It is not known in what extent admission glucose improves risk stratification of the GRACE Score in patients with non-ST-segment elevation acute coronary syndromes (ACS). We tested the hypothesis that admission glucose adds relevant prognostic information to the GRACE Score.
Consecutive patients admitted with ACS had plasma glucose measured at admission and cardiovascular events were defined as death, non-fatal myocardial infarction or non-fatal refractory angina during hospitalization.
Among the 148 patients studied, 11.5% developed cardiovascular events. Patients in the forth quartile of admission glucose (> or =175mg/dl) had a greater incidence of events, compared with those in the first 3 quartiles (22% vs. 8.1%; RR=2.7; 95%CI 1.1-6.4; P=0.03). Plasma glucose remained a predictor of events, after adjustment for diabetes (P=0.03). After adjustment for the GRACE Score, glucose in the forth quartile lost its predictive value (P=0.29). Plasma glucose added to GRACE did not improve the C-statistics (0.82; 95%CI 0.75-0.88), as compared with the original Score (0.81; 95%CI 0.74-0.87). Net reclassification improvement by new score was -0.03 (P=0.86), indicating no useful reclassification.
Despite its association with adverse events, admission plasma glucose does not improve GRACE's accuracy to predict in-hospital events in patients with ACS.