Abstract WP231: Predictors for Early Atrial Fibrillation Detection Among Cryptogenic Stroke Patients with Insertable Cardiac Monitors

    loading  Checking for direct PDF access through Ovid

Abstract

Background: The use of insertable cardiac monitors (ICM) has increased the rate of detection of atrial fibrillation (AF) among cryptogenic stroke (CS) patients. We describe a single-center experience for AF detection among CS patients receiving ICMs upon discharge after the index stroke event and attempt to identify predictors for early AF detection.

Methods: From April 2014 to April 2016, patients receiving ICMs upon discharge for CS who underwent >90 days of monitoring were reviewed. Time from ICM placement to AF detection, chronic underlying medical illnesses, presence of left atrial dilatation (LAD) on echocardiography, and PR interval on admission EKG were assessed as predictors of early AF detection.

Results: A total of 114 patients met inclusion criteria and were followed for a median of 415 [268, 557] days. Among these 32 patients (28.1%) were found to have AF at a median of 53 [5, 132] days from ICM placement. Patients with AF detected <30 days from ICM placement had lower rates of hyperlipidemia (35.7% vs 88.9%, p=0.003) and higher rates of hypertension (100% vs 66.7%, p=0.02), tobaccoism (85.7% vs 33.3%, p=0.005), LAD (64.3% vs 16.7%, p=0.01), and prolonged PR interval (195.3±43.2 ms vs 170.3±23.4 ms, p=0.04) compared to patients with AF detected >30 days from ICM placement.

Conclusion: More than one-quarter of CS patients monitored for >90 days with an ICM were found to have previously undiagnosed AF. The majority of patients with AF detected were identified >30 days after their index CS event. Among patients in whom AF was ultimately detected by the ICM, AF may be identified earlier among patients with hypertension and tobaccoism in combination with LAD and prolonged PR interval. Prospective studies are needed to better identify predictors for early AF among the broader population of all CS patients.

Related Topics

    loading  Loading Related Articles