans and other insurers became less able, and therefore less willing, to pass along steep cost increases through their premiums. Moreover, although some major teaching institutions offer higher quality services that are unavailable elsewhere, and thus may remain attractive to managed care plans, the greater cost of services at many of these facilities may make it increasingly difficult for them to compete for patients (Reuter, 1996a).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund Appendix B History and Current Status of Medicare Graduate Medical Education Funding FINANCING GRADUATE MEDICAL EDUCATION Graduate medical education (GME) is financed by a variety of revenue sources that flow through a variety of related institutions. Academic health centers may consist of a medical school, an affiliated hospital and other patient care facilities, and a faculty practice plan. The medical school may receive funds from research grants, aid from state and local governments, and endowments and gifts; the hospital and other patient care facilities may also receive funds from these sources, as well as payments for patient care. Support for the expenses related to GME activities may, in turn, flow among these institutions in a number of ways (Reuter, 1996a). Most of the revenue of teaching hospitals comes from payments for patient care. Private payers traditionally have paid a premium above the cost of treating the patients they cover. These higher payments are used to offset the costs of care provided to patients who cannot pay—which tend to be a greater share of total costs at teaching hospitals—as well as to support the expenses related to the training of residents. Medicare and Medicaid originally based their payments on the costs at each institution—which were higher for teaching hospitals—and later provided a system of direct and indirect subsidies for medical education and related costs; for Medicaid, the structure of these subsidies varies widely from state to state (Boex and Garg, 1993). The environment in which teaching hospitals function has changed dramatically over the past few years. Surpluses from private payers, after growing steadily as a percentage of costs through the late 1980s, began to decline as managed care plans and other insurers became less able, and therefore less willing, to pass along steep cost increases through their premiums. Moreover, although some major teaching institutions offer higher quality services that are unavailable elsewhere, and thus may remain attractive to managed care plans, the greater cost of services at many of these facilities may make it increasingly difficult for them to compete for patients (Reuter, 1996a).

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund Medicaid programs have rapidly adopted managed care approaches to control their acute care costs; managed care plans rarely have explicit mechanisms to cover the higher costs of teaching facilities, and the efforts of individual states to maintain these payment differentials are subject to rising pressure on their budgets and the willingness to finance the payments through higher taxes. Medicare is the only payer with an explicit mechanism nationwide to support GME and the special services and patient care missions provided by teaching hospitals. Because of this, Congress focused on the methods used by Medicare in developing their 1995 proposal for the creation of a Graduate Medical Education and Teaching Hospital Fund. This proposal was included in the Balanced Budget Act of 1995 that was passed by Congress and vetoed by the President. This section also focuses on the methods used by Medicare to provide support for GME and teaching hospitals over the years. It contains an overview of the policies that determine Medicare's payments to hospitals with GME programs and the implications of these policies for the distribution of Medicare payments and the financial status of teaching hospitals. Medicare Teaching Payments The Medicare program provides two types of extra payments to hospitals with graduate medical education programs. First, teaching hospitals receive an adjustment to their Medicare Prospective Payment System (PPS) payments to reflect the added patient care costs associated with operating an intern and resident training program. The IME adjustment accounted for about 6.2% of total PPS operating payments in FY 1996, or about $4.3 billion. A separate IME adjustment is applied to PPS capital payments. Second, Medicare also pays teaching hospitals an amount, separate from PPS payments, to offset the direct costs of GME programs. These payments cover resident salaries and benefits, the salaries of supervising physicians, the cost of office space, and other overhead. Medicare direct GME payments for resident training totaled about $2.2 billion in FY 1996; in addition, $0.3 billion was paid to hospitals with nursing and allied health training programs. In addition to Medicare's payments to teaching hospitals, physicians who directly supervise interns and residents are allowed to bill under Part B of Medicare and be paid for the services they are supervising. Although it is not part of Medicare's explicit support for GME and teaching hospitals, it can be an important part of the revenues that are used by these hospitals to support their training programs. Payment Policies Prior to Prospective Payment At its inception in 1965, the Medicare program reimbursed each hospital for the reasonable operating and capital costs attributed to the treatment of Medicare

