Inaccuracy of available surgical risk scores to predict outcomes after transcatheter aortic valve replacement

Introduction Despite encouraging short-term and midterm results, transcatheter aortic valve implantation (TAVI) interventions are still burdened from high rates of adverse events, stressing the need for accurate predictive risk instruments. We compared available surgical risk scores to describe unfavorable outcomes after TAVI. Methods The Age, Creatinine, and Ejection fraction (ACEF) score, the logistic Euroscore, and the Society of Thoracic Surgeons Mortality score (STS) were appraised for their independent power of prediction and for their accuracy (Cindex) to predict 30-day and medium-term mortality, according to the Valve Academic Research Consortium. Results Nine hundred and sixty-two patients were included. All the scores demonstrated a moderate positive correlation. The closest correlation was observed between the STS score and Euroscore. After logistic regression analysis, STS score and Logistic Euroscore provided independent prediction for short-term all-cause mortality [PU0.02, odds ratio (OR) 1.1; 95% confidence interval (CI) 1.06-1.31 and PU0.027, OR 1.03; 95% CI 1.01-1.405]. For in-hospital complications, only STS score performed significantly (PU0.005, OR 1.05; 95% CI 1.01-1.06). ACEF, Euroscore, and STS score showed low accuracy for 30-day all-cause mortality (area under the curve 0.6, 0.44-0.75; vs. 0.53, 0.42-0.61; vs. 0.62, 0.52-0.71, respectively), whereas STS score performed better for in-hospital complications (0.59, 0.55-0.64). Moreover, after Cox-multivariate adjustments, only ACEF score was near to significance to predict all-cause mortality at mid-term (OR 1.7; 0.8-2.9; PU0.058), showing the highest accuracy (0.63, 0.55-0.71). Conclusion In TAVI patients, ACEF score, STS score and Logistic Euroscore provided only a moderate correlation and a low accuracy both for 30-day and medium-term outcomes. Dedicated scores are needed to properly tailor time and kind of approach.