The stress hyperglycemia ratio improves the predictive ability of the GRACE score for in-hospital mortality in patients with acute myocardial infarction.

Abstract

The Global Registry of Acute Coronary Events (GRACE) score is a powerful tool used to predict in-hospital mortality after acute myocardial infarction (AMI) and does not include a glycometabolism-related index. We investigated whether the addition of the stress hyperglycemia ratio (SHR) provides incremental prognostic value in addition to the GRACE score. A retrospective cohort of 613 AMI patients was enrolled in the present analyses. The patients were stratified according to the primary endpoint (in-hospital mortality) and the tertiles of the SHR. During hospitalization, 40 patients reached the primary endpoint, which was more frequently observed in patients with a higher SHR. The SHR, but not admission blood glucose (ABG), adjusted for the GRACE score independently predicted in-hospital mortality [odds ratio 2.5861; 95% confidence interval (CI), 1.3910-4.8080; P=0.0027]. The adjustment of the GRACE score by the SHR improved the predictive ability for in-hospital death (an increase in the C-statistic value from 0.787 to 0.814; net reclassification improvement, 0.6717, 95% CI 0.3665-0.977, P<0.01; integrated discrimination improvement, 0.028, 95% CI 0.0066-0.0493, P=0.01028). The likelihood ratio test showed that the SHR significantly improved the prognostic models, including the GRACE score. Adding the SHR to the GRACE score presented a larger net benefit across the range of in-hospital mortality risk than the GRACE score alone. The SHR, but not the ABG, is an independent predictor of in-hospital mortality after AMI even after adjusting for the GRACE score. The SHR improves the predictability and clinical usefulness of prognostic models containing the GRACE score.