Predictive value of the signal-averaged electrocardiogram for early mortality after acute myocardial infarction

Background: Signal-averaged electrocardiogram (SAECG) has been shown to identify patients at risk for ventricular tachycardia and sudden death after acute myocardial infarction (AMI). The aim of the present study was to assess and compare the diagnostic and prognostic value of abnormal SAECG and low left ventricular ejection fraction (LVEF) in the early (≤30 days) postinfarction period. Methods: We prospectively studied 112 consecutive patients with first AMI. SAECG of six precordial leads was performed within 48 hours and LVEF was measured within 7 to 10 days of AMI. Results: Eight patients (group I) died within 30 days of the AMI: three patients with documented ventricular tachycardia degenerating into ventricular fibrillation, one of unwitnessed sudden death, and four of electromechanical dissociation or cardiogenic shock. One hundred and four patients (group II) survived the initial 30 days of hospitalization. Logistic regression analysis identified abnormal root mean square of the last 40 ms (RMS₄₀; adjusted relative odds ratio = 11.0; P = 0.06) but not filtered QRS duration (QRSD; P = 0.8) as independent predictors of early cardiac mortality. Abnormal LVEF (≤30%) was found to be the single most powerful predictor of cardiac death in this subsample (adjusted relative odds ratio = 19.9; P = 0.02). Enhanced predictive value was achieved by combining abnormal LVEF with either abnormal QRSD or RMS₄₀ (from 42% to 67%). The combination of abnormal LVEF and QRSD had an excellent positive predictive value (100%) for sudden cardiac death. Conclusions: SAECG and LVEF, recorded soon after AMI, have independent significant predictive power for early (≤30 days) arrhythmic and nonarrhythmic cardiac death, which is enhanced by their combination. LVEF has better predictive value than SAECG for identifying patients at risk for early cardiac death after AMI.