ERCP-related perforation: an analysis of operative outcomes in a large series over 12years.

Abstract

Perforation is a rare but serious adverse event of endoscopic retrograde cholangiopancreatography (ERCP). The aim of this study was to determine the predictors of morbidity and mortality after surgical management of ERCP-related perforation (EP). The records of patients with EP requiring surgical intervention at a tertiary referral center in a 12-year period (2004-2016) were retrospectively analyzed for demography, indications for ERCP, risk factors, timing and type of surgical repair, post-operative course, hospital stay, and outcome. Multiple logistic regression was used to identify the parameters predicting survival. Of 25,300 ERCPs, 380 (1.5%) had EP. Non-operative management was successful in 330 (86.8%) patients. 50 (13.2%) patients were operated for EP. Out of 50, the perforation was detected during ERCP (intra-procedure) in 32 patients (64%). In 30 patients (60%), the surgery was performed within 24h of ERCP. Twenty patients underwent delayed surgery (after 24h of ERCP) following the failure of initial non-operative management. The delayed surgery after an unsuccessful medical treatment had a detrimental effect on morbidity, mortality and hospital stay. Post-operative duodenal leak was the only independent predictor of 90-day mortality (p = 0.02, OR = 9.1, 95% CI 1.52-54.64). Addition of T-tube duodenostomy (TTD) to the primary repair for either type I or type II perforations increased post-operative duodenal leak (type I, p = 0.048 and type II; p = 0.001) and mortality (type I, p = 0.009 and type II, p = 0.045). Although EP is a rare event, it has a serious impact on morbidity and mortality. Delaying of surgery following failed non-operative management worsens the prognosis. Addition of TTD to the repair is not helpful in these patients.