15Early prolonged ambulatory cardiac monitoring in stroke (EPACS) – an open-label randomised controlled trial

Abstract

Introduction: Cardioembolism in paroxysmal atrial fibrillation (PAF) is a preventable cause of transient ischaemic attack (TIA) or ischaemic stroke, however, due its transient nature, a short‐duration Holter monitor may miss a significant proportion of events. There are current limitations in ECG recording beyond 3‐7 days. While implantable loop recorders require a minor surgical procedure, event‐triggered loop recorders are cumbersome to use. We investigated whether there was a role for early prolonged cardiac monitoring using a novel patient‐friendly cardiac monitoring system on patients presenting early after a stroke or TIA. Methods: We conducted an open‐label randomised controlled trial of cardiac monitoring after a TIA or ischaemic stroke (recruited within 72hours of index event) comparing a wearable water‐proof adhesive 14‐day ECG monitoring patch (Zio® Patch, iRhythm Technologies) with a standard 24‐hour Holter ECG for the detection of PAF (Figure). PAF was defined as AF lasting at least 30 seconds. A recruitment target of 120 patients in total was determined following sample size and power analyses. Patients were recruited across two sites; a suburban district hospital (Princess Royal University Hospital, Orpington, UK) and an urban teaching hospital (Kings College Hospital, London, UK). An intention‐to‐treat analysis was performed to compare PAF detection rates using both monitoring techniques. ISCRTN Registration 50253271. Results: The primary outcome of PAF detection rate at day 90 was 7/43 (16.3%) cases in the ZioPatch arm and 1/47 (2.1%) cases in the Holter monitoring arm (OR 8.9, 95% CI 1.1‐76.0; p=0.047). There were no differences in secondary outcomes at 90 days, including second TIA or stroke (ZioPatch vs Holter; 2.3% vs 2.1%; p=0.99) or mortality (ZioPatch vs Holter; 2.3% vs 0%; p=0.96). Forty‐six patients had paired Holter and Ziopatch recordings. Episodes of non‐sustained ventricular tachycardia (NSVT) were found in 15 patients (32.6%), with a significant difference in detection rates between the two monitoring techniques (ZioPatch vs Holter; 32.6% vs 2.17%, p<0.0001). The longest duration of NSVT in those with recorded episodes was 3.26±1.28 seconds. There was an overall drop‐out rate of 20% which was almost entirely due to Holter ECG service provision. Conclusions: Early prolonged monitoring after an index stroke or TIA is superior to 24‐hour Holter monitoring in the detection of PAF. Patch‐based monitoring is a more efficient and patient‐friendly approach to cardiac monitoring compared to Holter‐based monitoring. Further appropriately powered studies are required to determine the impact of increased AF detection on recurrent stroke rates. (Figure Presented) .