Abstract 14779: A Simple, Highly-Sensitive Clinical Prediction Rule to Exclude Acute Coronary Syndrome in a Population with a High Burden of Cardiac Risk Factors
Background: A minority of patients presenting to emergency departments (ED) with chest pain have acute coronary syndrome (ACS). The ADAPT study showed that patients with TIMI Risk Score (TRS) of 0, no ischemic ECG changes and contemporary sensitive troponin assays below the 99th percentile on 2 assessments 2 hours apart were at low risk for 30-day ischemic events. We hypothesized that older patients with traditional cardiac risk factors (and thus, TRS > 0) could also be identified as being low risk if they had troponin I (TnI) levels below the limit of detection (LOD) of the assay, no ischemic ECG changes and no history of coronary artery disease (CAD), and potentially could be discharged without stress testing.
Methods: We performed a single-center retrospective cohort study of 246 consecutive patient care episodes in which a first TnI was drawn in the ED and a second TnI was drawn 4 to 12 hours later. Patients were categorized as potentially eligible for discharge from the ED without stress testing if they had no known CAD, no ischemic ECG changes and both TnI measurements were below the LOD, irrespective of other clinical risk factors. We determined the 30-day incidence of major cardiac events (MI, revascularization, or cardiovascular death).
Results: The median age of our population was 64 years, 95% were male, 54% had pre-existing CAD, and 27% had TRS ≤ 1. At 30 days, 23% had a major cardiac event and 40% underwent stress testing and/or left heart catheterization (LHC) before discharge. Applying our rule (no CAD, no ischemic ECG changes and 2 TnI determinations below LOD), 22% (n=55/246) of patients would have been eligible for discharge without stress testing; 44% of these patients eventually underwent stress testing and/or LHC. The negative predictive value for 30-day cardiac events in this subset was 98%. By comparison, the ADAPT rule identified 8% (p < 0.001 by χ2 test) of patients as low risk with a similar negative predictive value (100%). Universal adoption of this new predictive rule would have eliminated 26% of stress tests and LHCs performed in our population.
Conclusions: If validated in prospective studies, applying this risk model to ED chest pain patients could reduce the number of stress tests performed and identify patients who can be safely discharged from the ED.
Author Disclosures: A.C. Fanaroff: None. R.D. Schulteis: None. S.V. Rao: None. L.K. Newby: None.
- © 2014 by American Heart Association, Inc.