Reply to letter to the editor: "methodological considerations in non-sustained Atrial Fibrillation and stroke risk"

Abstract

We would like to thank the authors 1 for their thoughtful comments on our article 2 entitled "Nonsustained Atrial Fibrillation in Ambulatory ECG Recording and Thromboembolic Events in Long-Term Follow-Up."We would like to clarify that the number "163" stated in the first sentence of the Results section was a typographical error; the correct number of patients with nonsustained atrial fibrillation (NS-AF) included in the study was 133, and this has been corrected in the published version of the article.After propensity score matching, 20 cases were excluded, resulting in 113 patients in the NS-AF group and 113 in the control group for the subsequent analyses.Regarding the statistical approach, we did not perform Cox proportional hazards or competing risk analyses because the exact timing of stroke events could not always be determined with sufficient precision from hospital records and telephone interviews.Given this limitation, binary logistic regression was selected as the most appropriate approach to assess the association between NS-AF and stroke.Finally, although age remained statistically different between groups after matching, age was included in the multivariate model, and NS-AF persisted as an independent predictor of ischemic stroke, suggesting that the association is not solely explained by age.We appreciate these constructive remarks and believe that our clarifications address the methodological concerns and improve the interpretability of our findings.