A novel adaptive insertable cardiac monitor algorithm improves the detection of atrial fibrillation and atrial tachycardia in silico. (31st July 2021)

• Record Type: Journal Article
• Authors: Saha, Sunipa; Perschbacher, David; Jones, Paul; Frost, Kate; Sharma, Arjun; Mittal, Suneet; Richards, Mark
• Is Part Of: Journal of cardiovascular electrophysiology. Volume 32:Number 9(2021)
• Page Start: 2536
• Page End: 2543
• Publication Date: 2021-07-31
• Subjects: atrial fibrillation -- atrial flutter -- atrial tachycardia -- cryptogenic stroke -- insertable cardiac monitor -- remote monitoring; Blood vessels -- Physiology -- Periodicals; Electrophysiology -- Periodicals; Heart -- Physiology -- Periodicals; 612.1
• DOI: 10.1111/jce.15178
• Languages: English
• ISSNs: 1045-3873

Abstract

Abstract: Introduction: Insertable cardiac monitors (ICMs) provide a minimally invasive method of continuous monitoring for abnormal heart rhythms. While the benefits of ICMs are clear, current algorithm performance can be improved. The objective of this study is to assess the performance of a novel adaptive atrial fibrillation (AF) detection algorithm and separately programmable atrial tachycardia (AT) algorithm. Methods: A dual‐stage detect‐and‐verify AF algorithm and separately programmable AT algorithm were developed. Sensitivity and PPV across a range of settings were determined in silico by comparison with an adjudicated Holter data set ( n = 1966 with 229 patient days). Finally, the ability to improve performance through simulated remote programming was assessed. Results: The dual‐stage algorithm detected AF in all true AF patients (76/76) resulting in a patient‐level sensitivity of 100%. Episode‐level sensitivity and PPV ranged from 97.6% to 100% and 79.1% to 98.5%, respectively. Thirty‐six false‐positive episodes were observed and 32 (88.9%) of these were corrected with programming changes. Decoupling of AF and AT durations improved PPV from a range of 10%–22% to a range of 95%–100%. Conclusions: AF and AT algorithms were designed with novel features including an adaptive morphology assessment for AF detection and separately programmable durations for AT detection. In silico performance yielded improved PPVs while maintaining high sensitivity across a range of settings. Importantly, programming changes that may be made remotely with this system reduced false positives. These algorithms allow clinicians to individualize arrhythmia detection settings thereby improving data management and reducing clinic burden.