Reducing interns' work hours led to fewer attentional failures and serious medical errors in intensive care units

Intervention: Both studies used a crossover design. Interns were randomly assigned to either a traditional or intervention call schedule. The traditional call schedule involved in-house call every 3 night with a limit of 30 hours of consecutive work. The intervention schedule involved call every 4 night with a limit of 16 hours of consecutive work. Interns recorded their sleep and work hours and were followed by physician observers to detect medical errors. In addition, nurse chart reviewers assessed medical records for medical errors and a computer error detection program was utilized. The interns also underwent continuous ambulatory polysomnography at least 3 times per week to assess for periods of sleep and inattentiveness. All data were reviewed by investigators blinded to the interns’ schedules.