Abstract 2437: Timing of Mobile Cardiac Outpatient Telemetry May Increase Diagnostic Yield of Atrial Fibrillation in Select Patients with Cryptogenic Strokes

Abstract

Introduction: Cryptogenic stroke accounts for 30-40% of ischemic strokes. Routine Holter monitoring detects atrial fibrillation (AF) in only 5-7% of patients with ischemic strokes. However, prolonged cardiac monitoring increases the diagnostic yield of AF. Objective: In this study we used mobile cardiac outpatient telemetry (MCOT)and sought to determine 1)what percentage of our patients with cryptogenic stroke had underlying AF, 2) whether a period of highest diagnostic yield for AF could be derived and 3) whether there are clinical and electrocardiogram(EKG) parameters that predict AF in a select population. Methodology: Between March 2009 and December 2010, the data for 50 consecutive patients with MCOT at the stroke follow up clinic were analyzed. All had imaging, echocardiogram, EKG, hypercoagulable panel and at least 48 hours of in-patient cardiac telemetry monitoring. Of these, 16 were excluded because they either had no restricted diffusion on Diffusion Weighted Imaging (DWI) or had procedure-related strokes, hypercoagulable state or TEE detected mobile plaques in the ascending aorta. Results: The final 34 patients (19 males, 15 females, mean age of 62.8±12.9) had a diagnosis of cryptogenic stroke.Magnetic rasonance imaging( MRI) detected anterior circulation strokes in 67.6% (23/34) of the cases with the remaining being posterior circulation strokes or a combination of both. MCOT of 21 days duration was initiated up to a median of 8 weeks (2-24 weeks) post discharge. Overall, AF was detected in 29.4% (10/34) of the patients within a median of 8.5 days. Of these, the detection rates were 8.8% within the first 7 days and 26.5% within 14 days. Beyond 14 days, AF detection rate increased by only 3%. Regression analysis indicated that enlarged left atrial size (>4 cm) was not a significant predictor of AF in our patient population. The combination of stroke, dyslipidemia, coronary artery disease, atrio-ventricular block II type 1, supraventricular tachycardia and regional wall motion abnormality on echocardiogram however was correlated with a higher diagnostic yield of detecting AF on MCOT (r 2 =0.72). Conclusion: Occult AF is common in patients with cryptogenic stroke. We found increased detection of atrial fibrillation with monitoring between 7 to 14 days, especially in those with an underlying history of stroke, coronary artery disease and certain EKG rhythm abnormalities.