Pre- and post-procedure risk prediction models for post-endoscopic retrograde cholangiopancreatography pancreatitis

Background Post-endoscopic retrograde cholangiopancreatography pancreatitis (PEP) is the most common serious adverse event. Given recent endoscopic advances, we aimed to develop and validate a risk prediction model for PEP using the latest clinical database. Methods We analyzed the data of patients with naive papilla who underwent endoscopic retrograde cholangiopancreatography (ERCP). Pre-ERCP and post-ERCP risk prediction models for PEP were developed using logistic regression analysis. Patients were classified into low- (0 points), intermediate- (1-2 points), and high-risk (>= 3 points) groups according to point scores. Results We included 760 and 735 patients in the derivation and validation cohorts, respectively. The incidence of PEP was 5.5% in the derivation cohort and 3.9% in the validation cohort. Age <= 65 (1 point), female sex (1 point), acute pancreatitis history (2 points), malignant biliary obstruction (2 points [pre-ERCP model] or 1 point [post-ERCP model]), and pancreatic sphincterotomy (2 points, post-ERCP model only) were independent risk factors. In the validation cohort, the high-risk group (>= 3 points) had a significantly higher risk of PEP compared to the low- or intermediate-risk groups under the post-ERCP risk prediction model (low: 2.0%; intermediate: 3.4%; high: 18.4%), while there was no significant between-group difference under the pre-ERCP risk prediction model (low: 2.2%; intermediate: 3.8%; high: 6.9%). Conclusions We developed and validated pre-ERCP and post-ERCP risk prediction models. In the latter, the high-risk group had a higher risk of PEP development than the low- or intermediate-risk groups. Our study findings will help clinicians stratify patient risk for the development of PEP.