The Institute of Medicine report that follows represents the suggestions of the Institute's Committee on Implementing a National Graduate Medical Education Trust Fund formed in response to a specific congressional request. We decided at the beginning that our charge required us to focus on defining a set of principles and, flowing from them, recommendations for a new payment structure that could be supportive of graduate medical education and teaching hospitals, open to additional entrants, less restrictive than at present, and more responsive to changes in the health care marketplace. We decided, further, not to attempt to define policies for international medical graduates or for the size or composition of the work force.

This report reflects compromises made among the disparate views of the committee members to reach language that all could accept. We are well aware that other responses to the initial questions could have been made by reasonable and knowledgeable people; such is the nature of the policy process. Some committee members were willing to recommend a general, non-Medicare source of funds for graduate medical education; others were reluctant to do so, either because they questioned tax support or because they felt it was beyond the committee's charge. In the end, the committee's plan supports the use of general non-Medicare funds, but notes that implementation of a Medicare-only trust fund (the present funding source) would go far to improve the current distribution.

Strong advocates for more incentives for primary care noted that payment inevitably influences work force and questioned reliance on the market to achieve a better balance among physician specialties. These members would have preferred a more prescriptive distribution plan, whereas others favored a more neutral plan that would permit greater responsiveness to market conditions. The report adopts a middle ground. We suggest retaining existing specialty weights for direct graduate medical education payments, while making other

changes that would provide freer response to market forces. Additional issues where different views might have prevailed include the relationship of an institution's Medicare caseload to payment; nursing education; and funding of institutions other than teaching hospitals. The discussions at the committee's meetings were vigorous and constructive, and encompassed a wide range of views.

As committee chair, I believe that we have formed a reasonable response in the short time we were given for completion and in the context of rapid, ongoing changes in the environment of graduate medical education. This was a challenging task given the variables and uncertainties as to the best course to follow. Although the committee members are unanimous in supporting our proposals as a feasible, fair approach to the next step in funding graduate medical education, these recommendations are made with the expectation that as experience accrues and is monitored, changes from our approach may appropriately be considered. We offer this report, therefore, as one among other contributions to the policy debate.

Rosemary A. Stevens,

Chair

The size and nature of the work force required for the health professions have presented difficult challenges to public-and private-sector planners in the United States for decades. Based upon predictions of serious shortages of physicians for the growing U.S. population after World War II, federal financial incentives were offered to increase medical school enrollments and to open new schools. As the number of U.S. graduates increased and the number of international medical graduates coming to the United States accelerated, work force planners predicted an oversupply of physicians—a conclusion reached in the 1996 Institute of Medicine (IOM) report on the physician work force (IOM 1996a).

Federal support of graduate medical education (GME) has been accepted as an important social value to be funded through the Medicare program, with substantial additional contributions from streams of clinical income generated by teaching hospitals and medical faculty practice plans. This acceptance was also based on a general recognition that there were legitimate costs to teaching hospitals associated with their educational mission even though such costs were hard to identify specifically. Hospitals that undertake education also provide high-quality care to Medicare beneficiaries as well as unsponsored research, complex care, and technology development, and they and their young physicians in training are important in providing care to underserved poor populations. Medicare has recognized these services, and states have added variously to Medicare and private payer financing through Medicaid and more direct granting mechanisms.

Recently, the cost consciousness of managed care purchasers operating in a competitive environment has constrained the income streams that were available from private payers for the support of medical education at the undergraduate and graduate levels. The margins substantially above cost that Medicare, in

particular, has been deliberately providing to nourish the teaching hospital sector are being questioned, and states have begun to review the structure and amount of their support. At the same time patient care, technology, and therefore opportunity for education have moved increasingly out of the hospital to ambulatory sites, many of which are at distant locations. Moreover the necessity to organize care employing multiple professional providers and in a variety of organizational settings which reemphasize primary care, has become increasingly apparent. These changes have been taking place in the context of what many consider an actual current oversupply and an imbalance in the nature of physicians in training and in practice.

In the study which follows, the IOM examines one specialized but critically important aspect of this health education/care enterprise, that is, the way in which Medicare funds, and potentially other sources of revenue, could be more effectively distributed in order to accomplish the most cost-effective use of resources while maintaining the highest quality of contemporaneous medical education and care which properly prepares health professionals for the future. We understand that many other aspects of the undergraduate and graduate health professional educational process will also require further serious and comprehensive consideration.

Medicare's support for institutions engaged in GME began in 1965, and although there have been several attempts, recently, to reform Medicare support of GME (Balanced Budget Act of 1995; HR 2425; HR 4069; and S 1870 of 1996), Congress has indicated an interest in continuing distribution of Medicare funds in some fashion. Depending on balanced budget and Medicare program priorities, the congressional focus on GME funding is likely to continue. In addition, changes in health care organization and delivery suggest the need for reevaluation of the current Medicare GME payment structure, which has become outdated. These changes include a shift of health care to nonacute and ambulatory settings, increased enrollment in managed care, and a new assessment of U.S. health care work force requirements. The current GME system provides little support for training in these settings or in managed care and rewards hospitals for expansion of residencies in the face of a possible national oversupply of physicians.

Following the attempts of Congress to bring about reform and the presidential veto of the Balanced Budget Act of 1995, the Ways and Means Committee of the U.S. House of Representatives requested the IOM to examine GME fund distribution. Specifically, the request was that the IOM assume a trust fund from which GME funding would be distributed and explore the principles and mechanisms that could govern that distribution (see Appendix A).

The IOM assembled a committee representing different perspectives in the fields of medical education, health services delivery, and health policy, research, and economics. The committee met twice during a 2-month period, solicited written comments from a wide array of health organizations, and hosted a public hearing of nine associations involved with GME. The principles that could best support the mission of teaching institutions and the public purposes of funding of

GME were explored. Within the context of those principles, this report offers policymakers guidance for the restructuring of GME support.

The Institute thanks the members of the Committee on Implementing a National Graduate Medical Education Trust Fund for their contributions. The IOM greatly appreciates the financial assistance of The Pew Charitable Trusts, The Commonwealth Fund, and The Burroughs Wellcome Fund, without which this report would not have been possible. For their kind assistance, the Institute thanks Carolyn Asbury, James Bentley, Brian Biles, David Blumenthal, Carmella Bocchino, Enriqueta Bond, Roger Bulger, Robert Dickler, Marvin Dunn, John Eisenberg, Robert Epstein, Linda Fishman, Ruth Hanft, Tim Henderson, Richard Knapp, Stephen Mick, Gordon Moore, James Reuter, Anne Schwartz, Mel Worth, and Don Young. The Institute is grateful to Stuart Guterman, deputy director, Prospective Payment Assessment Commission (ProPAC), for his commissioned work on the history and current status of Medicare support of GME (see Appendix B). The IOM also appreciates the work of the IOM staff: Roger Herdman, study director; Kathleen Nolan, research assistant; Heather Callahan, project assistant; and Clyde Behney, division director.

Kenneth I. Shine, President

Institute of Medicine