ABSTRACT
OBJECTIVES: The SYNTAX score (Ssc) assessing the complexity of coronary anatomy predicts survival after percutaneous coronary intervention (PCI). We tested the hypothesis that the newly developed euroSCORE II (eSC2) can improve the prediction of outcome after complex PCI by the Ssc. METHODS AND RESULTS: Our study comprised 1262 consecutive patients with triple vessel disease or left main stenosis, who were contacted 3 years after elective PCI with drug-eluting stents. We calculated eSC2, Ssc, logistic euroSCORE, and ACEF score. Prediction of 3-year all-cause mortality by these scores was assessed by Cox proportional hazard models. Models were compared by the Hosmer-Lemeshow test for calibration (HL), the C-statistics (AUC) for discrimination and by net reclassification indices (NRI). eSC2 and Ssc were significant predictors of 3-year mortality (unadjusted hazard ratios [95%-confidence limits], 1.050 [1.033-1.067], 1.180 [1.146-1.215], respectively, P<0.001). The predictive value of eSC2 was improved by logarithmic transformation. Adding eSC2 to the model with Ssc improved calibration (HL 7.4 vs. 11.1) and discrimination (increase in AUC [95%-confidence limits] 0.12 [0.07 to 0.17]) and yielded a significant NRI of 0.38 (95%-confidence limits 0.28 to 0.47). The absolute difference in 3-year mortality between strata of Ssc (≤22, >22-32, >32) was smaller with eSC2<1% (1.4%, 3.4%, 9.7%, respectively), than with eSC2>1.6% (11.2%, 20.2%, 30.6%, respectively). The predictive ability of eSC2 was similar to that of the other clinical scores. CONCLUSIONS: eSC2 predicts 3-year mortality after complex PCI and modifies the impact of angiographic complexity on outcome.
