Reproducibility of Ventricular Fibrillation Characteristics in Patients Undergoing Implantable Cardioverter Defibrillator Implantation

INTRODUCTION: The purpose of this study was to evaluate the immediate reproducibility of local electrogram characteristics recorded during repeated episodes of induced ventricular fibrillation (VF) in patients undergoing implantable cardioverter defibrillator (ICD) implantation. METHODS AND RESULTS: Power spectral analysis (using a fast Fourier transform algorithm) of electrograms recorded during 3 seconds of VF were analyzed in 24 patients undergoing ICD implantation using a Medtronic Transvene lead. Patients had 2 to 7 episodes of VF that were induced during defibrillation threshold testing. VF was induced by burst pacing (n = 20) or T wave shock (n = 4). Simultaneous electrograms during VF were recorded from a Medtronic Transvene lead with the following configurations: (1) a narrow spaced (12 mm) dedicated bipole used clinically for sensing; (2) a unipolar electrogram from the right ventricular coil; and (3) a widely spaced (18.3 mm) integrated bipole using the distal tip and the coil. Intraclass correlation coefficients (ICCs) were determined to examine the reproducibility of these VF characteristics among VF episodes in each patient. Recordings from both bipolar configurations had ICCs from 0.40 to 0.55, whereas unipolar recordings ICCs were below 0.40. Reproducibility was similar for dedicated and integrated recordings. CONCLUSIONS: Frequency characteristics of repeated episodes of VF induced in the same subjects show fair-to-good but not excellent reproducibility. Bipolar recordings were far more reproducible than unipolar recordings, but both bipolar configurations had similar reproducibility. These findings have implications for both the pathophysiology of induced VF and the design of VF detection algorithms.