Background: In 2002, the Accreditation Council on Graduate Medical Education enacted regulations, effective 1 July 2003, that limited work hours for all residency programs in the United States.
Objective: To determine whether work-hour regulations were associated with changes in mortality in hospitalized patients.
Design: Comparison of mortality rates in high-risk teaching service patients hospitalized before and after July 2003, with nonteaching service patients used as a control group.
Setting: 551 U.S. community hospitals included in the Healthcare Cost and Utilization Project's Nationwide Inpatient Survey between January 2001 and December 2004.
Patients: 1,511,945 adult patients admitted for 20 medical and 15 surgical diagnoses.
Measurement: Inpatient mortality.
Results: In 1,268,738 medical patients examined, the regulations were associated with a 0.25% reduction in the absolute mortality rate (P = 0.043) and a 3.75% reduction in the relative risk for death. In subgroup analyses, particularly large improvements in mortality were observed among patients admitted for infectious diseases (change, -0.66%; P = 0.007) and in medical patients older than 80 years of age (change, -0.71%; P = 0.005). By contrast, in 243 207 surgical patients, regulations were not associated with statistically significant changes (change, 0.13%; P = 0.54).
Limitations: Teaching status was assigned according to hospital characteristics because direct information on each patient's provider was not available. Results reflect changes associated with the sum of regulations, not specifically with caps on work hours.
Conclusions: The work-hour regulations were associated with decreased short-term mortality among high-risk medical patients in teaching hospitals but were not associated with statistically significant changes among surgical patients in teaching hospitals.