Abstract 13402: Continuous Rhythm Monitoring in Patients After Embolic Stroke of Undetermined Source Yields High Evaluation Burden

Abstract

Introduction: Embolic stroke of undetermined source (ESUS) accounts for up to 20% of ischemic strokes. Undetected atrial fibrillation (AF) represents one important underlying cause. Once diagnosed, initiation of oral anticoagulation is recommended for secondary stroke prevention. Yet, detection of paroxysmal AF is tedious. Implantable loop recorders (ICM) have evolved as the diagnostic gold standard. Prior studies have indicated that ICM implantation and remote monitoring in particular result in a relevant financial and personnel burden in unselected patients. ESUS patients are considered at higher risk of an AF diagnosis. Hypothesis: We hypothesized that AF diagnosis by ICM and remote monitoring puts considerable strain on health care providers, even in patients at high risk for AF after ESUS. Methods: We investigated all patients of our hospital-based, prospective, single-center, observational ESUS cohort that were equipped with an ICM after ESUS between January 2018 and December 2019. Follow-up was performed through July 2020. Results: Overall, 179 patients were eligible for analysis. Of these, 38 patients (22,1%) were diagnosed with AF by ICM and remote monitoring. During the study period a total of 11,975 episodes were transmitted through remote monitoring, of which 3,766 were alarms for AF. Response to alarms resulted in 307 on site visits and 132 telephone calls to inquire about symptoms, to inform about the diagnosis of AF, and to initiate a change in antithrombotic regimen from antiplatelet therapy to oral anticoagulation. On average, a single ESUS patient generated 40.9 episodes per year for review, triggered one additional on-site visit, and every other patient required additional telephonic follow-up. The number needed to diagnose one ESUS patient with AF within one year was 5. Conclusions: ICM and remote monitoring can successfully diagnose AF in a high-risk cohort after ESUS. Yet, the number of alarms other than for AF, and alarms inappropriately suggesting AF by far exceed appropriate AF diagnoses. Even in a cohort at high risk for AF and with an immediate therapeutic consequence upon AF diagnosis, the resulting burden on personnel and health care resources is high. An optimization of diagnostic ICM algorithms is therefore desirable.