The HEART score has less utility with high sensitivity troponin

Background Clinical risk scores for acute chest pain evaluation were largely validated using older troponin assays. We sought to analyze the HEART score and clinical decision pathways utilizing high sensitivity troponin. Methods This was a retrospective study of all adult patients presenting with nontraumatic chest pain to a single center, quaternary-care emergency department over a 6-month time period. Patients were risk-stratified by the HEART score using an institution-specific clinical decision pathway that aligns with current guidelines. Predictive models and optimal cutoff values were constructed via logistic regression for the primary outcome of acute coronary syndrome at presentation or major adverse cardiovascular events within 30 days and additional analysis was performed among risk groups assessing associations of HEART score with negative troponin and electrocardiogram compared to positive troponin or electrocardiogram. Results Data from 1,014 patients were included for analysis with 46 patients (4.5%) meeting the primary outcome. The model most associated with the primary outcome utilized electrocardiogram and high-sensitivity troponin only (C-statistic = 0.883), which reliably predicted the absence of the primary outcome in intermediate risk patients. The addition of medical history, risk factors, and age to this model (via the continuous HEART score) did not increase its value (C-Statistic 0.876). Conclusion Our study suggests that high-sensitivity troponin and electrocardiogram alone can be used to risk stratify most patients with acute chest pain appropriately. Traditional scoring systems developed with contemporary cardiac troponin assays have less utility in the era of high sensitivity troponin, especially with patients previously deemed intermediate risk.