Automatic Identification of Clinical Lead Dysfunctions

Abstract

Implantable cardioverter defibrillators (ICD) lead dysfunctions can cause inappropriate shocks. Current ICDs store lead diagnostics and detected episodes. This stored information with intracardiac electrograms (EGM) and sensed RR interval patterns may characterize the ICD lead performance. The aim of this analysis was to determine the sensitivity and positive predictive value (PPV) of an automatic lead dysfunction identification algorithm. This algorithm uses RR and EGM data to distinguish noncardiac oversensing (OS), for example, due to conductor fracture, and cardiac OS, for example, T-wave OS, from detected episodes. The algorithm also uses lead diagnostics: sensing integrity counter trends (e.g., RR intervals <140 ms), nonsustained tachyarrhythmias episodes with a mean RR <200 ms and impedance trends to identify lead fractures. The PPV was determined using the stored memory from 1,756 ICD patients enrolled in a 13-center long-term lead study with an average follow-up of 18.3 patient-months. Sensitivity was determined in 35 patients who presented with OS or lead fracture-related adverse events confirmed by stored ICD diagnostics. The algorithm sensitivity was 97.1% (34/35). There were 43 additional patients identified by the algorithm without an adverse event. Stored ICD diagnostics confirmed lead dysfunctions in 32 of 43 patients corresponding with an 85.7% PPV (66/77). ICD memory diagnostics and episodes with intracardiac EGM may be used to identify ICD lead dysfunctions with high sensitivity and PPV. This algorithm may be implemented in postprocessing ICD environments (e.g., remote server, programmer) to rapidly identify lead dysfunction prior its clinical manifestation.