Pathological ECG findings in 24-h-Holter-ECGs in patients with ischemic stroke

Abstract Background Electrocardiogram (ECG)-monitoring covering several days is recommended by current guidelines to detect atrial fibrillation (AF) and other arrhythmias in stroke patients. In clinical practice, the extent of rhythm monitoring varies. In this analysis, we use data from the 24-hour screening-Holter-ECGs of the ongoing randomized multicenter trial Find-AF 2 to assess the detection rate of AF and other sustained arrhythmias. Methods and Results Find-AF 2 (NCT04371055) is a randomized, controlled, open-label, parallel arm, multicenter trial with central AF adjudication (intervention arm) and blinded endpoint assessment 1. Patients ≥60 years with recent (≤30 days) ischemic strokes according to the AHA/ASA definition of any etiology are screened for eligibility. All eligible patients received a 24-hour Holter-ECG prior to randomization which was reviewed by the ECG core lab following a predefined standard operation procedure. The primary analysis focused on the detection of AF ≥30 seconds and the extent of excessive supraventricular ectopic activity. Bradycardic arrhythmias were deemed relevant if they led to pauses 2.5s or if the heart rate 40 bpm was accompanied by higher-degree AV-blocks. Tachycardic arrhythmias were reported if they persisted ≥ 30 seconds, had a mean rate of 120 bpm or had a ventricular origin. A total of 5411 Screening-Holter-ECGs from 51 study centers were analyzed. The mean age was 72.4±7.6 years and 2194 (41%) patients were female. Holter ECGs were performed after a median IQR of 3 2; 5 days post stroke. We found sustained arrhythmias or pauses in 176 patients (3.3%). AF was seen in 78 patients (1.4%) with a median IQR burden of 51% 19; 95 and a median IQR duration of the longest episode of 690 240;1170 minutes (Figure 1). Oral anticoagulation was initiated in all patients with AF. In 72 patients (1.3%), pauses 2.5 s were seen with a mean duration of the longest pause of 3.3±0.9 s and the longest pause lasting for 7.1 s. Seventeen pauses and two bradycardic episodes occurred due to higher-degree AV-blocks, 41 due to sinus node arrest or sino-atrial block and fourteen during AF. One patient received a pacemaker. We observed regular supraventricular tachycardias in 39 patients (0.7%) and one ventricular tachycardia (0.02%). Conclusion Overall, 3.2 % of ischemic stroke patients had relevant arrhythmias, with an AF detection rate of 1.4%. Potentially life-threatening arrhythmias were rare, and the number needed to screen for a change in therapy was 69. Figure 1: Distribution of relevant arrhythmias. SVT=Supraventricular tachycardia. Figure 2: The empirical cumulative distribution is shown for the duration of the longest AF episode. AF= Atrial Fibrillation. Figure 1 Figure 2