Prediction of 1-year mortality with different measures of ST-segment recovery in all-comers after primary percutaneous coronary intervention for acute myocardial infarction.

BACKGROUND Post hoc analyses from several randomized, controlled trials have established the prognostic importance of different measures of ST-segment recovery in highly selected patients undergoing primary percutaneous coronary intervention (PCI) for ST-segment-elevation myocardial infarction (STEMI). In this single-center registry, we investigated whether various measures of ST-segment recovery can be applied to unselected STEMI patients undergoing primary PCI. METHODS AND RESULTS We analyzed 12-lead ECGs from 2124 consecutive STEMI patients who underwent primary PCI at our institution between November 1, 2000, and January 1, 2007. ECGs were recorded at the catheterization laboratory immediately before arterial puncture and at the end of PCI. We examined measures assessing ST-segment recovery on the postprocedural ECG and measures comparing both ECGs and related these to 1-year, all-cause mortality. Cumulative ST-segment recovery (∑ST-D resolution) at a 50% cutoff had the highest unadjusted accuracy (C statistic, 0.646; 95% confidence interval, 0.602 to 0.689; P<0.001) as compared with the other 8 measures evaluated. Furthermore, ∑ST-D resolution was the strongest contributor to both the net reclassification and integrated discrimination improvement. CONCLUSIONS Although each measure of ST-segment recovery provided univariable prognostic information, the ∑ST-D resolution measure comparing summed ST-segment deviations on the preprocedural and postprocedural ECG was the best independent predictor of 1-year mortality in all-comer STEMI patients after primary PCI.