Predicting recurrence of biliary colic after conservative management in the emergency department: a prospective cohort study with a parametric Weibull survival model

Background Biliary colic is usually managed conservatively in the emergency department (ED), yet recurrence is frequent and its timing is poorly characterized, particularly in Latin American settings with limited access to elective cholecystectomy. We aimed to develop and internally validate a survival-based model predicting recurrence of biliary colic after conservative ED management. Methods Single-center prospective cohort of adults with uncomplicated biliary colic and ultrasound-confirmed gallstones discharged from the ED of a tertiary hospital in Cuenca, Ecuador, followed for up to 365 days. The primary outcome was symptomatic recurrence requiring an ED visit; cholecystectomy and biliary complications before recurrence were censored. Predictors were screened with Cox regression; multivariable Cox and parametric Weibull models were compared by Harrell's C-index and the Akaike information criterion. Internal validation used 1000 bootstrap resamples; calibration was assessed at 7, 15, 30 and 60 days overall and by predicted-risk group. A sensitivity analysis used Fine–Gray competing-risks regression. Results Of 507 patients (median age 38 years; 76.5% women), 223 (44.0%) recurred; 25% of recurrences occurred within 7 days of discharge and recurrence-free survival was 67.8% at 30 days and 43.1% at 365 days. Biliary complications developed in 10.7% and only 7.3% underwent cholecystectomy before recurrence. In the final Weibull model, a positive Murphy's sign (hazard ratio [HR] 7.09; 95% CI 4.73–10.61) and diabetes mellitus (HR 2.18; 95% CI 1.33–3.58) were independent predictors. The shape parameter (γ = 0.419; 95% CI 0.376–0.468) indicated a monotonically decreasing baseline hazard. Discrimination was acceptable (C-index 0.692; bootstrap-corrected 0.692, 95% CI 0.665–0.723) with good calibration from 30 days onwards and underestimation of very early risk in the highest-risk group. Fine–Gray estimates were consistent (subdistribution HR 5.73 and 2.06, respectively). Conclusions Recurrence after conservative management of biliary colic is frequent and concentrated in the first days after discharge. Two bedside variables, Murphy's sign and diabetes mellitus, identify patients at high risk of early recurrence who may benefit from prioritized surgical evaluation and structured early follow-up. External validation in multicenter cohorts is required. Trial registration: Not applicable (observational study).