TPT 8.14 Index Laparoscopic Common Bile Duct Exploration Versus ERCP-First Strategy for Choledocholithiasis: A Retrospective Comparative Cohort Study with Propensity-Based Analyses

Abstract

ERCP followed by delayed cholecystectomy is a common pathway for choledocholithiasis, but exposes patients to recurrent biliary events and repeat interventions while awaiting definitive surgery. Laparoscopic common bile duct exploration (LCBDE) provides definitive single-stage management, but is variably adopted. We compared the outcomes of single-stage management versus two-stage management of choledocholithiasis. We analysed prospectively maintained database of patients undergoing index ERCP or index LCBDE for choledocholithiasis. Patients with prior cholecystectomy, ERCP for non-stone indications, or a recent ERCP within 6 months were excluded; patients with multiple ERCPs contributed only the index ERCP. Outcomes were further acute biliary admissions, repeat ERCP procedures, additional imaging (ultrasound/MRI/CT), and additional hospital bed-days after the index admission. We used negative binomial regression with log (person-time) offsets; propensity overlap weighting estimated effects in the overlap population (ATO). The cohort comprised 390 patients (ERCP-first 352, LCBDE-first 38) contributing 731.0 person-years of follow-up. Crude readmission rates were 3.9 versus 18.7 per 100 person-years and bed-day rates were 9.1 versus 161.5 per 100 person-years (LCBDE-first vs ERCP-first). In adjusted models, LCBDE-first was associated with fewer readmissions (IRR 0.16, 95% CI 0.05-0.51) and fewer bed-days (IRR 0.05, 95% CI 0.02-0.18). Overlap-weighted analyses were directionally consistent; estimates for repeat ERCP and imaging were less precise. In this observational cohort, the LCBDE-first strategy was associated with materially fewer subsequent admissions and bed-days than an ERCP-first pathway. Propensity-based analyses support these differences in the overlap population, but limited propensity overlap and unmeasured confounding warrant cautious causal interpretation.