Multisite Validation of a Strategy to Identify Very Low Risk Emergency Department Patients Without Troponin

Abstract

Background

Patients with chest pain who are very low risk, defined by a History, Electrocardiogram, Age, and Risk factors (HEAR) score ≤1, may not require troponin testing.

Objectives

The aim of this study was to determine whether troponin testing is needed in patients with HEAR scores ≤1 in a multisite U.S. cohort.

Methods

We conducted an observational cohort study using the Wake Forest Chest Pain Registry. Patients ≥18 years old with HEART Pathway assessments and high-sensitivity troponin testing were accrued from 5 U.S. emergency departments (November 1, 2020-July 7, 2022). HEAR scores were prospectively completed by the treating clinician for patients with no known coronary artery disease and a nonischemic electrocardiogram. The outcome was 30-day major adverse cardiovascular events (MACE) (death, myocardial infarction [MI], and revascularization). The proportion of patients with HEAR scores ≤1 with MACE within 30 days was determined, and test characteristics were calculated. The net reclassification improvement index for troponin testing among patients with HEAR scores ≤1 was determined.

Results

Among 9,105 patients, 17.2% (1,565/9,105) had a HEAR score ≤1. At 30 days, MACE occurred in 0.7% (11/1,565; 95% CI: 0.4-1.3), with 3 deaths, 8 MIs, and 1 revascularization. The sensitivity and negative predictive value for 30-day MACE in patients with a HEAR score ≤1 were 97.9% (95% CI: 96.2-98.9) and 99.3% (95% CI: 98.7-99.6). Troponin testing correctly reclassified 8 with death, MI, or revascularization. Troponin was elevated among 74 without MACE, yielding a nonsignificant net reclassification improvement index of 0.7% (95% CI: −0.4 to 1.8).

Conclusions

Patients with no known coronary artery disease, a nonischemic electrocardiogram, and a HEAR score ≤1 had a missed MACE rate <1%. Troponin testing identified additional patients with MACE but did not significantly improve risk stratification accuracy.