Additional file 2 of Postoperative delirium prediction using machine learning models and preoperative electronic health record data

Additional file 2 Sensitivity Analyses. 1. Sensitivity Analysis in Patients Age 65 and Over. Supplementary Figure S1.1.: Inclusion Flow Diagram. Supplementary Table S1.1: Baseline Demographics. Supplementary Figure S1.2: Receiver Operating Characteristic (ROC) Curve For 5 Models. Supplementary Table S1.2: Comparison of Model Characteristics. Supplementary Figure S1.3: Feature Importance Summary of XGBoost Model. Supplementary Table S1.3: Comparison of Most Important Variables Chosen by XGBoost And Neural Network. Supplementary Table S1.4: Multivariable Logistic Regression Using Variables Selected by Expert Clinicians. Supplementary Table S1.5: Multivariable Logistic Regression Using Variables Chosen by The XGBoost Algorithm. 2. Sensitivity Analysis with Neurosurgery Patients Excluded. Supplementary Figure S2.1.: Inclusion Flow Diagram. Supplementary Table S2.1: Baseline Demographics. Supplementary Figure S2.2: Receiver Operating Characteristic (ROC) Curve For 5 Models. Supplementary Table S2.2: Confidence Intervals Of AUC-ROC. Supplementary Figure S2.3: Feature Importance Summary of XGBoost Model. Supplementary Table S2.3: Comparison of Most Important Variables Chosen by XGBoost And Neural Network. Supplementary Table S2.4: Multivariable Logistic Regression Using Variables Selected by Expert Clinicians. Supplementary Table S2.5: Multivariable Logistic Regression Using Variables Chosen by The XGBoost Algorithm.