Stability: An ICD Detection Criterion for Discriminating Atrial Fibrillation From Ventricular Tachycardia

INTRODUCTION: The purpose of this study was to review a new implantable cardioverter defibrillator (ICD) detection criterion, stability, to determine if it can effectively discriminate rapid rhythms of atrial fibrillation from ventricular tachycardia. Inappropriate shocks for rapid atrial fibrillation limit the acceptance of ICDs. The advent of an additional detection criterion, stability, has been postulated to be of value in discriminating rapid atrial fibrillation, which may not warrant treatment, from ventricular tachycardia, which obviously does warrant therapy deliver. METHODS AND RESULTS: Twenty-six patients were studied during 32 episodes of rapid atrial fibrillation and 24 episodes of monomorphic ventricular tachycardia below 220 beats/min. Each rhythm was repeatedly evaluated by the device at each of the seven stability values available (8, 16, 23, 31, 39, 47, and 55 msec) and then classified as stable or unstable. Upon completion of this acute study, 32 ICD patients had the stability feature activated and were followed for proper arrhythmia treatment by the device. Using stability windows from 8 to 47 msec, all atrial fibrillation rhythms were appropriately classified as unstable. Three of 6 were classified correctly for the 55-msec window. All ventricular tachycardia rhythms were appropriately classified as stable from all stability windows from 8 to 55 msec. Clinical follow-up confirmed appropriate therapy delivery when coupled with sustained rate duration (SRD). Thirty-two patients followed for 292 patient-months had no episodes of untreated ventricular tachycardia with 428 successfully classified as stable and treated. Only three episodes of suspected atrial fibrillation resulted in therapy delivery as the rhythm duration exceeded the SRD of 30 seconds. CONCLUSIONS: The CPI Ventak PRx ICD is highly reliable in appropriately classifying atrial fibrillation as unstable and monomorphic ventricular tachycardia as stable for most stability windows evaluation tachycardias below 220 beats/min. As a result, when testing of atrial fibrillation is not possible, we recommend the routing programming of this stability feature at the 31-msec window with an SRD of 30 seconds. The reliability of this device in discriminating atrial fibrillation from monomorphic ventricular tachycardia may have important clinical implications for other tiered therapy ICDs with this feature as well as for future ICDs in development.