Direct graduate medical education payments to teaching hospitals by medicare: Unexplained variation and public policy contradictions

Purpose. To comprehensively examine both inter- and intrastate variations i n Medicare's cost-rate structure for teaching hospitals and to assess the M edicare payment system for graduate medical education (GME), as presently c onfigured, as an instrument to promote physician workforce reform, specific ally sufficient public access to primary care physician services. Method. Using Public Use Files of hospital cost reports from the Health Car e Financing Administration for fiscal year 1997, 648 hospitals that met inc lusion criteria for moderate GME volume were identified. The average and ra nge of direct costs of resident training were computed for these teaching h ospitals to illustrate differences within and between the 45 states that ha d at least two teaching hospitals that qualified for comparison. The cost r ate upon which direct medical education (DME) payments are based was then c orrelated with the percentage of a state's counties that were wholly design ated primary care health personnel shortage areas (PCHPSAs) in 1997 and wit h its primary care physician to population ratio, as determined from the Ar ea Resource File. Results. Variations in inter and intrastate DME costs exist. In some states , the range in DME rates substantially exceeded the mean cost. DME funding policies are more generous toward teaching hospitals in states with greater primary care physician tu-population ratios and smaller proportions of cou nties wholly designated PCHPSAs. Conclusion. Inherent inequities in DME funding seriously undermine the curr ent Medicare GME payment system's capacity to contribute to U.S. physician workforce reform and to improve access to care. There is actually a financi al incentive to train residents in areas in which there is relatively less need for their services.