Feasibility of 90-day self-managed wearable Holter monitoring during the blanking period after atrial fibrillation ablation
T Nakano, T Oka, K Okayama, T Sekihara, K Ozu, Y SakataAbstract
Background
Early recurrence of atrial fibrillation (ERAF) after ablation has been reported to be associated with late recurrence, indicating the clinical value of rhythm monitoring during the 90-day blanking period. Although implantable cardiac monitors (ICMs) have been used for intensive recurrence monitoring, their routine use is not feasible because of cost and invasiveness. Wearable Holter monitoring is less invasive and potentially suitable for prolonged monitoring; however, its practical utility throughout the blanking period remains unclear.
Purpose
We aimed to evaluate the feasibility of 90-day self-managed garment-type Holter monitoring during the blanking period after AF ablation, including patient compliance, clinical-grade ECG data yield, and agreement between AI-estimated and physician-reviewed AF burden (reference standard).
Methods
This prospective pilot study included consecutive patients undergoing pulsed field ablation (PFA) at a single tertiary centre. As a sub-study of a prospective investigation examining the relationship between early and late recurrence after PFA, all patients were offered a self-managed Holter system for continuous use for 90 days after ablation. Feasibility endpoints were acceptance rate, completion rate, and median daily analysable ECG time (IQR). ECG recordings were labelled as AF, non-AF, or noise/missing; noise/missing segments were excluded from analysable time. AF burden was defined as AF time divided by analysable time (AF / [AF + non-AF]). AI-estimated AF burden was generated using an in-house AF detection model and compared with reference AF burden by Pearson correlation and linear regression.
Results
Among 54 consecutive post-PFA patients approached, 10 declined participation and 7 discontinued monitoring; 37 completed the 90-day protocol (acceptance 44/54 (81%); overall completion 37/54 (69%); completion among those who accepted 37/44 (84%)). The main reasons for declining were heat intolerance during summer and difficulty operating the system, whereas the most frequent reason for discontinuation was discomfort or itching related to the wearable electrode. Median daily analysable ECG time was 16.2 h (IQR 13.4 - 18.5 h), indicating substantial clinical-grade data yield under self-management. AI-estimated AF burden showed strong agreement with reference AF burden (Pearson r = 0.974, p <0.001). Linear regression showed AI = 0.0162 + 0.982 × reference burden (R2 = 0.949). The mean difference (AI - physician) was + 1.53 percentage points, with an error IQR of 1.09 percentage points.
Conclusion
Ninety-day self-managed Holter monitoring after AF ablation was feasible and provided substantial analysable ECG data. AI-estimated AF burden closely agreed with physician-reviewed burden, supporting this approach as a practical non-invasive monitoring strategy during the blanking period.Figure1Figure2