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund patients. Teaching hospitals, because of their more expensive mixes of patients and staff, practice style, and scope of services, tended to have higher costs than other hospitals and therefore received higher payments for patient care. In addition, Medicare paid the program's share of the direct operating cost of each hospital's teaching program. This payment was intended to recognize the broader social benefit of GME: Educational activities enhance the quality of care in an institution, and it is intended, until the community undertakes to bear such education costs in some other way, that a part of the net cost of such activities (including stipends of trainees, as well as compensation of teachers and other costs) should be borne to an appropriate extent by the hospital insurance program. (House Report 213, 89th Congress, 1965, as cited by Fishman, 1996) However, Medicare only covers its share of the graduate medical program costs that are allocated to the hospital; its payment does not cover the clinical portion of undergraduate medical education in these facilities. To slow the growth of Medicare spending, Section 223 of the Social Security Amendments of 1972 authorized the Secretary of Health, Education, and Welfare to determine prospective limits on the hospital costs reimbursed by the program. In 1974, per diem limits were set for hospitals' routine costs (i.e., for nursing and room and board, but not for ancillary services). Over time, the methods used to set the cost limits were revised. In 1979, it was decided that in calculating its routine per diem costs, a hospital could exclude all direct costs of GME programs. In 1980, the limits applied to teaching hospitals were adjusted to reflect the higher operating costs associated with treating patients in those hospitals (Lave, 1985): We believe these increases in per diem costs occur because the provision of graduate medical education causes increases in certain types of costs that are only indirectly related to education programs. (Department of Health and Human Services, 1980) This IME adjustment was based on the observation that teaching hospitals, particularly those with large residency programs, would be particularly vulnerable under the planned reductions in the routine cost limits. Moreover, the difference in costs across hospitals was found to increase with the intensity of the teaching program, as measured by the ratio of interns and residents to beds (Pettengill and Vertrees, 1982). Although both the name of the adjustment and the measure used to determine it implied that it was related primarily to the extra patient care costs generated by each resident, in fact it reflected all the reasons that teaching hospitals tend to be more expensive than other hospitals—including some that are not related to their teaching activities. The Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) extended the hospital cost limits to cover total operating costs per discharge and set a limit on the allowable rate of increase in those costs, but it retained the adjustment of

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund the cost limits according to teaching intensity. In addition, direct GME costs continued to be reimbursed on a reasonable cost basis under TEFRA. TEFRA also contained a provision that required the Secretary of Health and Human Services to develop a proposal for the prospective payment of hospitals under Medicare. The Secretary proposed to retain the payment for direct GME costs as a separate payment: Although graduate medical education is not directly related to the delivery of patient care to Medicare beneficiaries, these costs have always been paid by the Medicare program. This is not required by law, although the legislative history of Medicare indicates congressional intent that medical education costs be reimbursed by Medicare until the community undertakes to bear these costs in some other way. (Schweiker, 1982) Further, these direct GME payments would continue to be based on the hospital's reasonable costs: Direct medical education costs . . . will be passed through by PPS. . . . The Department believes that the direct costs of approved medical education programs should be excluded from the rate and be reimbursed as per the present system. This approach will assure that the base rate is related to a patient care outcome and not significantly influenced by factors whose existence is really based on objectives quite apart from the care of particular patients in a particular hospital. This approach will allow for continued Federal support of medical education through the Medicare program while clearly identifying that support as separate from patient care. (Schweiker, 1982) The ''pass-through'' payment for direct GME costs continued until the passage of the Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA), as described below, and the promulgation of rules based on this legislation. The Secretary's proposal for IME payments also continued the previous policy of recognizing the higher patient care costs at teaching hospitals: The indirect costs of graduate medical education are higher patient care costs incurred by hospitals with medical education programs. Although it is not known precisely what part of these higher costs are due to teaching . . . and what part is due to other factors . . . the Medicare Cost Reports clearly demonstrate that costs per case are higher in teaching hospitals. . . . Thus, not wanting to penalize these hospitals, an adjustment methodology has been developed which permits Medicare to pay teaching hospitals the same standard prices as other hospitals, while passing through the higher patient care costs associated with teaching hospitals. The Department believes that recognition of these indirect costs should be accomplished as a lump-sum payment, separate and distinct from the base rate. . . . The hospital's cash flow will be preserved by some sort of periodic payment. (Schweiker, 1982)

