The revised EuroSCORE II for the prediction of mortality in patients undergoing transcatheter aortic valve implantation

Abstract

The assessment of procedural risk is crucial in patients with severe symptomatic aortic stenosis. Logistic EuroSCORE and STS score are currently used to estimate procedural risk and mortality for surgical and transcatheter aortic valve implantation (TAVI). The recently published EuroSCORE II might provide a helpful tool. The new EuroSCORE II was calculated in 206 patients undergoing transfemoral TAVI and compared to the established logistic EuroSCORE and STS mortality score. Discriminative power and calibration of each test was statistically evaluated. 30-day and 1-year mortality rates were 6.8 % (14/206) and 27.2 % (56/206). In-hospital mortality rate was 29.2 ± 17.8 % with logistic EuroSCORE, 9.5 ± 6.8 % with STS score, and 9.22 ± 7.12 % with EuroSCORE II: Logistic EuroSCORE and EuroSCORE II were significantly increased in non-survivors compared to survivors at 30 days and at 1 year. EuroSCORE II and STS score (r = 0.49, p < 0.001) showed moderate correlation, whereas strong correlation was found between EuroSCORE II and logistic EuroSCORE (r = 0.71, p < 0.001). ROC curve analyses for the prediction of 30-day mortality (AUC 0.79 vs. 0.69 vs. 0.71) and 1-year mortality (AUC 0.72 vs. 0.70 vs. 0.70) were performed. Statistical comparison revealed no difference between the AUCs (p > 0.05). In percutaneous TAVI patients, 30-day mortality was best approximated by the new EuroSCORE II, whereas the logistic EuroSCORE best reflected 1-year mortality. However, EuroSCORE II does not provide additional prognostic information beyond the established logistic EuroSCORE. For more exact risk prediction in TAVI, a distinct TAVI risk score would be desirable.