External validation of the GRACE risk score in patients with myocardial infarction in Hungary

Abstract

Background: Literature confirms that the Global Registry of Acute Coronary Events (GRACE) risk score provides a better risk evaluation than clinical judgment in patients with acute myocardial infarction. We aimed to externally validate the GRACE risk score in unselected patients with myocardial infarction in Hungary. Methods: Data from the comprehensive Hungarian Myocardial Infarction Registry (HUMIR), a national registry that collects data on consecutive acute myocardial infarction (AMI) patients, were used. Hospitals registered 102,939 infarction events in the HUMIR between January 1, 2014, and December 31, 2020. The data required to calculate GRACE risk score were available for 75,199 events. We studied the 6-months, 1-year, and 3-year outcomes. We calculated widely used metrics to characterise calibration (calibration curve, calibration intercept and slope, Eavg, Emax, and E90) and discrimination (c-score, equivalent to AUC, and Somer's Dxy). Results: The risk of low-risk patients was underestimated, and the risk of high-risk patients was overestimated. However, the deviation was small, especially for the three-year survival (E90 was 0.15, 0.22, and 0.08). Discrimination was good, with an AUC of approximately 0.8, and was very similar in all the periods. Conclusions: These data confirmed the usefulness of GRACE risk score in selecting high-risk patients with myocardial infarction in the Hungarian population.