Prediction of Arrhythmic Events After Acute Myocardial Infarction Using Two Methods for Late Potentials Recording

One hundred consecutive patients recovering from an acute myocardiai infarction underwent, prior to home discharge, signal‐averaged electrocardiography (ECG), left ventriculography. and 24‐hour Holter ECG recording. The signal‐averaged ECG was recorded and analyzed using two procedures: the orthogonal bipolar XYZ lead configuration with a bidirectional filter: and a precordial unipolar lead configuration with a uonrecursive digital filter. An abnormal signal‐averaged ECG was seen in 40% of patients with the XYZ system and in 30% of patients in the precordial method, abnormal ejection fraction (< 40%) in 24% of patients and high grade ectopy activity in 22%. During the 24‐month follow‐up period, 12 patients (12%) had an arrhythmic event defined as either sudden death (11 patients) or sustained ventricular tachycardia (1 patient). Neither the signal‐averaged ECG with the XYZ configuration, the abnormal ejection fraction, nor the high grade ectopy were able to statistically predict a higher arrhythmic event rate. The signal‐averaged ECG with the precordial configuration was able to statistically predict a higher arrhythmic event rate, P < 0.03; odds ratio = 3.96. The combination of the orthogonal XYZ configuration signal‐averaged ECG with the ejection fraction (P < 0.01, odds ralio = 7.33), or with ejection fraction and Holter monitoring (P < 0.06. odds ratio = 6.17) was able to predict a higher arrhythmic event rate. The combination of the precordial configuration signal‐averaged ECG with the ejection fraction (P < 0.002, odds ratio = 14.4), or with ejection fraction and Holter monitoring (P < 0.06. odds ratio =10) was able to better predict a higher arrhythmic event rate. The combination of a normal or abnormal signal‐averaged ECG and ejection fraction gave a sensitivity, specificity, positive, or negative value prediction of arrhythmic events of 60%, 90.6%, 37.5%, and 96%, respectively. It must be emphasized that the number of arrhythmic events during the 2‐year follow‐up was small and further study is required to determine the true predictive value of each method for arrhythmic events.