Selective vs Routine Cholangiography Across a Health Care Enterprise.

Importance: There is sparse literature on whether routine cholangiography (RC) vs selective cholangiography (SC) during cholecystectomy is associated with improved perioperative outcomes, regardless of whether an intraoperative cholangiogram (IOC) is performed. Objective: To compare perioperative outcomes of cholecystectomy between surgeons who routinely vs selectively perform IOC. Design, Setting, and Participants: This retrospective cohort study was conducted from January 2015 through June 2023 and took place within the Cleveland Clinic Enterprise, which includes 18 hospitals and 9 ambulatory surgery centers in 2 states (Ohio and Florida). Participants included adult patients who underwent cholecystectomy for benign biliary disease. Data analysis was conducted between July 2023 and August 2024. Exposure: Routine cholangiography, defined as more than 70% of cholecystectomies performed with IOC per surgeon over the study period. Main Outcome(s) and Measure(s): The primary outcome was major bile duct injury (BDI). Hierarchical mixed-effects models with patients nested in hospitals adjusted for individual- and surgeon-level characteristics were used to assess the odds of major BDI and secondary outcomes (minor BDI, operative duration, and perioperative endoscopic retrograde cholangiopancreatography [ERCP]). Results: A total of 134 surgeons performed 28 212 cholecystectomies with 10 244 in the RC cohort (mean age, 52.71 [SD, 17.78] years; 7102 female participants [69.33%]) and 17 968 in the SC cohort (mean age, 52.33 [SD, 17.72] years; 12 135 female participants [67.54%]). Overall, 26 major BDIs (0.09%) and 105 minor BDIs (0.34%) were identified. Controlling for patient and surgeon characteristics nested in hospitals, RC was associated with decreased odds of major BDI (odds ratio [OR], 0.16; 95% CI, 0.15-0.18) and minor BDI (OR, 0.83; 95% CI, 0.77-0.89) compared with SC. Major BDIs were recognized intraoperatively more often in the RC cohort than the SC cohort (76.9% vs 23.0%; difference, 53.8%; 95% CI, 15.9%-80.2%). Lastly, RC was not significantly associated with increased perioperative ERCP utilization (OR, 1.01; 95% CI, 0.90-1.14) or negative ERCP rate (RC, 27 of 844 [3.2%] vs SC, 57 of 1570 [3.6%]; difference, -0.3%; 95% CI, -1.9% to 1.0%). Conclusions and Relevance: In this study, RC was associated with decreased odds of major and minor BDI, as well as increased intraoperative recognition of major BDI when it occurred. RC could be considered as a health systems strategy to minimize BDI, acknowledging the overall low prevalence but high morbidity from these injuries.