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund The Secretary's proposal formed the basis for Medicare's PPS, which was enacted in the Social Security Amendments of 1983. IME Payment Under PPS Under PPS, each hospital is paid a fixed rate per Medicare discharge to cover its operating costs. A separate PPS payment, determined in a similar but not identical way, is made for the hospital's capital costs. The per discharge payment is based on a standardized payment amount (there is one amount for hospitals in large urban areas and one amount for all other hospitals), but it is adjusted to reflect the local input prices faced by the hospital and the relative costliness of cases in the diagnosis-related group (DRG) to which each patient is assigned. This is the hospital's basic DRG payment. In addition, each hospital receives an outlier payment for cases that are exceptionally costly relative to other cases in the same DRG. The IME payments received by each teaching hospital are a percentage add-on to the sum of its basic DRG operating payments plus any outlier payments. The percentage for each hospital is determined by its ratio of residents to beds. There is a similar add-on to PPS capital payments, which is determined by the ratio of residents to occupied beds. Although the Secretary's proposal described the IME payment as a "lump-sum payment . . . separate and distinct from the base rate," it generally has been thought of as an adjustment to the hospital's payment rate for each case under PPS. The size of the IME operating adjustment has been modified over time. The Secretary proposed an IME adjustment of 5.795% for every increment of 0.1 in each hospital's resident-to-bed ratio. That is, a hospital with one resident for every 10 beds would receive about 5.8% in additional PPS payments; a hospital with twice as many residents per bed would receive twice as much. This proposal was based on an analysis of the relationship between Medicare operating costs per discharge and the resident-to-bed ratio across all hospitals. However, in response to projections that a vast majority of teaching hospitals would experience a reduction in Medicare payments under PPS, the Congress increased the IME adjustment: This adjustment is provided in light of doubts . . . about the ability of the DRG case classification system to account fully for factors such as severity of illness of patients requiring the specialized services and treatment programs provided by teaching institutions and the additional costs associated with the teaching of residents . . . the adjustment for indirect medical education costs is only a proxy to account for a number of factors which may legitimately increase costs in teaching hospitals. (House Ways and Means Committee and Senate Finance Committee Reports, March 1983, as cited in Fishman, 1996) The level of the adjustment was doubled to 11.59%. Since its implementation, the IME adjustment has been reduced twice, each time to reflect other changes in the payment system. In 1986, Congress added an

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund adjustment to PPS payments for hospitals that treat a disproportionate share of indigent patients. Because part of the rationale for doubling the original IME adjustment was that teaching hospitals tend to treat large numbers of poor people, and because teaching hospitals were expected to receive a high proportion of these disproportionate share (DSH) payments, the IME adjustment was reduced to 8.1% for every 10% increment in teaching intensity. In 1989, Congress made substantial changes in the formulas for determining DSH payments, and further reduced the IME adjustment to 7.7%. Changes in Direct GME Payments Even after the implementation of PPS, Medicare continued to pay its share of each hospital's actual direct GME costs. However, the enactment of COBRA changed the way that Medicare pays hospitals for the direct costs of teaching programs. The direct GME payment is a per-resident amount based on the hospital's per-resident costs in its base year (i.e., Medicare cost reporting periods beginning during FY 1984). The count of residents used in determining direct GME payments differs from that used in the IME adjustment formula in two ways. First, the hospital is allowed to count the time that residents spend outside of the hospital (under certain conditions). Second, residents beyond their period of initial residency (the minimum period of training required for board eligibility in a specialty or 5 years, whichever is less) are counted as 0.5 full-time equivalents. The base year per-resident amounts are updated to the current year using the Consumer Price Index for Urban Areas. However, in FY 1994 and FY 1995, the update was eliminated for residents in non-primary care specialties (i.e., specialties other than primary care, family practice, general internal medicine, obstetrics and gynecology, geriatrics, and public health and preventive medicine). Although the update has been applied to all specialties in subsequent years, this has resulted in different payment amounts depending on the resident's specialty. Payments to Teaching Hospitals Teaching hospitals treat a large share of Medicare patients and receive an even larger share of Medicare payments. There were 1,061 hospitals receiving IME payments in FY 1996, representing 21% of all PPS hospitals. These hospitals were responsible for 43% of discharges and 53% of all PPS payments. Teaching hospitals received almost $37 billion in PPS operating payments (Table B-1). They also received payments for their capital costs, and outpatient and other services they furnish, in addition to $2.5 billion for their direct GME costs.

