Share this article on:

Complexity and Uncertainty in Financing Graduate Medical Education

Knapp, Richard M. PhD


The current sources and structure of graduate medical education (GME) financing policies are being challenged. To foster understanding of these challenges, the author explains the current GME financing system and explores the impacts that various policy changes and legislation under consideration would have on residency programs and the health of already financially stressed teaching hospitals. Specifically, the wide variety of revenue sources (mostly third-party payers) are discussed, particularly the present Medicare direct graduate medical education (DGME) payments and indirect medical education (IME) adjustment payments, and their policy implications. The author concludes that a serious problem in advocating a new approach to the financing of GME (which would focus on changes in Medicare support) is the absence of a common vision, not only among policymakers but also within the academic medical community. Too much energy has been expended on how GME and the special missions of teaching hospitals should not be funded, and not enough attention focused on how to achieve a dependable source of revenue to help finance them. The public policy challenge is how to sustain Medicare's traditional support for GME in the face of intense pressure to “save” Medicare through fundamental reforms.

Dr. Knapp is executive vice president, Association of American Medical Colleges, Washington, D.C.

Correspondence and requests for reprints should be sent to Dr. Knapp, AAMC, 2450 N Street, NW, Washington, DC 20037.

For another article on a related topic, see page 1069.

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 considered as an element in the cost of patient care, to be borne to an appropriate extent by the hospital insurance program.1

This statement in the House of Representatives and Senate reports accompanying the enactment of the Medicare program in 1965 has served as a guiding principle for the financing of graduate medical education (GME) by Medicare and other purchasers of health care services over the last 35 years. Importantly, this construct provided a rationale for purchasers' explicit support for GME. In its practical application, the financing policy has produced a reliable and steady stream of revenue to finance the costs of educating and training the nation's physicians and other health care professionals.

Beginning in 1995, legislators began a serious debate over this policy. Reports of the National Bipartisan Commission on the Future of Medicare and the Medicare Payment Advisory Commission (MedPAC), both released in 1999, raised the stakes of this debate by calling into question why and how Medicare supports GME and recommending major policy changes to that support.

In order to evaluate these challenges to the current source and structure of GME financing policies, it is important to understand clearly the current system and the changes being recommended. The purpose of this paper is to explain the current GME financing system and to explore the impacts that various policy changes under consideration would have on financing the nation's physician training programs and the teaching hospitals that sponsor and support them.

Back to Top | Article Outline


The current financing of direct graduate medical education (DGME) costs consists of revenues from many diverse sources, but the vast majority of support comes from revenues paid to teaching hospitals by third-party payers. These revenues include payments from commercial insurance and managed care companies, government-financed programs such as Medicare and Medicaid, and state and local appropriations. Other revenue sources that also provide support for direct GME costs include fees from faculty physician practice plans, foundation and federal government grants, and other diverse non-hospital sources, in addition to the Veterans Health Administration (VHA)2 and the Department of Defense (DoD).

Purchasers of health care services participate in the financing for the graduate training of physicians using different payment mechanisms.

  • The Medicare program makes explicit payments to teaching hospitals for its share of the direct costs of GME.
  • The Medicaid program, financed through a federal—state partnership, makes payments in many, but not all, states for GME-related costs through a variety of explicit and implicit mechanisms.3
  • Private insurance and managed care companies have contributed implicitly to direct GME costs by paying higher prices for services provided to their beneficiaries or policy holders than these companies pay to non-teaching hospitals. The portion of these higher prices that goes to defraying the costs of GME (as well as other activities that justify a higher payment level) is neither separately negotiated nor specifically identified. As result, determining the level of private payers' participation in GME financing is virtually impossible.

The Medicare program has become the principal battle-ground for debate about financing GME, not only because of its explicit, proportionate support of GME's direct costs, but also because the program provides teaching hospitals with a second explicit payment with an education label, known as the “indirect medical education” (IME) adjustment.

The complexity and importance of these two Medicare payments with an educational label underscore the need for a full understanding of the rationale for these federal dollars and the method by which they are paid. These payments are critical to the financial viability of teaching hospitals. (See Table 1.) In 1999, according to MedPAC, total margins for major teaching hospitals were only 2.4%, compared with 4.0% for other teaching hospitals and non-teaching hospitals.

