In their article Itsik Ben-Dor et al. compare the accuracy of the STS ( http://209.220.160.181/STSWebRiskCalc261/de.aspx ) and the initial logistic EuroSCORE in predicting 30-day mortality in high-risk patients with severe aortic stenosis who were referred for possible transcatheter aortic valve implantation (TAVI). The total group was divided into three subgroups: group A patients treated by medical therapy, or balloon aortic valvuloplasty, group B, patients undergoing a ‘classic’ aortic valve replacement (AVR), and group C patients with a TAVI. The ability of the two risk scores to predict 30-day mortality for the three groups is poor (C-statistics <70). However the authors misuse the STS score and the EuroSCORE. Both risk stratification models are developed to calculate the risk of mortality of adult cardiac surgery. There is nowadays no discussion that this EuroSCORE overpredicts the risk of mortality of contemporary cardiac surgery and that the STS score is superior for patients undergoing aortic valve replacement . But these two models are not useful to predict mortality of TAVI procedures, and even not of patients undergoing a medical treatment. The most important reason is because these models are based on development data set without patients undergoing TAVI procedures, simply because these procedures didn’t exist at that moment. And I agree with the authors that a specific dataset must be developed for these patients .

The only thing that the STS- or EuroSCORE can predict is the risk of death of a patient in case of a surgical AVR. This predicted risk is one of the criteria that can help the decision to perform a TAVI procedure if the predicted risk of a surgical procedure is ‘not-acceptable’ high. For the STS a surgical risk of 10% is the cutoff point and for the logistic EuroSCORE 20% is accepted as high mortality .

Recently the EuroSCORE II has been launched. ( http://www.euroscore.org ). This EuroSCORE II is an update of the initial EuroSCORE risk. One of the reasons for this update was that also with the years, cardiac surgical mortality, even in high risk patients, decreased. Using this EuroSCORE II, the predicted risk is mostly lower than with the initial EuroSCORE. Interesting now is what shall be the impact of a lower predicted mortality of the surgical procedure (AVR) on the acceptation of patients for TAVI? It is not because the risk of the surgical procedure is nowadays lower than predicted with the initial EuroSCORE, that the percentage of a ‘high mortality risk’ must be lowered.