What predicts failed cannulation and therapy at ERCP? Results of a large-scale multicenter analysis

Study background and aims: Predicting outcome at endoscopic retrograde cholangiopancreatography (ERCP) remains difficult. Our aim was to identify the risk factors for failed ERCP. Patients and methods: A prospective multicenter study of ERCP was performed in 66 hospitals across England. Data on 4561 patients were collected using a structured questionnaire completed at the time of ERCP. Results: In total 3209 patients had not had an ERCP prior to the study period. Considering their first ever ERCP, 2683 (84 %) were successfully cannulated, 2241(70 %) had all intended therapy completed, 360 (11 %) had some intended therapy completed, and 608 (19 %) were considered to have had a failed procedure. For first ever ERCP, factors associated with incomplete procedure (odds ratio and 95 % confidence interval) were: Billroth surgery (9.2, 3.2 – 26.7), precutting (2.0, 1.6 – 2.7), common bile duct (CBD) stone size and number (3.2, 2.1 – 4.8 for multiple, large stones), interventions in the pancreatic duct (3.4, 1.6 – 7.0), and CBD stenting (2.8, 2.2 – 3.5). Analysis of the 1352 patients who had undergone an ERCP prior to the study period indicated previous failed ERCP was also predictive of incomplete therapy (1.5, 1.1 – 2.1). The modified Schutz score correlated with ERCP completion, as did the Morriston score, even when modified to include only variables measurable before the procedure. Conclusion: This study confirms that patient- and procedure-based variables are key predictors of technical success and validates current methods of rating ERCP difficulty. Of note, a correlation between outcome and institutional factors, such as unit and endoscopist caseload, was not demonstrated.