Development and external validation of an interpretable machine learning model for predicting prolonged postoperative ICU length of stay in coronary artery bypass grafting patients using MIMIC-IV 3.1 and eICU-CRD 2.0.
An AI model flags patients at risk for prolonged ICU stays after heart surgery with transparent reasoning, supporting better planning.
CatBoost model using 8 bedside features (24h fluid intake, CCI, SOFA, SAPS-II, GCS, vasopressor use, CHF, AF) predicts prolonged post-CABG ICU stay with AUC 0.77 (internal, n=6,919) and 0.65 (external eICU-CRD, n=5,972); SHAP per-patient attribution enables transparent bedside decision support.
What the study was
- Study design
- retrospective_cohort
- Category
- ai_ml_diagnostics
- Maturity
- clinical
- Journal
- BMC Medical Informatics and Decision Making
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.