Survival posT Tavi: the STT score. Derivation and external validation.

Introduction:Surgical risk scores have failed to accurately predict events after TAVI, stressing the need for a dedicated tool incorporating all clinical variables useful to discriminate risk of patients undergoing TAVI.

Methods:All consecutive patients undergoing TAVI at our institutions were enrolled. The primary endpoint was one year all-cause mortality (as for VARC definition), and secondary endpoint was all-cause death after 30 days. All clinical variables significantly related to the primary end point at multivariate analysis were incorporated. Accuracy of the score was derived, and externally validated. Net classification improvement compared to STS score was appraised.

Results: A total of 1064 patients constituted the derivation cohort, and 180 the external validation cohort. 165 patients (15%) died at one year follow up, and previous stroke (OR 1.80; 1.4-3), inverse of renal clearance (OR 8; 6-14) and systolic pulmonary arterial pressure more than 50 mmHg (OR 2.10; 1.5-3) were independently related to all-cause mortality. Area Under the Curve (AUC) of the Survival posT TAVI (STT) for the primary end point was of 0.68 (0.62-0.71). At 30 days, 65 (7%) died, and AUC for them was 0.66: (0.64-0.75). On external validation cohorts, AUC curves of STT were, respectively, 0.66 (0.56-0.7) for 30 days and 0.67 (0.62-0.71) for all-cause mortality at follow-up. Net reclassification improvement (NRI), for 30 days all-cause mortality for STT compared to STS was of 31% (p<0.001), and of 14% (p<0.001) for all-cause mortality at follow up.

Conclusions: STT represents an easy and accurate score to define risk of 30 days and of mid-term mortality for patients undergoing TAVI, improving STS score performance.

On the left ROC of STT  for all cause 30 days death on external validation cohort (0.66 0.56-0.7). On the right ROC of STT  for all cause death at follow up on external validation cohort (ROC 0.67: 0.62-0.71)

writed  at 19-10-2013 16:15:04