Xiaodong Zhang, Jieyu Zhang, Lu Wang
AIM: To identify predictors of post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) and to develop a multivariate prediction model by comprehensively evaluating clinical characteristics, procedure-related factors, and inflammatory biomarkers, including C-reactive protein (CRP)-to-lymphocyte ratio (CLR).METHODS: A total of 335 patients who underwent endoscopic retrograde cholangiopancreatography (ERCP) at Zhejiang Jinhua Guangfu Cancer Hospital from December 2022 to December 2025 were retrospectively analyzed. Patients were divided into the PEP group (n = 52) and the non-PEP group (n = 283) according to whether PEP occurred. Demographic characteristics, comorbidities, ERCP-related procedural factors, and preoperative laboratory indices were collected and evaluated. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of PEP. A nomogram was subsequently developed based on the identified independent predictors. The discriminative ability, calibration, and clinical utility of the model were evaluated using the receiver operating characteristic (ROC) curve, Hosmer–Lemeshow test, and decision curve analysis, respectively.RESULTS: Compared with the non-PEP group, patients who developed PEP exhibited significantly higher CLR levels and a higher prevalence of periampullary diverticulum, pancreatic duct cannulation, and pancreatic duct guidewire passage ≥2 times (all p < 0.05). Multivariate logistic regression analysis revealed that a history of acute pancreatitis (odds ratio [OR] = 3.43, 95% confidence interval [CI]: 1.32–8.93), periampullary diverticulum (OR = 4.10, 95% CI: 2.00–8.39), pancreatic duct cannulation (OR = 3.11, 95% CI: 1.46–6.64), and CLR (OR = 1.11, 95% CI: 1.04–1.19) were independent predictors of PEP. The nomogram incorporating these independent predictors demonstrated satisfactory discriminative ability (area under the curve [AUC] = 0.80, 95% CI: 0.73–0.87). The Hosmer–Lemeshow test indicated good model calibration (p = 0.239), and the decision curve analysis showed that the model provided a net clinical benefit within the threshold probability range of 0.2–0.7.CONCLUSIONS: The multivariate nomogram incorporating CLR and other independent predictors demonstrated good discriminative ability and calibration, and may serve as a practical tool for facilitating individualized risk assessment and guiding preventive strategies in patients undergoing ERCP.