Table 1

Table 1

Back to Top | Article Outline


DGME costs include residents' stipends and fringe benefits, the salaries and fringe benefits of faculty who supervise residents, allocated overhead costs and other direct costs, including clerical personnel who work exclusively in the GME administrative office, and other directly assigned costs. Institutional overhead costs, such as portions of maintenance, cafeteria, and depreciation costs, are assigned to the DGME cost centers through an allocation process. Together, these various costs are defined as a hospital's “Medicare-allowable” DGME costs.

From 1965 until 1985, Medicare paid for its share of each hospital's historical, allowable DGME costs. Reimbursement was open-ended: if a hospital increased its DGME costs, the Medicare program would pay its share of the increased allowable costs incurred.

In April 1986, Congress passed the Consolidated Omnibus Budget Reconciliation Act of 1985 (P.L. 99–272), which dramatically changed the DGME payment method in two ways. First, Medicare established hospital-specific per-resident amounts. Second, in general, the Medicare program limited the number of years for which it fully supports its share of the training of residents to the amount necessary for residents to complete their initial residency program.

To determine the per-resident amount, a hospital calculates its base-year (usually FY 1984) Medicare-allowable DGME costs, and divides that figure by the average number of full-time-equivalent (FTE) residents present in all areas of the hospital complex in that year. The per-resident amount is then updated from the base year to the payment year using the Consumer Price Index (CPI) and multiplied by the number of FTE residents in that year. Medicare's share of this total dollar figure is calculated by multiplying it by the ratio of Medicare inpatient days to total inpatient days.4 Because the costs included in individual hospitals' base-year amounts differ substantially, wide variation exists among hospital per-resident amounts. This variation has been a source of tension and a policy focus in recent years. The extent to which some residents can be fully counted in this formula and a few other nuances also affect DGME payments.

In the context of Medicare DGME payments, six policy issues have dominated the debate over the 17 years since the passage of P.L. 99–272:

  • DGME payments to non-hospital entities
  • The duration of training for which DGME payment is made
  • Limiting payments to certain types or a defined number of residents
  • Differential payments by specialty
  • Payments directly to programs versus hospitals
  • Limits on the variation in per-resident amounts

None but the last of these issues has received significant attention, although, in the Balanced Budget Act (BBA) of 1997 (P.L. 105–33), a provision was included that allowed HCFA (i.e., the Health Care Financing Administration, now known as the Centers for Medicare and Medicaid Services) to pay a limited number and type of non-hospital sites for DGME costs. The BBA also added another major policy issue when it capped the number of residents who can be counted for purposes of Medicare DGME payments. The cap also applies to IME payments.

Back to Top | Article Outline

Medicare Resident Limits

In a provision that at the time generated no opposition, resident limits were imposed by the BBA. In general, the law mandated that, for purposes of Medicare IME reimbursement—42 U.S.C. 1395 ww(d)(5)(B)(v)—and direct GME reimbursement—42 U.S.C. 1395 ww(d)(h)(4)(F)—a hospital's number of allopathic and osteopathic residents may not exceed the number reported on the hospital's most recent cost report that ended on or before December 31, 1996. Dental and podiatry residents are excluded from the resident-limit provision.

To provide context for the environment in which this provision was enacted it is important to note that in 1996 six major national organizations with a major stake in physician workforce policies issued a consensus statement that said:

During the past few years, studies of the physician workforce have produced compelling evidence that the United States is on the verge of a serious oversupply of physicians. The attendant consequences of physician oversupply—the under-employment and unemployment of physicians—are highly undesirable from the perspective of both the society at large and the individual physicians who are affected. Given this, the current rate of physician supply (the number of physicians entering the workforce each year) is clearly excessive.5

Since 1996, the consensus of opinion has changed substantially. In fact, in June 2002 the Association of American Medical Colleges (AAMC) revised its physician workforce position statement. The policy now states:

After reviewing the range of studies over the past several decades, and the currently available analyses of the potential supply and demand for physicians, the AAMC has concluded that no definitive conclusions can be drawn about the adequacy of the workforce, nor can specific recommendations be made about the rate of supply of new physicians.

It is also the view of many medical center leaders and residency program directors that because Medicare is such an important source of financial support for training programs, the cap is beginning to impede the continued development of the educational missions at many institutions.

