Machine learning for predicting in-hospital mortality in elderly patients with heart failure combined with hypertension: a multicenter retrospective study
Authors' conclusion
This study developed eight machine learning models to predict in-hospital mortality in elderly patients with hypertension as well as heart failure. Compared to other algorithms, the Random Forest model performed significantly better. Our study successfully predicted in-hospital mortality and identified the factors most associated with in-hospital mortality.
Study context
- Population
- Elderly patients (≥ 65 years) with HF and hypertension, Chongqing Medical University Medical Data Platform, January 2012 to December 2021
- Participants
- 4,647 patients
- Exposure
- Eight machine learning algorithms for predicting in-hospital mortality
- Outcome
- Prediction of in-hospital mortality (model chosen by AUC)
Methodological detailsmachine learning, multicenter, LASSO variable selection, eight machine learning algorithms compared, SHAP and LIME interpretation
- Machine learning
- Multicenter
- LASSO variable selection
- Eight machine learning algorithms compared
- SHAP and LIME interpretation
Main finding
As reported in the abstract
Prediction of in-hospital mortality
| Group | AUC | Accuracy | Recall | Specificity | Brier score |
|---|---|---|---|---|---|
| Random Forest | 0.850 (95% CI 0.789-0.897) | 0.738 | 0.837 | 0.734 | 0.178 |
Random Forest
AUC0.850 (95% CI 0.789-0.897)
Accuracy0.738
Recall0.837
Specificity0.734
Brier score0.178
Observational data: associations, not effects.
Further findings
Factors most related to in-hospital mortality (SHAP)
- Most related factors: urea, LOS, neutrophils, albumin and HDL-C
Original publication
Machine learning for predicting in-hospital mortality in elderly patients with heart failure combined with hypertension: a multicenter retrospective study. Cardiovasc Diabetol. Published 15 November 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →