Modified HEART score to improve ruling out acute coronary syndrome at the emergency department

Background Beyond high-sensitivity troponin (hs-Tn) dosage, risk scores are recommended (e.g. HEART score) to identify patients which could be safely discharged from ED. A modified and simplified version of HEART score (mHEART score) has been developed for the same goal. Objective We aimed to prospectively validated the mHEART score and compared its performance with that of two other commonly used score, GRACE and classic HEART scores. Methods Prospective observational study carried out from November 2019 to November 2020. Consecutive adult patients admitted to the ED of University Hospital of Bologna for non-traumatic chest pain were enrolled. For each enrolled patient GRACE, classic HEART and mHEART scores were calculated. The main outcome was the occurrence of major cardiovascular events and overall mortality over 90 days (MACE). Results The mHEART score had higher AUC (0.891) as compared to classic HEART (0.884) and GRACE (AUC 0.804) scores (p < 0.001). The classic HEART score identified 48% patients at low risk, mHEART score 51% and GRACE 63%. The cumulative MACE incidence in these groups were 1.0%, 1.0% and 6.1%, respectively. The modified and classic HEART scores showed the highest negative predictive value (99%). If combined with multiple hs-Tn testing, the mHEART allowed to identify half of the population (49.8%) that could have been discharged with a MACE miss-rate of 0.5%. Conclusions mHEART score showed the best performance for ruling out ACS among patients with chest-pain presenting in the ED. Integration of this score with current hs-Tn algorithm might enhance management of patients with chest-pain in ED.