Abstract WP199: Left Atrial Volume Index and PR Interval Are Independent Predictors of Atrial Fibrillation in Embolic Stroke Patients With Insertable Cardiac Monitor

Abstract

Introduction: The use of insertable cardiac monitors (ICM) improved the detection rate of atrial fibrillation (AF) in patients with embolic strokes (ES). However, given cost of the device, current indications and availability of implant procedure, the optimal patient population that would benefit from the device remains unknown. Objectives: We sought to determine 1) the incidence of AF in patients with ES in our population and 2) whether certain variables are predictive of AF in patients with ES. Methods: A case control study was conducted from January 1, 2015 to June 23, 2017. Patients admitted to the Gates Vascular Institute with ES on MRI and CT angiogram but no evidence of AF on telemetry underwent Reveal LINQ ICM implant prior to discharge. Demographics, EKG and echocardiographic variables were assessed. Statistical analysis was conducted using SPSS version 22 and R v.1.0.44. Independent T tests were applied for continuous variables and Chi Square or Fisher’s exact test for categorical variables. Cox proportional hazard model was used to analyze whether certain variables were associated with increased odds of AF. A p-value of <0.05 was considered statistically significant. Results: A total of 177 patients were enrolled and followed for 478 ± 179 days. AF was detected in 31% (n=55). Mean age ± SD (positive vs. negative AF on ICM) was 71.1 ± 12.0 vs. 68.6 ± 11.8, p=0.20. Age, gender, race, BMI, systolic, diastolic blood pressures, cholesterol and HBA1c were not significantly different between the two groups. Significant univariate predictors of AF included PR interval (hazard ratio [HR] 1.19 per 10 milliseconds, p < 0.001) and Left Atrial Volume Index (LAVI) (HR 1.04 per 1 mL/m 2 , p = 0.01). In multivariate analysis, both PR interval and LAVI remained significant predictors of AF (HR 1.3, p < 0.001; HR 1.05, p < 0.001, respectively). Using a risk-score formulated from these predictors, we found that patients in the top quartile had a five fold higher rate of AF detection compared to the lowest quartile. Conclusion: In summary, our atrial fibrillation detection rate was 31%. PR interval and Left Atrial Volume Index were found to be both independent and multivariate predictors of atrial fibrillation in embolic stroke patients with insertable cardiac monitor.