Abstract WMP61: Recurrent Stroke Rates in Short versus Long Term Cardiac Monitoring in Patients With Embolic Stroke of Unknown Source

Abstract

Background: Outpatient cardiac monitoring for 30 days or longer increases detection rates of paroxysmal atrial fibrillation (AF) after cryptogenic stroke, but the ideal duration of monitoring remains unclear. Two commonly used methods of prolonged outpatient cardiac monitoring are 30-day continuous telemetry and insertable cardiac monitors (ICM). We aim to compare rates of AF detection and recurrent stroke between patients with Embolic Stroke of Unknown Source (ESUS) monitored for approximately 30 days (short term) vs. beyond 30 days with ICM (long term). Methods: We analyzed a single center retrospective cohort of patients discharged with a diagnosis of ESUS over a two-year period. Patients were divided into two groups based on implemented cardiac monitoring method: short-term (30-day non-invasive monitor) and long-term (ICM Results: We identified 117 ESUS patients; 71 patients underwent short-term monitoring and 46 patients underwent long-term monitoring. After a median follow-up of 259 days (IQR 166-468 days), AF was detected in 11.2% (8/71) of patients undergoing short-term cardiac monitoring and 19.6% (9/46) of patients on long-term monitoring; 77.8% of AF was detected beyond 30 days. Recurrent stroke occurred in 12.8% (15/117) of patients. Rates of recurrent stroke were lower in patients undergoing long-term cardiac monitoring vs. only short-term monitoring [adjusted hazard ratio 0.12, 95% confidence interval 0.02-0.89, p = 0.038]. Conclusion: In a real world ESUS patient cohort, long term monitoring was associated with increased detection of AF and reduced risk of recurrent stroke. Large multicenter prospective studies are needed to confirm our findings.