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund TABLE B-1. Distribution of PPS Hospitals and Discharges and Estimated FY 1996 PPS Operating Payments, by Hospital Group PPS Operating Payments ($billions) Hospital Group Number of PPS Hospitals Percentage of PPS Discharges Percentage of PPS Operating Payment Total IME DSH All hospitals 5,153 100% 100% $69.1 $4.3 $4.3 Major teaching 246 11 19 13.0 2.8 1.4 Other teaching 815 32 34 23.7 1.4 1.4 Nonteaching 4,092 57 47 32.4 0.0 1.5 NOTE: DSH = disproportionate share; FY = fiscal year; IME = indirect medical education; and PPS = Prospective Payment System. SOURCE: ProPAC estimates based on ProPAC PPS Payment Model and FY 1994 MedPAR file data from the Health Care Financing Administration.

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund TABLE B-2. PPS and Total Margins, by Hospital Group, 12th Year of PPS Hospital Group PPS Inpatient Margin Percentage with Negative Margin Total Hospital Margin Percentage with Negative Margin All hospitals 7.9% 39.3% 5.6% 19.4% Major teaching 18.6 5.3 3.7 14.0 Other teaching 7.6 32.1 5.6 17.7 Nonteaching 3.7 42.7 6.5 20.0 NOTE: Data are preliminary and subject to revision. PPS = Prospective Payment System. SOURCE: ProPAC analysis of Medicare Cost Report data from the Health Care Financing Administration. In FY 1996 there were 246 hospitals with 25 or more interns and residents per 100 beds, sometimes referred to as major teaching hospitals. These hospitals represent about 5% of all hospitals, but are responsible for 11% of PPS discharges and received 19% of PPS payments in FY 1996. Medicare paid teaching hospitals $4.3 billion in IME payments in FY 1996. Many teaching hospitals also receive Medicare DSH payments which are related to the amount of care they furnish to poor patients. Teaching hospitals accounted for about 65% of the $4.3 billion in DSH payments in 1996. Both IME and DSH payments have risen rapidly since the implementation of PPS. GME payments also have risen rapidly. Moreover, there is wide variation in per-resident payment amounts across teaching hospitals. In the 10th year of PPS (1993), the average per-resident payment amount (of which Medicare pays a share) was more than $65,000. But the per-resident payment amount varied widely across hospitals: 10% of residents were in hospitals with per-resident amounts below $38,200, while another 10% were in hospitals with per-resident amounts above $98,600. One of the primary factors driving both IME and direct GME spending growth is a continuing increase in the number of interns and residents. Moreover, virtually all of the growth in the number of residents in recent years is due to a rise in the number of graduates of foreign medical schools receiving graduate medical training in U.S. hospitals. FINANCIAL CONDITION OF TEACHING HOSPITALS Medicare has more than adequately compensated teaching hospitals for the costs of treating Medicare patients through PPS. Since the beginning of prospective payment, teaching hospitals' PPS margins have exceeded those of other hospitals. Further, over the years the gap between the margins of teaching and nonteaching hospitals has widened. Preliminary data for 1995 indicate that major teaching hospitals had the highest PPS margins of any group of

