Participation of international medical graduates in graduate medical education and hospital care for the poor

Objective. —To determine the impact of limiting international medical graduate (IMG) participation in US graduate medical education (GME) on the delivery of hospital care to the poor. Methods. —To ascertain the pattern of IMG participation in GME and the degree to which the principal teaching hospitals with programs with large IMG enrollments provide care to the poor, we used data from the American Medical Association 1993 Annual Survey of Graduate Medical Education Programs and Teaching Institutions to analyze the pattern of IMG participation in GME in the six core specialties of internal medicine, family practice, obstetrics and gynecology, surgery, pediatrics, and psychiatry. Main Outcome Measures. —Programs were identified as IMG dependent if at least 50% of the resident physicians enrolled in the first year of the program were IMGs. All programs were linked to their principal teaching hospitals, and hospitals were assessed according to the number of programs based at each institution, the number of IMG-dependent programs at the institution, and whether no-pay patients and/or Medicaid/public assistance beneficiaries constituted more than 20% of the patients served. Results. —Of the 20170 first-year resident physicians in the six core specialties, 31.8% were IMGs. The proportion of programs dependent on IMG enrollment was 27.7%, ranging from 5.2% in obstetrics and gynecology programs to 49.5% in psychiatry programs. About 72% of all first-year IMGs were in IMG-dependent programs. Of the 688 hospitals serving as principal teaching sites for programs in at least one of the six core specialties, 106 were categorized as dependent on IMG programs, but only 77 of those provided a disproportionate amount of care to the poor. Finally, 40% of IMG-dependent GME programs and 36% of first-year IMG residents were based in hospitals that did not provide a disproportionate amount of care to the poor. Conclusions. —Based on this analysis, 77 hospitals can arguably be considered dependent on IMG resident physicians to provide care to the poor. Moreover, a large number of IMG residents and IMG-dependent programs are in hospitals that do not provide a disproportionate amount of care to the poor. These findings show the scale of the problem policymakers must address if they choose to limit IMG access to GME while maintaining access of the poor to needed hospital care. (JAMA. 1995;274:696-699)