How will graduate medical education reform affect specialties and geographic areas?

To project specialty and geographic impacts of workforce reform proposals on the practice output of graduate medical education (GME).

A demographic life-table model to predict GME output was developed using 1987 cohort data from the Association of American Medical Colleges Annual GME Census. The 1992 GME cohort was used as a baseline to compare the simulated impact of alternate specialty and regional policies.

Allopathic and osteopathic GME programs in the United States.

Projected number of physicians (MDs and DOs) entering nine categories of practice specialty at the conclusion of GME.

If GME input is reduced to 110% of US medical graduates with 55% entering practice as generalists (including obstetrics and gynecology), then the total number of first-year positions will decline from 24,433 to 18,783, and the total number of residents in GME would decline from 103,858 to 80,699 at equilibrium. Even with a 110% restriction on GME input, the overall physician-to-population ratio will continue to grow, albeit at a much slower rate. The number of generalists leaving GME annually would increase by 742 (9%) and the number of specialists would decline by 6517 (44%). At the regional level, allocating GME positions by prorating to the current distribution results in less change than would prorating positions to regional populations.

Achieving national goals of reduced aggregate physician production, reduced specialist supply, and generalist increases will require significant alterations in the GME pool. Adequate time and funding for resident substitution will be required for hospitals to develop alternate models of providing service to allow national workforce goals to be met.