A 13-variable machine learning model predicted 1-year mortality post-TAVR with an AUC of 0.84 in external validation, outperforming EuroSCORE II (AUC 0.70), FRANCE-2, and TAVI2-SCORE.
Cohort (n=1,295)
Yes
Does a parsimonious machine learning model using preprocedural variables improve prediction of 1-year all-cause mortality in TAVR patients compared to existing risk scores?
A 13-variable machine learning model significantly outperforms existing risk scores in predicting 1-year mortality after TAVR, potentially aiding in identifying procedural futility.
Effect estimate: NRI 0.62 (95% CI 0.49 to 0.75)
Absolute Event Rate: 0.84% vs 0.7%
BACKGROUND: Current risk scores inadequately predict long-term mortality after transcatheter aortic valve replacement (TAVR), limiting their ability to guide decisions around procedural futility. We aimed to develop and externally validate a machine learning (ML) model using only preprocedural variables to predict 1-year all-cause mortality. METHODS: An ML model was trained on a retrospective cohort of 1025 TAVR patients using 52 clinical and echocardiographic variables. Feature selection and model tuning were performed via a multiobjective evolutionary algorithm to optimise predictive performance and model simplicity. The final model used 13 preprocedural variables and was externally validated in an independent cohort of 270 patients. Performance was compared with European System for Cardiac Operative Risk Evaluation II (EuroSCORE II), FRANCE-2 and TAVI2-SCORE using the area under the curve (AUC), calibration and net reclassification improvement (NRI). RESULTS: The ML model demonstrated excellent discrimination, with AUCs of 0.81 in the discovery cohort and 0.84 in the external validation cohort. This exceeded the performance of EuroSCORE II (AUC: 0.61 and 0.70), FRANCE-2 (0.52 and 0.58) and TAVI2-SCORE (0.56 and 0.64). Calibration plots showed strong agreement between predicted and observed risks. NRI in the test set compared with FRANCE-2 was 0.62 (95% CI: 0.49 to 0.75); compared with TAVI2-SCORE, it was 0.36 (95% CI: 0.14 to 0.61). The final model incorporated age, atrial fibrillation, creatinine, haemoglobin, pulmonary function, frailty markers (Katz Index, poor mobility) and tricuspid regurgitation. Misclassification analysis revealed that most errors were clustered near the decision threshold, with no evidence of systematic bias. Performance was consistent across subgroups and robust to temporal and institutional variation. CONCLUSION: This externally validated ML model, using 13 routinely available variables, significantly outperforms existing risk scores in predicting 1-year mortality post-TAVR. Its simplicity and generalisability support its potential use in real-world clinical decision-making to identify patients at high risk of procedural futility.
Bharucha et al. (Fri,) conducted a cohort in Transcatheter aortic valve replacement (TAVR) (n=1,295). Machine learning model (13 preprocedural variables) vs. EuroSCORE II, FRANCE-2, and TAVI2-SCORE was evaluated on 1-year all-cause mortality prediction (AUC) (NRI 0.62, 95% CI 0.49 to 0.75). A 13-variable machine learning model predicted 1-year mortality post-TAVR with an AUC of 0.84 in external validation, outperforming EuroSCORE II (AUC 0.70), FRANCE-2, and TAVI2-SCORE.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: