The predictive value of left atrioventricular coupling index for ILR-detected Atrial Fibrillation following an Embolic Stroke of Undetermined Source in patients with sinus rhythm

Abstract The prediction of atrial fibrillation in high-risk populations is an important research area of modern atrial fibrillation care. One such population is the embolic stroke of undetermined source population, where appropriate anticoagulation might help to reduce the annual stroke recurrence risk of 5%. In this study, individuals were categorised according to their tertile of left atrioventricular coupling index, a composite echocardiographic parameter that reflects both left atrial and ventricular size. Its association with the detection of implantable loop recorder detected sub-clinical atrial fibrillation was investigated. A retrospective single-centre cohort study was performed. A total of 296 embolic stroke of undetermined source patients, referred for implantable loop recorder, who also underwent echocardiography were screened. Of these, 230 cases had echocardiograms of sufficient quality to measure left atrioventricular coupling index. Patients were categorised according to their left atrioventricular coupling index tertile, with tertile 1 being less than 16.97, tertile 2 being 16.97-23.9, and tertile 3 being greater than 23.9. Patients in the third left atrioventricular coupling tertile were more likely to develop atrial fibrillation of any duration, with a 1.9 fold increased univariable hazard ratio though this association was not significant in longer duration of atrial fibrillation episode. This effect was no longer significant with stepwise multivariable analysis, suggesting that age was primarily mediating the effect seen in this study. Left atrioventricular coupling index was not an independent predictor of future implantable loop recorder-detected atrial fibrillation after adjustment for age in this cohort.