Development and testing of an algorithm to detect implantable cardioverter-defibrillator lead failure

Area of Science:

• Biomedical Engineering
• Cardiology
• Medical Device Technology

Background:

• Implantable cardioverter-defibrillators (ICDs) can deliver inappropriate shocks due to lead failures mimicking arrhythmias.
• Distinguishing lead failure noise from true ventricular tachyarrhythmias is crucial for patient safety and device efficacy.

Purpose of the Study:

• To evaluate an algorithm designed to differentiate noise from lead failure in ICDs from actual ventricular tachyarrhythmias.
• To assess the algorithm's accuracy in preventing unnecessary shocks.

Main Methods:

• The algorithm analyzes the ventricular electrogram baseline, specifically the far-field electrogram, to detect lead failure noise.
• A threshold of <0.35 mV-mV for the minimum baseline measure over 12 beats was used to classify noise.
• Data from 24 patients with lead failure and 56 patients with ventricular arrhythmias were analyzed.

Main Results:

• The lead failure episodes showed significantly lower minimum baseline measures (0.28 ± 0.34 mV-mV) compared to ventricular tachycardia (40.8 ± 43.0 mV-mV) and ventricular fibrillation (19.1 ± 22.8 mV-mV).
• The algorithm achieved 83% sensitivity and 100% specificity in detecting lead failures using the established threshold.

Conclusions:

• A far-field electrogram baseline measure effectively identifies lead failure noise.
• This algorithm demonstrates high accuracy in differentiating lead failure from life-threatening arrhythmias, potentially preventing inappropriate ICD shocks.