LAUSR

We are grateful to Ayubi and Safiri for their comments for our recent article [1]. In this large observational study, we investigated the performance of the CHA2DS2-VASc score in predicting ischemic bowel disease (IBD) in a nationwide cohort consisting of 212,058 patients with atrial fibrillation (AF).We used the CHA2DS2-VASc score to classify the AF patients and found that AF patientswith a score of ≧2 had an adjusted hazard ratio of 3.35 (95% CI= 2.71–4.13) for IBD, compared to those with a score of b2. Our study highlights both the advantages and the limitations of using the CHA2DS2-VASc score.Well known thrombotic risk factors are also used to estimating the CHA2DS2-VASc score. It is not surprising to find an association between cardiovascular risk factors captured in the CHA2DS2-VASc score and the risk of IBD for AF patients. However, the discriminative power of the score was poor, with a Cstatistic of 0.56 (95% CI = 0.55–0.57). We agreed that a prediction model with internally or externally validated is necessary. We further performed an internal validity analysis by bootstrapping a sample of 1000 and have a measured C-statistic of 0.54 (95% CI = 0.53–0.56).