Predicting ICU Transfer and Short-term Mortality in Emergency Department Atrial Fibrillation Patients: An Enhanced Machine Learning Model Using MIMIC Data
Preprint
Abstract
Atrial fibrillation (AF) is a prevalent condition in emergency department (ED) patients and is associated with an elevated risk of intensive care unit (ICU) transfer and short-term mortality. Early identification of high-risk patients is critical for timely intervention and improved clinical outcomes. We constructed a combined cohort from the MIMIC-IV-ED and MIMIC-IV databases, comprising ED admissions of patients with AF, and developed an interpretable machine learning (ML) framework to predict ICU transfer and mortality within 3, 7, and 30 days using clinical variables obtained at triage. The preprocessing pipeline included imputation of missing data, z-score normalization of numerical features, one-hot encoding of categorical variables, and correction for class imbalance using the Synthetic Minority Over-sampling Technique (SMOTE). A hybrid feature selection strategy combining Recursive Feature Elimination with Cross-Validation (RFECV) and Least Absolute Shrinkage and Selection Operator (LASSO) regression reduced the initial set to 19 clinically relevant predictors. Among six evaluated machine learning algorithms, LightGBM demonstrated the highest performance for ICU transfer (AUROC = 0.7979, 95\% confidence interval (CI): 0.7916–0.8041) and for 7- and 30-day mortality (AUROC = 0.8316, 95\% CI: 0.8156–0.8476; AUROC = 0.8010, 95\% CI: 0.7898–0.8123), while CatBoost achieved the best performance for 3-day mortality (AUROC = 0.8444, 95\% CI: 0.8237–0.8644). SHAP(SHapley Additive exPlanations) analysis identified O\textsubscript{2}sat, acuity, and resprate as key determinants, underscoring the clinical plausibility and interpretability of the models. These findings highlight the potential of interpretable machine learning approaches to enable early, time-sensitive risk stratification and support informed clinical decision-making for AF patients in the ED.Keywords
This publication has 25 references indexed in Scilit:
- The Scoring Model to Predict ICU Stay and Mortality After Emergency Admissions in Atrial Fibrillation: A Retrospective Study of 30 366 PatientsClinical Cardiology, 2025
- Prediction of mortality risk in patients with severe community-acquired pneumonia in the intensive care unit using machine learningScientific Reports, 2025
- Association between inflammation markers and all-cause mortality in critical ill patients with atrial fibrillation: Analysis of the Multi-Parameter Intelligent Monitoring in Intensive Care (MIMIC-IV) databaseIJC Heart & Vasculature, 2024
- Global Aging and our FuturesWorld Futures, 2023
- MIMIC-IV, a freely accessible electronic health record datasetScientific Data, 2023
- Early Rhythm-Control Therapy in Patients with Atrial FibrillationNew England Journal of Medicine, 2020
- Scalable and accurate deep learning with electronic health recordsnpj Digital Medicine, 2018
- Validation of the Risk Estimator Decision Aid for Atrial Fibrillation (RED-AF) for Predicting 30-Day Adverse Events in Emergency Department Patients With Atrial FibrillationAnnals of Emergency Medicine, 2015
- A Population-Based Description of Atrial Fibrillation in the Emergency Department, 2002 to 2010Annals of Emergency Medicine, 2013
- Independent Risk Factors for Atrial Fibrillation in a Population-Based CohortJAMA, 1994