A new algorithm for rhythm discrimination in cardioverter defibrillators based on the initial voltage changes of the ventricular electrogram

Aims Ventricular activation onset is faster in supraventricular beats than in ventricular rhythms. The aim of this study was to evaluate a criterion to differentiate supraventricular (SVT) from ventricular tachycardia (VT) based on the analysis of the initial voltage changes in ICD-stored morphology electrograms. Methods Far field ICD-stored EGMs were obtained from 68 VT and 38 SVT episodes in 16 patients. The first EGM peak was detected, three consecutive time epochs were defined within the preceding 80 ms window and the voltage changes with respect to a sinus template were analysed during each time period and combined into a single parameter for rhythm discrimination. Results The algorithm was tested in an independent validation group of 442 VT and 97 SVT spontaneous episodes obtained from 22 patients with a dual chamber ICD. The area under the receiver-operator characteristics (ROC) curve indicated that the arrhythmia separability with this method was 0.95 (tolerance interval: 0.85-0.99) and 0.98 (0.87-0.99) for the control and validation groups respectively. A specificity of 0.91 was obtained at 95% sensitivity in the validation group. Conclusion The analysis of voltage changes during the initial ventricular activation process is feasible using the far field stored electrograms of an ICD system and yields a high sensitivity and specificity for arrhythmia discrimination. (C) 2003 The European Society of Cardiology. Published by Elsevier Science Ltd. All rights reserved.