Spectral turbulence analysis of the signal-averaged electrocardiogram for predicting inducible sustained monomorphic ventricular tachycardia in patients with and without bundle branch block.

Abstract

Spectral turbulence analysis of the signal-averaged electrocardiogram is a new method for identifying patients prone to sustained monomorphic ventricular tachycardia. In contrast to analysis in the time domain, it has been claimed to be applicable in patients with bundle branch block. The aim of this study was to assess the predictive value of spectral turbulence analysis, in relation to the inducibility of sustained monomorphic ventricular tachycardia, in patients with and without bundle branch block. One hundred and sixty nine patients, of whom 120 had a QRS duration > or = 120 ms, were studied. Forty-seven patients had inducible sustained monomorphic ventricular tachycardia and were compared to 122 control patients. The overall sensitivity of the spectral turbulence analysis for predicting inducible ventricular tachycardia was 77%, the specificity 35% and the total predictive accuracy 47%. The limited predictive accuracy was mainly due to a lack of difference between patients with and without ventricular tachycardia in patients with a QRS duration > or = 120 ms. In patients with QRS < 120 ms, however, there were significant differences in all spectral turbulence parameters and the method had a sensitivity of 75%, a specificity of 72% and a total predictive accuracy of 73%. The diagnostic usefulness of spectral turbulence analysis is dependent upon normal QRS duration and the method is applicable only to patients without bundle branch block.