Validation of a risk factors-based nomogram predicts cholangitis post-ERCP, suggesting enhanced clinical decision-making.
Key Points
This research aims to identify key risk factors for post-ERC cholangitis and develop a predictive model to enhance risk assessment.
Cohort randomly divided into training and validation sets at a 7:3 ratio.
Screened risk factors using univariate analysis, LASSO regression, and multivariate logistic regression.
Developed a nomogram based on independent risk factors identified from analyses.
Assessed model performance using receiver operating characteristic curves and calibration curves.
Evaluated clinical utility through decision curve analysis and clinical impact curve.
Identified significant risk factors including diabetes mellitus, previous ERCP, malignant biliary obstruction, high biliary obstruction, and low albumin levels.
Achieved an area under the curve of 0.903 in the training set and 0.884 in the validation set.
Calibration curve demonstrated strong alignment between predictions and actual outcomes.
No significant deviation observed in predicted vs. actual values per the H–L test results.
Model showed significant net clinical benefit across various risk thresholds.