continued innovation requires a sound underpinning of basic and clinical research. It is important that the level of payments for graduate medical education recognize all the costs involved in graduate training, including residents' stipends and benefits, salaries and benefits related to faculty teaching, and overhead costs related to the graduate education process to preserve the most creative and innovative health care and health research systems in the world.
Graduate medical education is not the only factor that accounts for the higher costs at AHCs. Health care professionals at AHCs often treat patients who are more seriously ill, provide infrastructure for biomedical and health services research, and offer clinical training experiences for undergraduate medical and other health professional students—costs that add to an institution's cost base. In recognition of these additional costs, the Medicare program has historically paid teaching hospitals an indirect medical education adjustment. Likewise, some modified version of an adjustment to account for those costs unique to AHCs, above and beyond the direct costs of graduate medical education, will be necessary in a managed competition environment. The adjustment should be formula driven rather than based on year-to-year appropriations and, again, should be funded by all payers.
Finally, all accountable health plans, when feasible, should include an academic health center as a centerpiece of their networks. Such a requirement would guarantee all citizens access to the latest in state-of-the-art care when medically necessary. At the same time, it would help to ensure an adequate flow to academic health centers of the patients on whom their educational and human research programs are completely dependent. Without access to suitable patient populations for clinical studies, the United States will lose its leadership in opening the frontiers of innovative and effective health care. This approach might be further complemented if regional ''centers of excellence" were established in AHCs for the evaluation of emerging technologies and specialized methods of treatment. AHCs would be the natural place for locating such centers, given their strong emphasis on research, particularly if they have a health services research component. The regionalization of certain emerging technologies would also help to control the widespread diffusion of new technologies absent sufficient outcomes research that could be used to judge their cost-effectiveness.