Use of baseline ST-vector magnitude to identify electrocardiographic injury in patients with suspected acute myocardial infarction

Abstract

No information is currently available regarding optimal cut-off values of the ST-vector magnitude (ST-VM) for predicting acute myocardial infarction (AMI) in emergency department (ED) chest pain patients undergoing vectorcardiographic (VCG) monitoring. A prospective observational study was performed in 1,722 chest pain patients with suspected acute coronary syndrome and absence of bundle branch block (BBB) and left ventricular hypertrophy (LVH) on initial ECG who underwent continuous VCG ST-segment monitoring during the initial ED evaluation. Three cut-off values for baseline ST-VM are reported and represent the smallest values in which the positive likelihood ratio (+LR) for AMI is greater than 5, 10, and 20, respectively. AMI occurred in 158 of 1,722 patients (9.2%) without BBB or LVH on initial ECG. Optimal cut-off values at the predetermined +LR values of 5, 10, and 20, were 121 μV (sensitivity, 41.8%; specificity, 92.0%), 151 μV (sensitivity, 29.1%; specificity, 97.1%), and 175 μV (sensitivity, 25.9%; specificity, 98.7%), respectively. Combining the earlier-mentioned cut-off values with physician judgment of initial pretest probability (high, intermediate, or low, respectively) resulted in a relative increase in identification of injury of 37.5% as compared with the ED physician's interpretation of initial ECG (41.8% v 30.4%; P <.0001), and 65.2% as compared with the official ECG interpretation (41.8% v 25.3%; P <.0001). Increasing ST-VM results in increasing likelihood of AMI. Clinical studies need to be performed to determine if ST-VM cut-off values of 121, 151, and 175 μV in conjunction with physician pretest probability of AMI can be used as criterion for emergent reperfusion therapy in patients without LVH or BBB on the initial ECG. (Am J Emerg Med 2002;20:535-540. Copyright 2002, Elsevier Science (USA). All rights reserved.)