Abstract WP200: Development and Validation of a Score to Detect Paroxysmal Atrial Fibrillation During Long-term Holter-monitoring After Acute Ischemic Stroke

Abstract

Introduction: Atrial fibrillation remains one of the most common causes of ischemic stroke. The diagnosis of paroxysmal atrial fibrillation (pAF) frequently escapes routine diagnostic due to its intermittent and asymptomatic occurrence. Currently prolonged monitoring times (72h) are recommended, but did not find their way into everyday practice. Therefore an individual patient selection for prolonged ECG-monitoring might increase the diagnostic yield of pAF in a resource-saving and cost-effective manner. Methods: We used individual patient data from three prospective studies (n total =1556) which performed a prolonged Holter ECG-monitoring (at least 72 h) and centralized data evaluation after ischemic stroke (IDEAS 72h monitoring, Find-AF 7 days and Find-AF randomised 3 times 10 days monitoring) in patients presenting with sinus rhythm. The score was developed on the IDEAS cohort and internally (bootstrapping) and externally (Find-AF and Find-AF randomised data) validated following the TRIPOD guideline. Results: pAF was detected in 77 of 1556 patients (49 (4.3%) IDEAS, 20 (9.0%) Find-AF, 8 (4.0%) Find-AF randomized ). After logistic regression analysis with variable selection, age and the qualifying stroke event (categorised as stroke with NIH-SS <=5 (OR 2.4; 95%CI 0.8-6.9, p=0.112) or stroke with NIH-SS >5 (OR 7.2; 95% CI 2.4-21.8, p<0.001) were found to be predictive for the detection of pAF within the prolonged Holter monitoring and were included in the final score ( A ge, S troke S everity: NIHSS > 5 , ASS5 ). The score was shown to keep its discriminative performance (IDEAS, AUC 0.76) on the external validation cohort (Find-AF, Find-AF-randomized, AUC 0.75). When ASS5-score and CHADS2-score were applied to classify patients into high and low-risk patients according to the score’s particular Youden-index, the ASS5-score was statistically significant better in discriminating between high- and low-risk patients than the CHADS 2 -score (NRI 0.22 with p=0.047; ROC-AUC 0.75 for ATISS5 and 0.61 for CHADS 2 -Score with p = 0.0032). Conclusion: ASS5-score can be used by clinicians to select patients for prolonged ECG-monitoring after ischemic stroke to detect pAF.