Fast-Rate Counters as Indicators of Signal Quality and Therapy Delivery in Subcutaneous Implantable Cardioverter-Defibrillators

Abstract

Subcutaneous implantable cardioverter-defibrillators (S-ICDs) prevent sudden cardiac death (SCD) without intravascular leads, reducing complications and infection risk. However, inappropriate therapy remains a key limitation, often driven by low signal quality, marked by reduced R-wave amplitude and diminished R:T ratio, which leads to sensing errors and false tachyarrhythmia detection. This study evaluates fast-rate counter behavior as a marker of signal quality to distinguish true spontaneous tachyarrhythmias from sensing abnormalities. Remote monitoring data from S-ICD patients were analyzed to assess transitions to the tachycardia therapy profile (TTP), triggered when heart rate exceeds the programmed treatment zone. Under adequate signal quality, spontaneous malignant tachyarrhythmias typically produce sustained and organized rate patterns prior to therapy delivery. In contrast, low signal quality may generate erratic and intermittent detections, that can misclassify non-pathological rhythms. To address variability in transmission intervals, an exponential smoothing filter was applied to the fast-rate counter signal: 𝑇𝑇𝑃𝑠𝑚(𝑖)=𝛼·𝑇𝑇𝑃(𝑖)+ (1− 𝛼)·𝑇𝑇𝑃𝑠𝑚(𝑖−1), where 𝛼 (0 < 𝛼 < 1) controls the degree of temporal smoothing. This study analyzed 318 records from the UNTOUCHED dataset [1]. Signals were classified using predefined low signal quality criteria, with independent expert adjudication as the reference standard. Episodes identified as low quality by both the algorithm and the adjudicator were labeled true positives, while those identified by neither were considered true negatives. Per-episode, the method achieved 84.83% sensitivity, 73.78% specificity, and a 97.69% negative predictive value for low signal quality detection. Mean TTPsm values differed significantly between groups (p < 0.01) (Table 1). Low signal quality significantly alters fast-rate counters, closely reflecting sensing performance. Their dynamics may help identify and monitor sensing variation, enabling early detection of signal integrity issues.