Objective: The HEART score has been proposed for emergency department (ED) prediction of major adverse cardiac events (MACE). We sought to summarize all studies assessing the prognostic accuracy of the HEART score for prediction of MACE in adult ED patients presenting with chest pain.Methods: We searched MEDLINE, PubMed, EMBASE, Scopus, Web of Science, and the Cochrane Database of Systematic Reviews from inception through May 2018 and included studies using the HEART score for the prediction of short-term MACE in adult patients presenting to the ED with chest pain. The main outcome was short-term (i.e., 30-day or 6-week) incidence of MACE. We secondarily evaluated the prognostic accuracy of the HEART score for prediction of mortality and myocardial infarction (MI). Where available, accuracy of the Thrombolysis in Myocardial Infarction (TIMI) score was determined.
Results:We included 30 studies (n = 44,202) in analysis. A HEART score above the low-risk threshold (≥4) had a sensitivity of 95.9% (95% confidence interval [CI] = 93.3%-97.5%) and specificity of 44.6% (95% CI = 38.8%-50.5%) for MACE. A high-risk HEART score (≥7) had a sensitivity of 39.5% (95% CI = 31.6%-48.1%) and specificity of 95.0% (95% CI = 92.6%-96.6%) for MACE, whereas a TIMI score above the low-risk threshold (≥2) had a sensitivity of 87.8% (95% CI = 80.2%-92.8%) and specificity of 48.1% (95% CI = 38.9%-57.5%) for MACE. A high-risk TIMI score (≥6) was 2.8% sensitive (95% CI = 0.8%-9.6%), but 99.6% (95% CI = 98.5%-99.9%) specific for MACE. A HEART score ≥ 4 had a sensitivity of 95.0% (95% CI = 87.2%-98.2%) for prediction of mortality and 97.5% (95% CI = 93.7%-99.0%) for prediction of MI.