The role of ROSE-enhanced ERCP-guided biopsy in diagnosing biliary stricture

To evaluate the diagnostic efficacy of rapid on-site evaluation (ROSE)-enhanced endoscopic retrograde cholangiopancreatography (ERCP)-guided biopsy for biliary strictures, and develop a predictive model for diagnostic accuracy. This retrospective study included 200 patients undergoing ERCP-guided biopsy (ROSE cohort, n = 100; NON-ROSE cohort, n = 100). Diagnostic performance metrics (accuracy, sensitivity, specificity, positive predictive value [PPV], negative predictive value [NPV]), specimen adequacy, biopsy attempts, procedure duration, and complications were compared between groups. Logistic regression identified independent predictors of diagnostic accuracy. A prediction model was developed and internally validated through receiver operating characteristic (ROC) curve analysis, calibration curves, decision curve analysis (DCA), and constructed as a nomogram. The ROSE cohort demonstrated significantly higher accuracy (92.0% vs. 78.0%), sensitivity (97.2% vs. 74.7%), and NPV (91.6% vs. 48.7%) (all P < 0.05). ROSE required fewer biopsy passes (2.43 vs. 3.61; P < 0.001), shorter sampling time (2.71 vs. 4.33 min; P < 0.001), without increased complications (P > 0.05). Multivariate analysis confirmed ROSE as an independent predictor (OR≈4.0, P = 0.032). The validated predictive model showed good discrimination (AUC≈0.80) and clinical utility. ROSE significantly enhances diagnostic yield and procedural efficiency in ERCP-guided biopsy for biliary strictures. The developed nomogram provides an evidence-based tool for individualized prediction, supporting clinical decision-making.