To identify time-windowed clinical predictors of in-hospital cardiac arrest (IHCA) and develop a temporally validated, calibrated machine-learning early warning for general wards.
Approach:
Study Design: Retrospective matched case–control study at a tertiary hospital in Taiwan (2019–2021) including 115 IHCA cases and 115 controls matched on age and ward.
Data Collection: Demographics, comorbidities, and laboratory results were extracted from electronic health records. Vital signs and level of consciousness were sampled in 16–24, 8–16, and 1–8-hour windows before the index time.
Model Development: Models were trained with fivefold cross-validation and isotonic probability calibration, tested on a temporally held-out 2021 cohort.
Key Findings:
Independent predictors of IHCA included atrial fibrillation (OR 6.30), heart failure (OR 2.98), end-stage renal disease (OR 2.54), potassium (OR 1.81 per 1 mEq/L), and white blood count (OR 1.06 per 1 k/µL).
MEWS was associated with IHCA up to 16 hours, with lower SpO₂ predominating in the final 8 hours (OR 1.30 per 1% decrease; 95% CI 1.08 to 1.54).
Calibrated XGBoost achieved AUROC curve 0.89 and average precision 0.88; at sensitivity 0.95, specificity was 0.47 in the matched test set.