In some parts of the country there is a need for more emergency room physicians (and other specialists), yet the ability to start emergency room residency programs is being thwarted by the resident limits. There also are emerging specialties that are in the process of being recognized by the Accreditation Council for Graduate Medical Education (ACGME). The resident limits are barriers to teaching hospitals' abilities to incorporate these “cutting edge” specialty residency programs without diminishing other programs. As the population ages, additional residency programs will also be needed to ensure a pipeline of specialists in areas such as cardiology, rheumatology, and oncology.

There are other problems associated with the resident limits, including the fact that the BBA provision is based on a snapshot of activity, essentially “freezing” the status of residency education at a random point in time—1996. A number of institutions were undergoing transitions during that time such that 1996 was an anomaly in terms of residency counts. For these institutions, the resident limits can have profound effects on their educational missions. The opinion that the cap on Medicare payments for residency positions should be removed or at least substantially modified has gained strength, particularly in the last 18 months.

(It is worth noting that on June 28, 2002, the U.S. House of Representatives passed legislation—entitled the Medicare Modernization and Prescription Drug Act of 2002, H.R. 4954—that includes a complex provision to redistribute a portion of a hospital's resident limit slots that are “unused” to teaching hospitals seeking to increase their resident limits. The slots that would be redistributed would be eligible for DGME payments but not IME payments. The Senate had not acted at the time this article went to press.)

Back to Top | Article Outline

Variation in Per-resident Payment Amounts

With respect to the issue of variation in per-resident amounts, the conference committee report accompanying the 1997 BBA stated,

The Conferees are aware of and concerned about studies and reports from the Prospective Payment Assessment Commission [ProPAC] and the Physician Payment Review Commission [PPRC] that describe wide variation in hospital-specific per resident payment amounts. The Conferees have directed the Secretary [of Health and Human Services] to study and report back to the Committees of jurisdiction on reasons for such variations and to provide recommendations to reduce such variation, as appropriate, to provide greater payment equity among all teaching facilities receiving direct graduate medical education payments.6

That study was due in August 1998, but has yet to be submitted to the committees.

The per-resident variation issue was addressed, however, by an unexpected initiative in the Medicare, Medicaid, and State Chilcren's Health Insurance Program (SCHIP) Balanced Budget Refinement Act of 1999 (P.L. 106–113), signed by President Clinton on November 29, 1999. In brief, the law created a payment floor and ceiling for hospital per-resident amounts. Beginning in FY 2001, a hospital's per-resident amount was adjusted if it was below or above a “corridor,” or range, surrounding a weighted standardized national average per-resident amount, modified by the geography-specific factor used to adjust physician payments (the average of the three geographic practice-cost indices).

The national average per-resident payment amount was calculated based on a hospital's per-resident amount for its cost report period ending in FY 1997, standardized by the 1999 physician geographic adjustment factor. The individual hospital amounts were then summed, each weighed by the hospitals' respective number of FTE residents, and then divided by the national number of residents. The national average per-resident amount was then updated to FY 2001 by the increase in the CPI. For each geographic area, a locality-adjusted national average was then calculated by applying the physician geographic adjustment factor for each area to the 2001 national average.

Beginning in FY 2001, a floor has been calculated for each hospital at 70% of its locality-adjusted national average per-resident amount. In addition, 140% of the locality-adjusted national average per-resident amount has been calculated. If a hospital's FY 2001 per-resident amount is below the 70% floor, its per-resident amount will be increased to the 70% level in FY 2001 and updated annually by the CPI increase. If a hospital's per-resident amount is above the 140% level of the locality-adjusted national average, the per-resident amount is frozen at FY 2000 levels for FYs 2001 and 2002, and in FYs 2003–2005, the per-resident amount will be updated by the CPI increase minus two percentage points. If a hospital's per-resident amount is between 70% and 140% of the locality-adjusted national average, the per-resident amount will continue at current levels, updated annually by the CPI increase.7 The Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act (BIPA) of 2000 (P.L. 106–554) raised the per-resident amount floor from 70% to 85% of the average effective in FY 2002; BIPA made no changes to the 140% provision. (Another provision of the previously mentioned H.R. 4954 includes a DGME provision to extend through 2012 the freeze on updating a hospital's per-resident amount above 140%.)

