Withdrawal and extended leave during residency training: Results of a national survey

Background. Although national figures for medical student withdrawal and extended leave have long been reported, similar data have not been available for residents in training. Method. Data for this study came from the American Medical Association survey of the 1991-92 residency year, in which program directors were asked for information about residents who had taken extended leave or had withdrawn or been dismissed from their programs prior to completion. Data are reported for 89,368 residents enrolled in 6,302 programs (89.2% of all surveyed programs). Results. During the 1991-92 year, 2,449 residents (2.7%) withdrew or were dismissed from their programs and 887 (1.0%) took extended leave. Specialty and program changes accounted for 56% of the withdrawals, while performance difficulties were implicated in 12.9%. Maternity or paternity leave was involved in 32.2% of extended leaves, followed by research sabbaticals (11.4%) and physical problems (10.5%). Women had higher rates of both withdrawal and extended leave than men. Withdrawal for performance difficulties was lowest among graduates of U.S. and Canadian allopathic schools as compared with graduates of osteopathic or foreign medical schools, and lowest among Caucasians as compared with those of other racial - ethnic identities. Conclusion. Although overall figures and percentages are low, there are small but persistent losses of residents annually that vary by specialty, gender, race-ethnicity, and education.