Can Machine Learning Aid the Selection of Percutaneous vs Surgical Revascularization?
Authors' conclusion
An ML-based approach for identifying individuals who benefit from CABG or PCI is feasible and effective.
Study context
- Population
- Patients with 3-vessel and/or left main coronary artery disease (CAD) (SYNTAX trial; CREDO-Kyoto registry)SYNTAX n = 1,800
- Comparison
- Machine learning (ML) models (Lasso regression, gradient boosting) vs SSII-2020 for predicting 5-year death and treatment benefit of coronary artery bypass grafting (CABG) vs percutaneous coronary intervention (PCI)
- Primary endpoint
- Discrimination (C-index) for 5-year all-cause death and treatment benefit between PCI and CABG
- Follow-up
- 5 years
Methodological detailsprognostic model development, machine learning, cross-validation, external validation in CREDO-Kyoto registry
- Prognostic model development
- Machine learning
- Cross-validation
- External validation in CREDO-Kyoto registry
Primary result
As reported in the abstract
5-year all-cause death (C-index)
| Group | C-index, cross-validation | C-index, external validation |
|---|---|---|
| Hybrid gradient boosting | 0.78 (95% CI: 0.75-0.81) | 0.77 (95% CI: 0.76-0.79) |
Hybrid gradient boosting
C-index, cross-validation0.78 (95% CI: 0.75-0.81)
C-index, external validation0.77 (95% CI: 0.76-0.79)
Further findings
Treatment benefit of CABG vs PCI
- ML models identified heterogeneity in the treatment benefit of CABG vs PCI
Original publication
Can Machine Learning Aid the Selection of Percutaneous vs Surgical Revascularization? J Am Coll Cardiol. Published 28 November 2023. doi:10.
Summarised from the published abstract; every number is checked against it