Recorded heart sounds for identification of ventricular tachycardia.

Abstract

The ECG discrimination of ventricular tachycardia (VT) vs. supraventricular tachycardia (SVT) is both important and often difficult. In this study, we tested the hypothesis that recorded digital cardiac acoustical data reflect hemodynamic changes that can be used for VT detection. We studied 57 subjects (42 males, mean age 57, range 24-83 years) who had undergone electrophysiological testing for known and suspected cardiac arrhythmias. Acoustic cardiography (Audicor, Inovise Medical, Inc.) was performed during each subject's electrophysiological study. We evaluated the ability of S1 intensity and S1 variability to discriminate between VT and supraventricular rhythm. The 57 subjects had 17 episodes of VT and 76 episodes of supraventricular rhythm--including 22 episodes of SVT. VT had a lower S1 intensity and higher S1 variability than supraventricular rhythm (2.63+/-1.78 mV vs. 4.70+/-5.03 mV and 0.45+/-0.24 vs. 0.21+/-0.11, respectively). Conversely, left bundle branch block, right bundle branch block or SVT did not affect either S1 intensity or its variability. Ventricular pacing increased S1 variability but did not affect S1 intensity. The sensitivity of S1 variability for detecting VT was 50% at 100% specificity. VT is associated with both decreased S1 intensity and increased beat-to-beat S1 variability. The electronic recording and digital processing of digital heart sound data is useful for identifying VT and may facilitate the differential diagnosis of clinically important tachyarrhythmias, particularly in emergency situations where advanced techniques such as electrophysiology studies are not available.