Multicenter investigation of an implantable cardioverter-defibrillator algorithm to detect oversensing.

Background: The SecureSense right ventricular (RV) lead noise discrimination algorithm is designed to detect lead fracture and other types of oversensing in order to decrease inappropriate therapy.

Objective: We studied the real-life accuracy of the SecureSense algorithm in implantable cardioverter-defibrillator (ICD) patients followed by remote monitoring across multiple centers.

Methods: Across 3 French centers, we studied 486 patients with a St Jude Medical device who were followed by remote monitoring and who had the SecureSense algorithm activated. We reviewed ≤10 of the most recent remote monitoring-transmitted electrograms of nonsustained oversensing, RV lead noise, and ventricular tachycardia/ventricular fibrillation that received therapy.

Results: SecureSense inhibited 22 inappropriate therapies (lead dysfunction in 10 cases, P-wave oversensing in 12 cases). A total of 57 patients (12%) sent ≥1 nonsustained oversensing episode (total of 393 episodes) with multiple etiologies: noise on the near-field channel (38%), oversensing of T waves during ventricular pacing (33%), oversensing of the sinus P wave (12%), and oversensing of the paced P wave (6%). Two episodes (0.5%) of nonsustained VT were undersensed by the far-field channel. Of 336 analyzed episodes of ventricular tachyarrhythmia, 15 episodes (4%) in 4 patients were related to oversensing of intrinsic P waves in 11 episodes or oversensing of external noise due to electrical cautery for the remaining 4 episodes.

Conclusion: Of ICD patients equipped with SecureSense, 12% developed episodes of oversensing. The SecureSense algorithm prevented inappropriate ICD therapies with accurate diagnosis of oversensing (caused by lead dysfunction or oversensing of physiological signals). P-wave oversensing in integrated bipolar leads, electrical cautery, and electromagnetic interference are prone to be missed by SecureSense.