Mortality risk after transcatheter aortic valve implantation: analysis of the predictive accuracy of the Transcatheter Valve Therapy registry risk assessment model

Aims: The risk assessment tools currently used to predict mortality in transcatheter aortic valve implantation (TAVI) were designed for patients undergoing cardiac surgery. We aimed to assess the accuracy of the TAVI dedicated risk score in predicting mortality outcomes. Methods and results: Consecutive patients (n=1,038) undergoing TAVI at a single institution from 2014 to 2016 were included. The ACC/TVT registry mortality risk score, the STS-PROM score and the EuroSCORE II were calculated for all patients. In-hospital and 30-day all-cause mortality rates were 1.3% and 2.9%, respectively. The ACC/TVT risk stratification tool scored higher for patients who died in-hospital than for those who survived the index hospitalisation (6.4 +/- 4.6 vs. 3.5-11.6, p=0.03, respectively). The ACC/TVT score showed a high level of discrimination, C-index for in-hospital mortality 0.74, 95% CI: (0.59-0.88). There were no significant differences between the performance of the ACC/TVT registry risk score, the EuroSCORE II and the STS-PROM score for in-hospital and 30-day mortality rates. Conclusions: The ACC/TVT registry risk model is a dedicated tool to aid in the prediction of in-hospital mortality risk after TAVI.