Machine learning-based models to predict one-year mortality among Chinese older patients with coronary artery disease combined with impaired glucose tolerance or diabetes mellitus
Authors' conclusion
This study proposes an accurate model to stratify patients with a high risk of one-year mortality. The gradient boosting machine model demonstrates promising prediction performance. Some interventions to affect NT-proBNP and albumin levels, and statins, are beneficial to improve survival outcome among patients with CAD combined with IGT or DM.
Study context
- Population
- Chinese older patients with coronary artery disease combined with impaired glucose tolerance or diabetes mellitus (DM)
- Participants
- 451 patients
- Comparison
- Training vs Validation
- Trainingn = 308
- Validationn = 143
- Outcome
- One-year mortality (prediction model)
- Follow-up
- One year
Methodological detailsmachine learning, least absolute shrinkage and selection operator (LASSO) with ten-fold cross-validation, random 70:30 split into training and validation cohorts, Web-based application available
- Machine learning
- LASSO with ten-fold cross-validation
- Random 70:30 split into training and validation cohorts
- Web-based application available
Performance
As reported in the abstract
Prediction of one-year mortality: gradient boosting machine model
One-year mortality
26.83%
Gradient boosting machine outperformed other models: Brier score
0.114
Gradient boosting machine: AUC
0.836
Further findings
SHAP feature importance
- Top three features associated with one-year mortality: NT-proBNP, albumin and statins
Original publication
Machine learning-based models to predict one-year mortality among Chinese older patients with coronary artery disease combined with impaired glucose tolerance or diabetes mellitus. Cardiovasc Diabetol. Published 14 June 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →