Endoscopic Retrograde Cholangiopancreatography Followed by Laparoscopic Cholecystectomy Versus Laparoscopic Common Bile Duct Exploration with Cholecystectomy: Comparative Outcomes and Predictive Models for Choledocholithiasis

Anotacija

Background. Choledocholithiasis can be treated by a two-stage approach – endoscopic retrograde cholangiopancreatography followed by laparoscopic cholecystectomy (ERCP+LC) – or a single-stage approach using laparoscopic common bile duct exploration with cholecystectomy (LCBDE+LC). Comparative evidence on perioperative outcomes and predictive modeling remains limited. Methods. This prospective observational study included 200 adults with gallbladder and CBD stones managed at a tertiary center (100 ERCP+LC, 100 LCBDE+LC). Standardized protocols guided preoperative assessment, interventions, and discharge. Primary outcome was successful CBD stone clearance with completion of cholecystectomy. Secondary outcomes included operative time, hospital stay, pain scores (VAS at 24 h and 3 days), complications, retained stones, and satisfaction. Logistic regression models predicted retained stones and complications; performance was assessed by ROC-AUC, Brier score, and calibration. Nomograms and an integrated risk score table were developed. Results. LCBDE patients had larger CBD diameters (13.18±2.01 mm vs. 10.94±2.46 mm) and stones (6.29±1.47 mm vs. 5.65±1.69 mm). Operative time and hospital stay were similar. Pain at 24 h was lower in LCBDE (median 5 [IQR 5–6]) vs. ERCP+LC (6 [6–7]; p < 0.001), equalizing by day 3. Postoperative direct bilirubin was lower after ERCP+LC (p = 0.036). Complications were infrequent; retained stones occurred in 1% vs. 3%. Predictive models showed AUC 0.92 for retained stones and 0.73 for complications. Nomograms and a points table enable bedside risk stratification. Conclusions. Both strategies are safe and effective. LCBDE offers lower early pain and suits high stone burden; ERCP+LC remains appropriate for urgent decompression or limited laparoscopic resources. Risk-based tools may support individualized decision-making.