Performance of an ICD algorithm to detect lead noise and reduce inappropriate shocks

• Published in: Journal of interventional cardiac electrophysiology Vol. 45; no. 2; pp. 225 - 232
• Main Authors: Beau, Scott, Greer, Stephen, Ellis, Christopher R., Keeney, Jeffrey, Asopa, Shubha, Arnold, Edith, Fischer, Avi
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.03.2016; Springer Nature B.V
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Algorithms; Cardiology; Defibrillators, Implantable - adverse effects; Defibrillators, Implantable - classification; Electric Injuries - etiology; Electric Injuries - prevention & control; Electrocardiography - methods; Electrodes, Implanted; Equipment Failure Analysis - methods; Female; Humans; Male; Medicine; Medicine & Public Health; Middle Aged; Signal-To-Noise Ratio; Young Adult; Inappropriate therapy; Software; Defibrillator lead; Implantable cardioverter-defibrillator
• ISSN: 1383-875X; 1572-8595
• DOI: 10.1007/s10840-015-0081-6

Background Implantable cardioverter-defibrillators (ICD) provide treatment for life-threatening ventricular tachyarrhythmias. Failure of the pace/sense conductor of an ICD lead can cause noise on the sensing electrogram (EGM) that may be misinterpreted as ventricular activity, triggering inappropriate therapy. An algorithm based upon the confirmation of ventricular activity from a far-field EGM has been developed to reduce inappropriate therapies resulting from this type of lead failure, while ensuring that appropriate therapy is delivered. The objectives of this study were to evaluate the algorithm’s ability to discriminate lead noise from ventricular tachycardia/ventricular fibrillation (VT/VF) and to determine whether it inhibits inappropriate shocks without delaying appropriate shocks. Methods The algorithm was prospectively tested using near- and far-field EGM recordings from patients in three conditions: normal sinus rhythm with sustained and non-sustained lead noise via manipulation of the ICD pocket or lead system, and VT/VF induced during defibrillation threshold testing. The recordings were played through a bench-top device running the algorithm with the diagnosis, time to diagnosis, and inhibition of therapy documented. Results The algorithm detected noise and withheld inappropriate therapy in 231 of 238 recordings of sustained lead noise that would otherwise have been diagnosed as VT/VF (97.1 %). Non-sustained lead noise was correctly diagnosed in 47 of the 52 recordings (90.4 %). The device appropriately identified all 853 recordings of VT/VF (100 %), without an increase in the time to detection (0.01 ± 0.14 s). Conclusions The SecureSense TM algorithm correctly diagnosed sustained and non-sustained lead noise recordings without compromising detection of VT/VF. Use of the algorithm may reduce inappropriate shocks and alert clinicians to lead noise indicative of lead failure.