Signal-averaged electrocardiogram for evaluation of recurrent syncope

Signal-averaged electrocardiography (ECG) was performed in 150 consecutive patients presenting with syncope, to determine its diagnostic role in identifying patients with ventricular tachycardia (VT) and in determining their long-term prognosis. Patients also underwent a standardized investigational protocol to independently determine a cause of syncope. Twenty-nine patients had a late potential, 107 had a normal signal-averaged electrocardiogram and 14 had bundle branch block on 12-lead ECG. Signal-averaged ECG identified a late potential in 16 of 22 patients with VT and was normal in 101 of 114 patients in whom syncope was attributed to causes other than VT or remained unexplained (sensitivity 73%, specificity 89%, predictive accuracy 55%). In patients with coronary artery disease, the predictive accuracy increased to 82%. Absence of a late potential identified a group of patients with a very low incidence of VT. During follow-up of 1 to 20 months (median 11), 15 patients (10%) died, 6 suddenly. There was no significant difference in survival or recurrence of syncope between patients with and without a late potential. Signal-averaged ECG can noninvasively identify patients with serious ventricular arrhythmias among an unselected group presenting with syncope.