Table 1.
Demographic comparison of those who died and those who survived.
Fig 1.
A) Area under the curve (AUC) and B) precision-recall curves comparing XGBoost and TRISS.
Fig 2.
Confusion matrices of XGBoost and TRISS models demonstrating results from testing data.
Fig 3.
Calibration curves comparing XGBoost with TRISS.
Fig 4.
Area under the curve (AUC) of secondary outcomes of interest with corresponding 95% confidence intervals.
PE: pulmonary embolism, ARF: acute respiratory failure, DVT: deep vein thrombosis, AKI: acute kidney injury, ICU: intensive care unit.
Fig 5.
SHAP plot demonstrating the 20 most important features in the XGBoost model.
Features ranked by descending importance. Red points designate higher values for that feature while blue points denote lower values.