According to an analysis by the Association of American Medical Colleges, of the 1,100 hospitals that provide some level of GME, the 1999 provision resulted in increased DGME payments for 292 hospitals and payment freezes for approximately 130 hospitals. Almost 700 teaching hospitals were unaffected by this change. By raising the minimum payment level from 70% to 85% of the average per-resident amount, BIPA resulted in increased DGME payments for the 292 hospitals affected by BBRA as well as for an additional 213 hospitals.8

While the 1999 provision came as a surprise to some institutional executives, others had been quietly advocating a legislative change of this type to what they considered to be an unfair payment policy that permitted such wide variation in payment rates. The issue also was highlighted in a Health Policy Report in the July 22, 1999, issue of the New England Journal of Medicine.9 At the same time, and as reported earlier, this issue had been on the agendas of ProPAC and PPRC. Additionally, many policy analysts point out that all Medicare payment policies have been moving in the direction of a wage-adjusted national average since the inception in 1984 of a national prospective per-admission payment strategy for inpatient care based on diagnosis-related groups (DRGs).

Importantly, recently enacted legislation also provides direct GME payments to independent children's hospitals and stipulates that these payments are to be made on the basis of a standardized national average per-resident amount adjusted for a hospital's area Medicare wage index.10 This development reinforces the view of many policy analysts that the issues of how much to pay, and on what basis to do so, will have to be resolved before any mechanism for financing GME that goes beyond the current Medicare method can be devised.

Back to Top | Article Outline


The label “IME adjustment” is misleading because the purpose of this payment is much broader than to support GME. It compensates teaching hospitals for their higher Medicare inpatient hospital operating costs due to a number of factors. One important factor is the failure of the DRG patient-classification system to capture fully the greater patient-specific severity of illness characteristic of teaching hospitals. Additionally, the operating costs associated with the presence of GME programs (but not allocable directly to them) are higher; these costs include factors such as lower staff productivity and the use of more diagnostic tests by less experienced residents. Moreover, in addition to attracting more severely ill patients, teaching hospitals experience high standby costs for low-volume services. The following statement, contained in the 1983 reports of the House Ways and Means and Senate Finance Committees, captures the intent of Congress:

The [IME] 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 is only a proxy to account for a number of factors which may legitimately increase costs in teaching hospitals.

Unlike DGME costs, “indirect” operating costs in hospitals with GME programs cannot be quantified specifically; they are imbedded in the costs of inpatient care and can only be estimated through statistical analysis. The IME adjustment is provided through the Medicare Hospital Inpatient Prospective Payment System (PPS) as a percentage add-on to the basic DRG amount. Teaching hospitals receive an additional payment for every Medicare inpatient they treat. A given hospital's IME payment is determined by its individual intern/resident-to-bed (IRB) ratio and a nation-wide adjustment factor. The adjustment factor is determined (and has been changed periodically) by Congress and is based on a statistical relationship between teaching hospital costs and the number of residents per hospital bed. For example, in FY 1999, each DRG payment a hospital received was adjusted upward by approximately 6.5% for every ten residents per 100 beds in that hospital.

The FY 1999 6.5% payment rate for IME is a result of the BBA, which reduced the IME adjustment from its 7.7% level in 1997. Although the BBA was to have continued reducing the IME adjustment to 6.0% in FY 2000 and to 5.5% in FY 2001 and beyond, the BBRA and BIPA postponed the continued decline. Under current law, IME payments will be maintained at 6.5% through FY 2002 before being reduced to 5.5% in FY 2003 and thereafter. H.R. 4954 would set the IME payment at 6.0% in FY 2003 and 5.9% in FY 2004 before reducing it to 5.5% in 2005.

The IME payment reductions, as well as other payment reductions, have occurred at the same time that pressures are increasing from the Medicaid program, as well as from commercial insurance and managed care companies, to pay teaching hospitals at the same rate as non-teaching hospitals. As a consequence, the financial position of teaching hospitals, as documented by countless news accounts from across the country, has become precarious.

While these financial woes cannot be laid totally at the doorstep of the 1997 BBA, the cuts called for by that Act have certainly been a major contributing factor. Consider the following. Assuming the same inpatient volume and DRG mix of Medicare patients, the vast majority of AAMC teaching hospital members received less money in aggregate for serving Medicare inpatients in FY 2000 than they did in FY 1997—and these are actual dollars, not adjusted downward for inflation. How can this be? Following a freeze in FY 1998, the Medicare inpatient DRG rate was increased by only 0.5% in FY 1999 and 1.1% in FY 2000. IME payments were reduced, as noted above, by 9.1% in FY 1998, and by another 6.5% in FY 1999. Disproportionate share (DSH) payments were also decreased during this period, reducing the amount of Medicare payments received by teaching hospitals for the care of the poor and uninsured. The steep IME cuts, plus the other cuts, more than offset the small increases in the DRG payment rates.