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund hospitals—18.6% (Table B-2). The PPS margin for other teaching hospitals—those with less than 25 interns and residents per 100 beds—was 7.6%, and for nonteaching hospitals it was 3.7%. Total hospital margins, however, show a very different pattern. The preliminary data for 1995 show a total margin for major teaching hospitals of 3.7%, the lowest of any group of hospitals. The total margins for other teaching hospitals and for nonteaching hospitals are closer together at 5.6% and 6.5%, respectively. The reasons for the lower total margins among major teaching hospitals are difficult to disentangle. Possible explanations include the large amount of uncompensated care many of these hospitals furnish, inefficiencies that are driving up costs per case, and difficulties obtaining the revenue from private payers to support the extra costs of maintaining teaching programs. This latter concern is evident in comparing the gains and losses by payer for hospitals grouped by teaching status (Table B-3). Major teaching hospitals (as defined above) fare somewhat better than other hospitals under Medicare. Medicaid DSH payments, which generally favor state hospitals, are reflected in the smaller Medicaid losses for the public major teaching group. Other government programs, as expected, also are targeted toward public hospitals. However, public major teaching hospitals have a substantially greater uncompensated care burden than do other hospitals. Moreover, major teaching hospitals appear to lack the ability to generate large surpluses from private payers. In particular, the public major teaching hospitals have only about half the nationwide share of private payer patient load, and so cannot leverage their surplus payments to offset their large uncompensated care losses. CONCLUSIONS In FY 1996, Medicare provided $6.8 billion in added direct and indirect GME payments to teaching hospitals. In addition, these hospitals received $2.8 billion in Medicare DSH payments. Thus a total of $9.6 billion in GME and DSH payments went to the entire pool of hospitals with teaching programs. Of that group, the subset of institutions (246) that qualify as major teaching hospitals received $4.2 billion in additional PPS (IME and DSH) payments. These payments led to very high PPS margins among major teaching hospitals—18.6% in 1995, compared to 3.7% for nonteaching hospitals. The total margins of these hospitals, however, are much lower than for other hospitals—only 3.7% in 1995, in contrast to 6.5% for nonteaching hospitals. Despite these payments, 14.0% of major teaching hospitals had negative total margins in 1995. The extra payments from Medicare have helped many major teaching hospitals avoid severe financial stress and have allowed them to continue providing care to Medicare enrollees. However, it is feared that accelerating price competition in the private sector will reduce the ability of teaching hospitals to obtain the higher patient care rates from other payers that

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund traditionally have contributed to financing the costs of medical education and the other missions that teaching hospitals pursue. At the same time, the methods used to distribute these payments across hospitals provide a strong incentive to hire more residents and a strong disincentive to reduce their number. Under current Medicare rules, each residency slot unused by the hospital results in a reduction of almost $70,000 in Medicare (IME and direct GME) payments. Moreover, the per-resident payment varies substantially across hospitals, with some hospitals receiving well in excess of $100,000 per resident. Although this payment is based on actual costs allocated by the training program to each hospital, it can severely distort decisions as to how many and what types of residents to hire, as well as where to train them.

OCR for page 54
On Implementing a National Graduate Medical Education Trust Fund TABLE B-3. Gains or Losses by Payer, by Hospital Group, 1994 Gains or Losses as a Percentage of Total Costs Private Payers Hospital Group Total Gainsa Medicareb Medicaidb Other Govt. Payers and Subsidiesc Uncompensated Carec Private Payers Nonpatient Care Share of Total Costs Payment-to-Cost Ratio Public, major teaching 2.3% -0.3% -0.5% 0.9% -8.0% 6.5% 3.8% 19.1% 133.9% Nonpublic, major teaching 3.8 -1.1 -1.7 0.2 -4.7 7.3 3.8 38.8 118.9 Other teaching 5.4 -2.9 -1.2 0.1 -4.6 10.8 3.1 38.1 128.4 Nonteaching 6.1 -3.7 -1.7 0.2 -4.5 13.1 2.6 36.6 135.8 NOTE: This analysis by hospital group is limited to data for hospitals that submitted all required data elements. Most revenue (and corresponding costs) from Medicaid health maintenance organizations and Medicare risk plans are included in the private payer category. a Gains or losses as a percentage of total costs for each of the six sources of revenue shown sum to total gains as a percentage of total costs. The values shown would be slightly lower if expressed as total margins, where total gains are calculated as a percentage of total revenue rather than total costs. b Medicare and Medicaid costs equal all costs, both inpatient and outpatient, attributed by hospitals to these programs' patients regardless of whether the costs are reimbursable by the programs. c Operating subsidies from state and local governments were considered as payments for uncompensated care, up to the level of each hospital's uncompensated care costs. Additional subsidies above this level were counted as separate revenue. SOURCE: ProPAC analysis of data from the American Hospital Association Annual Survey of Hospitals.