Abstract
INTRODUCTION: Early cholecystectomy shortly after admission for mild gallstone pancreatitis has been proposed based on observational data. We hypothesized that cholecystectomy within 24 hours of admission versus after clinical resolution of gallstone pancreatitis that is predicted to be mild results in decreased length-of-stay (LOS) without an increase in complications. METHODS: Adults with predicted mild gallstone pancreatitis were randomized to cholecystectomy with cholangiogram within 24 hours of presentation (early group) versus after clinical resolution (control) based on abdominal exam and normalized laboratory values. Primary outcome was 30-day LOS including readmissions. Secondary outcomes were time to surgery, endoscopic retrograde cholangiopancreatography (ERCP) rates, and postoperative complications. Frequentist and Bayesian intention-to-treat analyses were performed. RESULTS: Baseline characteristics were similar in the early (n = 49) and control (n = 48) groups. Early group had fewer ERCPs (15% vs 29%, P = 0.038), faster time to surgery (16 h vs 43 h, P < 0.005), and shorter 30-day LOS (50 h vs 77 h, RR 0.68 95% CI 0.65 - 0.71, P < 0.005). Complication rates were 6% in early group versus 2% in controls (P = 0.613), which included recurrence/progression of pancreatitis (2 early, 1 control) and a cystic duct stump leak (early). On Bayesian analysis, early cholecystectomy has a 99% probability of reducing 30-day LOS, 93% probability of decreasing ERCP use, and 72% probability of increasing complications. CONCLUSION: In patients with predicted mild gallstone pancreatitis, cholecystectomy within 24 hours of admission reduced rate of ERCPs, time to surgery, and 30-day length-of-stay. Minor complications may be increased with early cholecystectomy. Identification of patients with predicted mild gallstone pancreatitis in whom early cholecystectomy is safe warrants further investigation.