Wright, Scott M.

ABSTRACT Background The Accreditation Council for Graduate Medical Education (ACGME) has mandated revisions to residents' work hours to improve patient safety and enhance resident education and wellness. The impact on clinical outcomes on a national level is poorly understood. Objective We examined data from before and after the ACGME 2011 duty hour revision and looked for differences between teaching and nonteaching US hospitals. Methods A retrospective observational study of patients admitted to hospitals in the 2-year periods before and after the 2011 duty hour revision was conducted, utilizing a nationally representative data set. We compared patient and hospital characteristics using standardized differences. With nonteaching hospitals serving as the control group, we used multiple group interrupted time series segmented regression analysis to test for postrevision level and trend changes in mortality, length of stay (LOS), and costs. Results We examined more than 117 million hospitalizations. At teaching and nonteaching hospitals, trends in mortality and LOS in prerevision and postrevision periods were not significantly different (all P > .05). A significant monthly reduction in cost per hospitalization was noted postrevision at teaching hospitals (P = .019) but not at nonteaching hospitals (P = .62). In the 2 years following the 2011 revision, there was a monthly reduction in cost per hospitalization (–$52.28; 95% confidence interval –$116.90 to –$12.32; P = .026) at teaching relative to nonteaching hospitals. Conclusions There were no differences in mortality or LOS between teaching and nonteaching hospitals. However, there was a small decrease in cost per hospitalization at teaching hospitals following the 2011 revision.