Reflex testing - II: evaluation of an algorithm for use of cardiac markersin the assessment of emergency department patients with chest pain

A reflex algorithm was developed and evaluated for the use of serum cardiac markers for the diagnosis and rule out of acute myocardial infarction (AMI ), and risk stratification of unstable angina patients for those who presen t to emergency departments (ED) with chest pain. The process begins with te sting of total CK and myoglobin at admission. Based on these results, the a lgorithm determines the need for subsequent testing for the CK-MB isoenzyme and cardiac troponin I (cTnI). The algorithm also directs the need for fur ther blood collection and cardiac marker testing at 4, 8, and 12 h after pr esentation. A total of eleven stopping points were identified. For some of these stopping points, the algorithm concluded that further blood collectio ns and testing was unnecessary and redundant. The algorithm was retrospecti vely evaluated on 101 non-consecutive chest pain patients who presented to the EDs at three hospitals. For the AMI group (n = 34), six of nine possibl e different stopping points were reached: 64.7% of cases were diagnosed wit h the first sample at admission, an additional 32.3% after 4 h, and 2.9% at 8 h. The 12-h sample was not necessary for any of the AMI patients. For th e non-AMI group (n = 67), most reached the stopping point of no cardiac inj ury or risk. There were five unstable angina patients who had minor myocard ial damage on the basis of a marginally increased cTnI. Of these, one patie nt subsequently suffered AMI, and three others required angioplasty or bypa ss surgery. Compared to performing four tests on all patient samples, the r eflex algorithm would have reduced the number of necessary tests from 442 t o 130 (71% reduction) for AMI patients, and 871 to 469 (46% reduction) for non-AMI patients, if prospectively implemented. (C) 1999 Elsevier Science B .V. All rights reserved.