ABSTRACT: Lack of sleep can impair performance in diverse situations. To prevent medical errors, work-hour restrictions on surgeons in training were imposed and are now being considered for practicing surgeons. This retrospective multicenter cohort study was undertaken to determine the association of operating the night before performing an elective cholecystectomy with complication rates when performed by community-based surgeons in 102 hospitals in Ontario.
Data from the Institute for Clinical Evaluative Sciences were used to identify all elective daytime laparoscopic cholecystectomies performed by a surgeon on a nonholiday weekday between 2004 and 2011. Elective laparoscopic cholecystectomy was defined as having started between 7 AM and 6 PM. At-risk surgeries were those performed by community-based surgeons who worked the night before. Each at-risk surgery was matched with 4 other laparoscopic cholecystectomies performed by the same surgeon in the same year when that surgeon had not worked the previous night. The primary outcome was conversion from laparoscopic to open cholecystectomy. Secondary outcomes were death or evidence of an iatrogenic injury within 30 days postoperatively.
Of 94,183 elective daytime cholecystectomies, 2078 were performed by a surgeon who had operated the night before. These at-risk procedures were randomly matched to 8312 procedures, resulting in a total sample of 10,390 procedures performed by 331 community-based surgeons. A given surgeon performed an average of 6 daytime laparoscopic cholecystectomies after operating the night before. Patient and hospital characteristics for the 2 groups of cholecystectomies were well balanced. Forty-six of 2031 procedures (2.2%) performed by a surgeon who operated the night before required conversion to an open procedure. Open conversion occurred in 157 (1.9%) of 8124 cases performed by a surgeon who had not operated the night before. These rates were not statistically different with an odds ratio of undergoing conversion to open cholecystectomy being 1.18 (95% confidence interval [CI], 0.85–1.64) when comparing these 2 groups. Fourteen (0.7%; 95% CI, 0.3%–1.0%) of 2031 procedures performed by a surgeon who operated the night before were associated with iatrogenic injuries, which also occurred in 72 (0.9%; 95% CI, 0.7%–1.1%) of 8124 cases performed by a surgeon who had not operated the night before; these rates were not statistically different. The odds of experiencing an iatrogenic injury were 0.77 (95% CI, 0.43%–1.37; P = 0.37) for surgeons operating when they had performed procedures the night before compared with procedures when they had not operated the night before. Fewer than 5 patients (0.2%) died when operated on by surgeons who had operated the night before the procedure. Seven of 8124 patients (0.1%; 95% CI, 0.0%–0.2%) who died were operated on by surgeons who had not operated the night before.
No significant association was found, indicating that operating the night before was associated with conversion to open cholecystectomy, risk of iatrogenic complications, or death for elective laparoscopic surgery performed the next day. Policies limiting attending surgeon work hours are controversial. Restructuring health care delivery to prevent surgeons operating during the day after they operated the previous night could have important cost, staffing, and resource implications.
Division of General Surgery, Department of Surgery, Western University; Institute for Clinical Evaluative Sciences; and Department of Epidemiology and Biostatistics, Western University, London; and Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada