Abstract TMP58: A Simple Score to Assess the Risk of Paroxysmal Atrial Fibirillation in Acute Ischemic Stroke Patients

Abstract

Introduction: Paroxysmal atrial fibrillation (PAF) is an important risk factor for ischemic stroke and has similar risk to permanent atrial fibrillation (AF), but detecting PAF is challenging. Holter ECG has a low sensitivity. Continuous ECG monitoring and implanted event recorders are high cost and inconvenient for elderly patients. Identifying patients at a particularly high risk of PAF by using scores may represent a reasonable alternative. We aimed to elucidate a clinical profile of patients with PAF including those risk factors by multiple variable analysis and create a score to detect patients with PAF. Methods: Consecutive patients with acute ischemic strokes from 2010 to 2011 were prospectively analyzed. We excluded patients with permanent AF. All patients without permanent AF had 24-hour Holter ECG and ECG monitoring on the ward within 7 days. Collected data included demographic data, clinical data including history of irregular rhythm or antiarrhythmic agent and transthoracic echocardiography data. PAF was documented by review of medical history, baseline ECG, ECG monitoring and Holter ECG. Results: We studied 197 stroke patients and excluded 45 (23%) patients with permanent AF. Thirty-five (23%) of 152 patients had PAF. Univariate analyses revealed that older age (p = 0.008), arrhythmia history (p <0.0001), higher brain natriuretic peptide (BNP) (p <0.0001) and left atrial dilatation (p = 0.007) were significantly associated with documented PAF more than without documented AF. Arrhythmia history (OR 11.4, 95% CI 2.4-54.0), BNP ≥ 85 pg/ml (OR 24.2, 95% CI 6.8-86.1) and left atraial (LA) diameter ≥ 40 mm (OR 3.3, 95% CI 1.1-9.5) were significantly independent predictive factors of PAF by logistic regression analysis. We calculated a score (BNP [≥ 150 pg/ml:3 point, 149-85 pg/ml:2], arrhythmia history[yes:2] and dilatation of LA [≥ 40 mm:1]. The area under the ROC curve for the documented PAF group was 0.908 and a total score ≥ 3 had a sensitivity of 83% and a specificity of 87%. Conclusions: In acute ischemic stroke patients without permanent AF, patients with documented PAF had significantly more history of arrhythmia, higher BNP and LA dilatation than those without documented AF. The new score can be useful to identify individuals at high risk of PAF.