Valvular & structural heart disease
Machine learning for prediction of all-cause mortality after transcatheter aortic valve implantation
Authors' conclusion
Machine learning based on readily available clinical data allows accurate prediction of 1-year all-cause mortality following a successful TAVI.
Study context
- Population
- Patients who underwent transcatheter aortic valve implantation (TAVI) and were discharged home after the index procedure604 derivation, 823 validation
- Comparison
- Machine learning model vs age and EuroSCORE II
- Outcome
- Prediction of all-cause mortality after hospital discharge
- Follow-up
- 1 year
Methodological detailsextreme gradient boosting, repeated 10-fold hold-out testing, external validation in two independent high-volume centres, Derivation at a high-volume centre, January 2013 to March 2019
- Extreme gradient boosting
- Repeated 10-fold hold-out testing
- External validation in two independent high-volume centres
- Derivation at a high-volume centre, January 2013 to March 2019
Performance
As reported in the abstract
AUC for 1-year all-cause mortality (external validation)
| Test or model | AUC |
|---|---|
| Machine learning model | 0.82 (0.78-0.87) |
| Age | 0.52 (0.46-0.59) |
| EuroSCORE II | 0.57 (0.51-0.64) |
AUC
Machine learning model0.82 (0.78-0.87)
Age0.52 (0.46-0.59)
EuroSCORE II0.57 (0.51-0.64)
- Machine learning model vs Age and EuroSCORE II: P < 0.001 — machine learning model superior
Further findings
1-year all-cause mortality · 12 months
Original publication
Machine learning for prediction of all-cause mortality after transcatheter aortic valve implantation. Eur Heart J Qual Care Clin Outcomes. Published 13 January 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →