Hospital Admission Despite Low-Risk History, Electrocardiogram, Age, Risk factors, and Troponin (HEART) Scores and 30-Day Outcomes in Emergency Department Chest Pain

Objective To evaluate whether hospital admission of emergency department (ED) patients despite low-risk History, Electrocardiogram, Age, Risk factors, and Troponin (HEART) scores (<4) identifies a subgroup at higher risk for 30-day major adverse cardiovascular events (MACE) and to determine whether clinician override provides prognostic information beyond the HEART score. Methods We conducted a retrospective, multicenter observational study across 6 EDs from January 1, 2021, through December 31, 2023. Adult patients presenting with chest pain and a documented HEART score were included. Analyses were restricted to first encounters among patients with low-risk HEART scores (<4). Clinician override was defined as hospital admission despite low-risk classification. The primary outcome was 30-day MACE, defined as myocardial infarction (MI), coronary revascularization, or all-cause mortality. Multivariable logistic regression identified factors associated with clinician override and evaluated the independent association between clinician override and 30-day MACE after adjustment for age, sex, race or ethnicity, study site, HEART score variables, troponin trend, chronic heart failure, and chronic kidney disease. Incremental prognostic value was assessed using the likelihood ratio and DeLong tests. Results Among 33,392 patients with low-risk HEART scores, clinician override occurred in 2101 (6.3%). Thirty-day MACE occurred in 120 patients (0.36%), including 40 of 2101 admitted patients (1.90%) and 80 of 31,291 discharged patients (0.26%). Clinician override was independently associated with higher odds of 30-day MACE (adjusted odds ratio, 5.16; 95% CI, 3.40 to 7.82). Clinician override significantly improved prediction of 30-day MACE beyond the baseline clinical model (AUROC, 0.8006 vs. 0.8287; DeLong P = .0026; likelihood ratio χ 2 = 49.82, P <.001). Conclusion Among ED patients with low-risk HEART scores, clinician override was uncommon but identified a subgroup at substantially higher risk for 30-day MACE. Although clinician override was largely associated with measurable clinical characteristics, it provided incremental prognostic information beyond the HEART score and major comorbidities, supporting the complementary role of clinician judgment in risk stratification.