Background. Emergency Department (ED) evaluation of patients with acute chest pain and non-diagnostic electrocardiography (ECG) remains a frequent and difficult problem. Stress testing is generally requested and may lead to costly, unnecessary and prolonged ED stay. We aimed to identify clinical characteristics that allow early discharge with outpatient evaluation of such patients.
Methods. Between 02/2006 and 02/2007, 459 adults [age 56±13 years, 228 (50%) men] who presented to ED with chest pain and had no ischemic ECG changes and 2 negative cardiac enzymes measurements on arrival and 6 hours later, underwent stress echocardiography (n=396) or coronary computed tomographic angiography (n=63) prior to discharge.
Results. Non-invasive evaluation for coronary artery disease (CAD) was negative in 367 (80%), indeterminate in 30 (7%) and positive in 62 (13%) [Table
]. Univariate predictors for a normal stress test were lower Framingham risk score, lower age, normal ECG (no ischemic changes), atypical chest pain and the absence of diabetes, hypertension, high cholesterol and previous CAD. Multivariate logistic regression identified normal ECG (RR=2.3, p=0.002), atypical chest pain (RR=3.0, p=0.001), no diabetes (RR=2.5, p=0.001) or absence of previous CAD (RR=2.8, p<0.001) as strong independent predictors for a negative non-invasive test. The same four variables were strong independent predictors of ED discharge. Of 459 patients, 186 (41%) had none of the four independent predictors. None of those had positive non-invasive test or obstructive CAD on angiogram (n=1).
Conclusion. A substantial portion (2/5) of patients seen in ED for chest pain evaluation can be safely discharged after 2 sets of negative cardiac enzymes (~6 hours) if they have a normal ECG, atypical chest pain, no diabetes or previous CAD. These patients can be evaluated for CAD as an outpatient. This strategy may significantly decrease length of stay, ED clutter and cost of in-patient evaluation.
Coronary CT angiography in emergency department patients with low-risk acute chest pain: diagnostic yield and clinical outcome