Performance of a dual-chamber implantable defibrillator algorithm for discrimination of ventricular from supraventricular tachycardia.

BACKGROUND

Inappropriate therapies remain a major problem in patients with implantable cardioverter defibrillators (ICDs). Decreasing the proportion of inappropriate therapies is a major objective. With the addition of atrial detection and advanced algorithms, dual-chamber ICDs are designed to offer better discrimination of ventricular (VT) and supraventricular (SVT) arrhythmias. The present multicentre, open study aimed to evaluate the performance of a dual-chamber detection algorithm, the Atrial View algorithm, incorporated in a dual-chamber ICD, the Ventak AV (Guidant Inc., St. Paul, Minnesota, USA).

METHODS AND RESULTS

Fifty-one patients (45 males, 62+/-11 years, ejection fraction 42+/-15%) with standard indications received a Ventak AV ICD which analyzes, within the VT zone RR stability, tachycardia onset, atrial rate and AV relationship. Predischarge enhanced-detection algorithms were prospectively programmed: stability 24 ms, onset 9%, atrial fibrillation threshold 200 beats/min, and Vrate>Arate. An additional sustained rate duration criterion was programmed at least at 30 s. ICDs were interrogated every 3 months or when patients received shocks. A blinded review of electrograms for arrhythmia diagnosis and appropriateness of therapy was performed by 2 experts. Over the follow-up period (12+/-3.6 months), a total of 400 tachycardia episodes was recorded within the VT zone. After the review of stored electrograms, 237 (59%) true positive, 143 (36%) true negative, 17 (4%) false positive and 3 (1%) false negative episodes were diagnosed. Considering the 3 VTs incorrectly detected by the detection algorithms, therapy was delivered in 2 cases after sustained rate duration and 1 VT reverted spontaneously. Inappropriate therapy occurred in 17 cases. All but 1 were related to SVT with 1:1 atrioventricular relationship. Finally, on a per episode basis, the detection algorithm sensitivity was 99% and specificity was 89%.

CONCLUSIONS

Programming of detection criteria based on stability, onset, atrial fibrillation rate threshold and Vrate>Arate allows a 99% sensitivity and an 89% specificity in Guidant ICDs. Discrimination of SVT with 1:1 atrioventricular relationship, however, remains a challenge for which new algorithms have to be designed.