Back to Top | Article Outline


Medicare Commission

Early in the National Bipartisan Medicare Commission on the Future of Medicare's deliberations, several members questioned whether the Medicare program should be financing GME at all or, for that matter, any other missions not directly related to the care of Medicare beneficiaries. Senator Bill Frist (R-TN) chaired a working group of commission members that filed a formal report on February 12, 1999. The Frist report outlined in some detail a proposal to move the financing of DGME payments from the Medicare Part A Trust Fund to a mandatory or discretionary appropriation. In the report, the Senator states, “I support maintaining the entitlement for DGME and recognize this would be accomplished through a mandatory appropriation. However, I recognize that others on the Commission would prefer DGME be subject to the same discretionary appropriations process as biomedical research or other education programs. In either scenario, I recommend stringent safeguards to ensure adequate levels of funding should Congress enact a premium support model.”11

Notwithstanding the fact that the final commission report did not achieve the 11 votes necessary to recommend it formally to the Congress, the report did surface an alternative concept for structuring GME funding. In the commission's final report, DGME and IME payments are dealt with as follows:

Congress should examine all non-insurance functions, special payments and subsidies to determine whether they should be funded through the Trust fund or from another source. For example, payments for Direct Medical Education (DME) would be financed and distributed independent of a Medicare premium support system. Since the Part A and Part B trust funds would be combined and the traditionally separate funding sources of payroll taxes and general revenue would be blurred, Congress should provide a separate mechanism for continued funding through either a mandatory entitlement or multi-year discretionary appropriation program. On the other hand, Indirect Medical Education (IME) presents a unique problem since it is difficult to identify the actual statistical difference in costs between teaching and non-teaching hospitals. Therefore, for now Congress should continue to fund IME from the Trust Fund as an adjustment to hospital payments.12

The financing of DGME through the appropriations process raises a host of worrisome potential consequences. First and foremost, the dependability of funding for DGME would be eroded. Under an appropriations process, GME would have to compete with other worthy priorities. The GME function would probably fall within the jurisdiction of the House and Senate Labor, Health and Human Services and Related Agencies Appropriations Subcommittees, where children's hospital GME and other health professions programs (Titles VII and VIII) of the Public Health Service are currently funded. Here it would compete with the National Institutes of Health and, indeed, the entire Public Health Service as well as programs in the Departments of Labor and Education. Some observers suggest GME would fare well in a competition for funds. However, “focus groups” and survey work conducted by the AAMC of voting Americans and congressional staff suggest that the financing of physician education is a distant third in competition with biomedical research and patient care programs.

Comparatively, the current entitlement status of Medicare financing provides a much more secure source of funding. While this entitlement method of funding can be changed, as it was in major ways in 1985, 1997, 1999, and 2000, it takes a specific affirmative act to amend the Medicare program, in contrast to the appropriations process, which is an annual event permitting many opportunities for legislators to intervene. Additionally, GME programs three to seven years in length need stable, predictable funding, which is unlikely to emanate from the success or failure of a particular year's funding fortunes in the appropriations process.

A good example of this problem is the fact that in FY 2002 the Children's Hospital Graduate Medical Education Program received an appropriation of $285 million. The administration has proposed only $200 million for FY 2003. This is the kind of budget battle that will be fought every year.

Further, the annual appropriations process would open the training of physicians to political influence. For example, funding DGME through the appropriations process will almost certainly make it a battleground for those legislators who would like to mandate what is taught in residency programs. The curriculum of GME has already received much attention from lawmakers, who seek to require or eliminate the teaching of such subjects as domestic violence, end-of-life-care, and abortion. The AAMC and the medical community vigorously support protecting the medical education curriculum as the purview of medical educators.

