Cardiac arrest & resuscitation
Prediction of neurologic outcome after out-of-hospital cardiac arrest: An interpretable approach with machine learning
Authors' conclusion
The XGBoost machine learning model with 10 features available at the time of hospital admission showed good performance for predicting neurologic outcome after OHCA, with no apparent signs of overfitting.
Study context
- Population
- Out-of-hospital Cardiac Arrest (OHCA) cases in the Swedish Cardiopulmonary Resuscitation Registry, 2010–2020
- Participants
- 55,615 OHCA cases
- Outcome
- Prediction of neurologic outcome after OHCA
Methodological detailsregistry-based, machine learning (XGBoost), stratified repeated 10-fold cross-validation, held-out test set
- Registry-based
- Machine learning (XGBoost)
- Stratified repeated 10-fold cross-validation
- Held-out test set
Performance
As reported in the abstract
Prediction of neurologic outcome after OHCA (test set)
| Test or model | AUC | Sensitivity | Specificity | PPV | NPV | Macro F1 |
|---|---|---|---|---|---|---|
| Final model | 0.964 (95% CI 0.960-0.968) | 0.606 (95% CI 0.573-0.634) | 0.975 (95% CI 0.972-0.978) | 0.664 (95% CI 0.625-0.696) | 0.969 (95% CI 0.966-0.972) | 0.803 (95% CI 0.788-0.816) |
Final model
AUC0.964 (95% CI 0.960-0.968)
Sensitivity0.606 (95% CI 0.573-0.634)
Specificity0.975 (95% CI 0.972-0.978)
PPV0.664 (95% CI 0.625-0.696)
NPV0.969 (95% CI 0.966-0.972)
Macro F10.803 (95% CI 0.788-0.816)
Further findings
Features with highest SHAP impact
- ROSC on arrival to hospital, initial rhythm asystole, conscious on arrival to hospital
Original publication
Prediction of neurologic outcome after out-of-hospital cardiac arrest: An interpretable approach with machine learning. Resuscitation. Published 12 August 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →