Risk Factors for Post-ERCP Pancreatitis in High- and Low-Volume Centers and Among Expert and Non-Expert Operators: A Prospective Multicenter Study

OBJECTIVES: Prospective studies have identified a number of patient- and procedure-related independent risk factors for post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis, with different conclusions, so various questions are still open. The endoscopist's expertise, case volume, and case mix can all significantly influence the outcome of ERCP procedures, but have been investigated little to date. METHODS: We identified patient-and procedure-related risk factors for post-ERCP pancreatitis and the impact of the endoscopist's experience and the center's case volume, using univariate and multivariate analysis, in a multicenter, prospective study involving low-and high-volume centers, over a 6-month period. RESULTS: A total of 3,635 ERCP procedures were included; 2,838 (78%) ERCPs were performed in the 11 high-volume centers (median 257 each) and 797 in the 10 low-volume centers (median 45 each). Overall, 3,331 ERCPs were carried out by expert operators and 304 by less-skilled operators. There were significantly more grade 3 difficulty procedures in high-volume centers than in low-volume ones (P < 0.0001). Post-ERCP pancreatitis occurred in 137 patients (3.8%); the rates did not differ between high-and low-volume centers (3.9% vs. 3.1%) and expert and non-expert operators (3.8% vs. 5.5%). However, in high-volume centers, there were 25% more patients with patient-and procedure-related risk factors, and the pancreatitis rate was one-third higher among non-expert operators. Univariate analysis found a significant association with pancreatitis for history of acute pancreatitis, either non-ERCP-or ERCP-related and recurrent, young age, absence of bile duct stones, and biliary pain among patient-related risk factors, and >10 attempts to cannulate the Vater's papilla, pancreatic duct cannulation, contrast injection of the pancreatic ductal system, pre-cut technique, and pancreatic sphincterotomy, among procedure-related risk factors. Multivariate analysis also showed that a history of post-ERCP pancreatitis, biliary pain, >10 attempts to cannulate the Vater's papilla, main pancreatic duct cannulation, and pre-cut technique were significantly associated with the complication. CONCLUSIONS: A history of pancreatitis among patient-related factors, and multiple attempts at cannulation among procedure-related factors, were associated with the highest rates of post-ERCP pancreatitis. Pre-cut sphincterotomy, although identified as another significant risk factor, appeared safer when done early (fewer than 10 attempts at cannulating), compared with repeated multiple cannulation. The risk of post-ERCP pancreatitis was not associated with the case volume of either the single endoscopist or the center; however, high-volume centers treated a larger proportion of patients at high risk of pancreatitis and did a significantly greater number of difficult procedures.