Objectives: This study was conducted to evaluate criteria for discrimination of ventricular tachycardia from atrial fibrillation and sinus tachycardia in a tiered-therapy cardioverter-defibrillator (Medtronic PCD).
Background: Interval stability algorithms discriminate ventricular tachycardia from atrial fibrillation. Onset algorithms discriminate ventricular tachycardia from sinus tachycardia. Neither has been validated clinically.
Methods: The stability criterion requires that a ventricular tachycardia interval not vary from any of the three previous intervals by more than the programmable stability value. The onset criterion detects initiation of ventricular tachycardia only if the ratio of an interval to the mean of four previous intervals is less than a programmed onset ratio and either the second or third preceding interval exceeds the ventricular tachycardia detection interval. We evaluated these criteria in 100 patients at electrophysiologic study and exercise testing (65 patients) and during a mean (+/- SD) follow-up of 16.2 +/- 7.9 months. The PCDs were programmed to tiered therapy in 54 patients. In the remaining 46 patients, the PCD's memory for detected ventricular tachycardia was used to study the specificity of the chosen onset criterion for rejecting sinus tachycardia. We used stored intervals preceding appropriate (n = 99) and inappropriate (n = 54) detections to test a new onset criterion that was less sensitive to a single index interval.
Results: Programmed stability of 40 ms decreased detection of induced atrial fibrillation by 95% (20 patients), paroxysmal atrial fibrillation by 95% (6 patients) and chronic atrial fibrillation by 99% (9 patients); all episodes of spontaneous (n = 877) and induced (n = 339) ventricular tachycardia were detected. A programmed onset ratio of 87% rejected sinus acceleration (98%) but caused underdetection of 0.5% of ventricular tachycardias. The onset criterion permitted inappropriate detection of premature ventricular complexes during sinus tachycardia, but the new criterion reduced these inappropriate detections by 98%.
Conclusions: The PCD's onset and stability criteria reduced inappropriate detection of atrial fibrillation and sinus acceleration while detecting 99.5% of ventricular tachycardias.