Development and validation of machine learning models to predict 30-day mortality in patients with cardiac arrest complicated by acute kidney injury

Abstract

Highlights • LASSO-LR predicts 30-day mortality in CA-AKI (AUROC = 0.751, Brier = 0.209).• LASSO-LR outperforms APACHE II (0.583) and SOFA (0.660).• SHAP and nomogram reveal oliguria and RDW as top predictors.