Buildings financed through the appropriations process not infrequently carry the names of particular legislators. While “programs” might be less susceptible to such “naming opportunities,” one can easily imagine a Ronald Reagan emergency medicine program in Washington, D.C. Even if such a visible action did not occur, it is certainly of interest to note 199 construction awards and 196 program projects in the health arena devoted to specific universities, hospitals, and other organizations were included in the FY 2001 Labor, Health and Human Services, Education and Related Agencies appropriations bill (P.L.106-554). This is not the way GME programs should be established or funded. Funding for primary care programs as well as for programs designed to train residents in short-supply specialties has been an issue in the current debate over Medicare funding. This issue would be intensified in an appropriations process in which many opportunities would exist for those favoring one discipline or another to get special treatment.

Senator Frist favored what is called a mandatory appropriation with substantial “safeguards” to protect GME financing in the context of reforming Medicare to a premium support system. Mandatory appropriations finance Medicaid, a portion of Medicare Part B funding, food stamps, agricultural subsidies, and other similar programs. Perhaps this is “safer ground” than the discretionary appropriations process. However, it is still very susceptible to the vagaries of the appropriations process previously identified.

In reviewing the history of financing GME in this context of a mandatory appropriation, it is useful to remember that in 1995, House Ways and Means Committee Chairman Bill Archer (R-TX) championed the Teaching Hospital and Graduate Medical Education Trust Fund. That trust fund would have derived its funding from the Medicare Trust Fund in addition to $13.5 billion in “mandatory” appropriations from FYs 1997–2002. The provision was included in the “Medicare Preservation Act of 1995” vetoed by President Clinton on December 6, 1995, for reasons unrelated to the trust fund initiative.

Back to Top | Article Outline

Medicare Payment Advisory Commission

The Balanced Budget Act of 1997 required the Medicare Payment Advisory Commission (MedPAC) to examine “…and develop recommendations on whether and to what extent Medicare payment policies and other Federal policies regarding teaching hospitals and graduate medical education should be changed.”

In MedPAC's August 1999 report13 to Congress on GME, the commission presented recommendations that focus on a new conceptual framework for Medicare's GME payment policy. Based on the commission's consensus view of residents' incurring the costs of their own training by accepting lower wages than they could otherwise receive in a competitive market, MedPAC argued that it would be more appropriate to combine the DGME and IME costs into a single adjustment to DRG payments that hospitals receive. Called a “teaching hospital adjustment,”14 it would be phased in over a period of years.

The commission's report states that the recommended new framework for changing historical GME payments is not intended to produce savings for the Medicare program. However, it is difficult to imagine a reduction in current payments would not occur, since the Commission has consistently recommended a reduction in the IME payment level. MedPAC's proposal appropriately recognizes the valuable patient care services provided by teaching hospitals, but discounts the critically important role teaching hospitals play in GME. MedPAC's report sends a very troubling message to Congress by concluding that Medicare should no longer explicitly support physician training.

Back to Top | Article Outline

An All-payer Approach to GME

In the 107th Congress there is legislative support, albeit slim and totally Democratic, to require all payers, including Medicare, to contribute to an “all payer” trust fund to support GME and the other special missions of teaching hospitals.

Two legislative proposals have been introduced that embrace this financial construct. The Medical Education Trust Fund Act (S. 743), introduced in the Senate by Jack Reed (D-RI), recognizes that all payers of the health care system should share the responsibility to fund GME. The bill requires Medicare, Medicaid, and the private payer community to contribute to the support of medical education through a 1.5% assessment of all health insurance premiums that would be pooled into the fund, providing total annual payments of $16 billion. The bill is co-sponsored by Senators Hillary Rodham Clinton (D-NY), Mary Landrieu (D-LA), and Chuck Schumer (D-NY).

The All Payer Graduate Medical Education (GME) Act (H.R. 2178), sponsored by Representative Ben Cardin (D-MD), establishes a trust fund through which private payers would contribute to GME; this bill would continue Medicare, Medicaid, and Veterans' Health care programs' commitments to physician training and the other special missions of teaching hospitals through their current financing mechanisms. The all-payer fund would be financed by assessing private health plans a 1% premium tax. Overall, the bill is estimated to increase total net DGME- and IME-designated hospital revenue by $4.1 billion.

The bill continues Medicare's contribution to GME, but makes changes to the DGME formula and reduces Medicare IME payments. The bill's changes to the Medicare DGME and IME payments would save Medicare an estimated $1.4 billion per year. However, $300 million of the $1.4 billion in Medicare savings would be used to reimburse payments for graduate education for non-physician health professionals.

