Establishment of a hybrid risk model to predict major cardiac adverse events in patients with non-ST-elevation acute coronary syndromes.

Abstract

The present study aimed to generate a hybrid risk model for the prediction of major cardiac adverse events (MACE) in patients with non-ST-elevation acute coronary syndromes (NSTE-ACS), by combining the Global Registry of Acute Coronary Events (GRACE) scoring system and the plasma concentration of N-terminal of the prohormone brain natriuretic peptide (lgNT-proBNP). A total of 640 patients with NSTE-ACS were randomly divided into either the model-establishing group (409 patients) or the prediction model group (231 patients). The clinical endpoint event was MACE, including cardiogenic death, myocardial infarction and heart failure-induced readmission. Among the 409 patients in the model-establishing group, 26 (6.6%) experienced MACE. The hybrid risk model was calculated using the following equation: Hybrid risk model = GRACE score + 20 × logarithm (lg)NT-proBNP + 15, in which the area under the receiver operating curves (ROCs) for the GRACE score and lgNT-proBNP were 0.807 and 0.798, respectively. From the equation, the area under the ROC for the hybrid risk model was 0.843; thus suggesting that the hybrid risk model may be better able to predict the occurrence of MACE compared with either of its components alone. Following re-stratification, 6% of patients in the hybrid risk model were re-grouped. A total of 15 patients in the prediction model group experienced MACE (6.5%). The areas under the ROCs for the hybrid risk model and the GRACE scores for the prediction model group were 0.762 and 0.748, respectively. The results of the present study suggested that the lgNT-proBNP and GRACE score-established hybrid risk model may improve the accuracy by which MACE are predicted.