Clinical predictive modeling for post-ERCP cholangitis in biliary stricture patients.

Aim: The objective of this study was to predict the occurrence of post-endoscopic retrograde cholangiopancreatography (ERCP) cholangitis (PEC) in patients with biliary stricture. Method: We collected clinical data and procedural records from 1,606 patients who underwent ERCP for biliary stricture over the last 10 years. Of these, 1,281 patients were randomly allocated to the training or validation group and 325 patients enrolled in other hospitals as an external validation set. LASSO logistic regression analysis was used to identify independent risk factors and establish a predictive model for PEC. The performance of the nomogram was evaluated using receiver operating characteristic (ROC) curves and calibration curves. At the same time, decision curve analysis (DCA) is used to determine the net benefit threshold of prediction. Results: The findings indicated that age, albumin levels, etiology of biliary stricture, stenosis site, diabetes, digestive tract reconstruction, obstruction length and the use of cholangioscopy may be factors that influence postoperative biliary tract infection in patients with biliary stenosis. The nomogram exhibited a higher net benefit in decision curve analysis. The calibration curve shows a strong agreement between expected and observed results. Conclusion: The proposed nomogram exhibits robust predictive performance. This tool can assist healthcare professionals in minimizing the risk of PEC in patients with biliary stricture undergoing ERCP.