The bill directs the Secretary of Health and Human Services, in concert with the AAMC and other affected community representatives, to develop and implement a plan to reduce the number of residency training positions to 110% of American medical school graduates by 2005. In addition, the bill directed the Secretary to consider the financial impacts of residency reductions on hospitals and to allow a portion of the money saved by the residency reductions to be used to support affected hospitals.

Congressman Cardin's bill currently has 17 co-sponsors.

Back to Top | Article Outline


A serious problem in advocating an alternative philosophical concept to finance GME is the absence of a common vision, not only among policymakers but also within the academic medical community.

The reports of the Bipartisan Medicare Commission and MedPAC as well as renewed calls for “Medicare Reform” ensure a broader debate of the issue of GME financing. Prior to the emphasis placed on the issue by these two commissions, only a handful of legislators and staff members had intimate knowledge of the facts and engaged in debates over the issue.

As noted above, there is now one report favoring the reassignment of Medicare DGME payments from the Medicare Trust Fund to the appropriations process, where they would be financed by general revenue, and a second report recommending that the DGME payment be combined with the IME payment into a single payment adjustment.

Too much energy has been expended on how GME and the special missions of teaching hospitals should not be funded. Not enough attention has been focused on how to achieve a dependable source of revenue to support those missions. Medicare has long provided its fair share of predictable, stable support of GME as well as of other essential functions of teaching hospitals, such as caring for the severely ill, maintaining highly specialized services around-the-clock, and providing an environment in which clinical research can flourish. The nation's teaching hospitals and their affiliated medical schools have achieved world leadership in physician education, research, and advanced patient care thanks in no small part to Medicare's indispensable support. The public policy challenge for the immediate future is how to sustain that support in the face of intense pressure to “save” Medicare through fundamental reforms.

Back to Top | Article Outline


1. House Report, Number 213, 89th Congress 1st Session. Report 32 (1965) and Senate Report, Number 404, 89th Congress 1st Session. Report 36 (1965).
2. The VHA is the largest single provider of training sites for GME in the United States today and funds about 9% of all residency positions each year at 107 VA medical centers.
3. Henderson TM. Funding of Graduate Medical Education by State Medicaid Programs [National Conference of State Legislatures for the Association of American Medical Colleges]. Washington, DC: AAMC, April 1999.
4. This brief description of the Medicare DGME payment and the IME payment draws heavily on the work of Linda Fishman when she worked at the Association of American Medical Colleges (AAMC). She wrote a detailed AAMC publication of these payment policies in 1996 entitled Medicare Payments with an Education Label.
5. Consensus Statement on the Physician Workforce. Statement from the American Association of Colleges of Osteopathic Medicine, American Medical Association, American Osteopathic Association, Association of Academic Health Centers, Association of American Medical Colleges, and National Medical Association. Washington, DC: 1996.
6. Balanced Budget Act of 1997, House Conference Report to Accompany H.R. 2015, 105th Congress, 1st Session, Report 105–217, July 30, 1997:823.
7. For a more thorough description of this method, see the fiscal year 2001 Medicare hospital inpatient prospective payment system final rule (65 Fed. Reg. 47054, 47090, Aug 1, 2000).
8. Moy E. Impact of 2000 Medicare legislation on teaching hospitals' Medicare direct graduate medical education payments. Analysis in Brief [short news bulletin of the Association of American Medical Colleges]. 2001;1(1).
9. Iglehart JK. Support for academic health centers. N Engl J Med. 1999; 341:30–4.
10. Healthcare Research and Quality Act of 1999 (S.580), Subpart IX, Section 340E.
11. Frist W. Protecting the future of graduate medical education: a look at financing within the premium support model. Paper presented at the GME Study Group February 12, 1999. Sen. First was chair of that group.
12. Building a Better Medicare for Today and Tomorrow. Report of the National Bipartisan Commission on the Future of Medicare. Washington, DC: Government Printing Office, March 15, 1999.
13. Medicare Payment Advisory Commission: Report to the Congress: Rethinking Medicare's Payment Policies for Graduate Medical Education and Teaching Hospitals. Washington, DC: Government Printing Office, August 1999.
14. This term was coined after the August 1999 report, and has been used by MedPAC and included in all of its publications since that time.
© 2002 Association of American Medical Colleges