AN EVALUATION OF A CHEST PAIN DIAGNOSTIC PROTOCOL TO EXCLUDE ACUTE CARDIAC ISCHEMIA IN THE EMERGENCY DEPARTMENT

Background: Although accelerated diagnostic protocols are being increa singly used in emergency departments to diagnose acute cardiac ischemi a, there have been no prospective evaluations of a chest pain diagnost ic protocol with serial determinations of creatine kinase MB isoenzyme and mandatory exercise electrocardiography (ExECG). Methods: Prospect ive cross-sectional study in which chest pain protocol results were co mpared with final (reference) diagnoses of acute cardiac ischemia (inc luding acute myocardial infarction and unstable angina). Patients in n eed of hospital admission but at low probability (by a validated algor ithm) for acute myocardial infarction were examined for exclusions: kn own coronary artery disease, cardiac complications, severe comorbiditi es, or inability to perform exercise testing. A 12-hour diagnostic pro tocol included serial measurements of creatine kinase MB, EGG, and cli nical assessments followed by ExECG for those with negative initial se rial testing. Reference diagnoses were established during hospitalizat ion and diagnostic accuracy was assessed. Results: The study group of 317 patients was 54% male and 65% black, and had a mean age of 46.6 ye ars; 9.5% had a final diagnosis of acute cardiac ischemia. For this di agnosis, the protocol had a sensitivity of 90.0% (95% confidence inter val, 72.3%-97.4%); specificity, 50.5% (95% confidence interval, 44.6%- 56.4%); positive predictive value, 16.0%; and negative predictive valu e, 98.0%. Creatine kinase MB, serial ECGs, and ExECG each made a contr ibution to improved sensitivity and accuracy, whereas clinical reasses sments were less discriminating, as indicated by protocol's receiver o perating characteristic curve. Conclusions: A chest pain diagnostic pr otocol achieved high sensitivity and improved specificity over the sta ndard emergency department workup. There were no adverse advents assoc iated with early ExECG.