Technical Feasibility of an Online Decision Support System for Acute Coronary Syndromes

Abstract

Wide variability presently exists in how patients are treated for suspected acute coronary syndromes (ACS) in the United States. Among the 6 million Americans presenting to the emergency department (ED) for evaluation of chest pain annually,1 of whom half are hospitalized,2 only 20% are found to have a heart attack.3 Moreover, 2% of ED patients actually having an acute myocardial infarction (MI) are mistakenly discharged home.4,5 Accordingly, 100 patients having chest pain must be evaluated in the ED, and 50 must be hospitalized, to diagnose the 10 who are having an acute MI, but this approach further fails to identify 1 additional patient discharged from the ED whose chest pain represented a heart attack. Thus, all patients reporting chest pain are treated as if they were at high risk, potentially resulting in overtreatment and higher costs. Guidelines developed by the American College of Cardiology (ACC) and the American Heart Association (AHA)6,7 permit many patients reporting symptoms of chest pain to be classified as having a low risk for death or nonfatal MI. However, decision support tools that might reliably establish this classification are not presently incorporated into the clinical workflow of EDs and other treatment settings, including urgent care clinics and other outpatient settings. To address the need for decisional support in the triage of patients reporting cardiovascular symptoms, we developed an online Decision Support System (DSS) that standardizes the initial evaluation of patients. The system incorporates 2 key functions: (1) It assists physicians and other health care professionals to comprehensively and accurately elicit patients' symptoms, irrespective of the patient's location; and (2) it transforms patients' symptoms of chest pain into a preliminary or “provisional” diagnosis linked to categories of high, moderate, and low risk for death or nonfatal MI based …