Comparisons of Changes in the Adapted Diabetes Complications Severity Index and CHA2DS2-VASc Score for Atrial Fibrillation Risk Stratification in Patients with Type 2 Diabetes Mellitus: A Nationwide Cohort Study

Abstract

Purpose: This study describes the risk prediction of atrial fibrillation (AF) after incident type 2 diabetes mellitus (DM) with either progression of adapted diabetes complications severity index (DCSI) or CHA2DS2-VASc score in a large registry from Taiwan. MethodsThe authors performed a retrospective nationwide cohort study by analyzing a Longitudinal Health Insurance Dataset, observing the ability of dynamic adapted DSCI and CHA2DS2-VASc score for AF risk discrimination in type 2 diabetic patients. The predictive performance of changes in the adapted DCSI and CHA2DS2-VASc score with regard to AF events was assessed using area under the curve of receiver operating characteristics (AUROC); and the difference between them was examined using the Delong test. ResultsA total of 81,655 new-onset type 2 DM patients were included in the study cohort. Among them, the AUROC for the adapted DCSI change in predicting AF (0.79, 95% CI = 0.78-0.80) was significantly higher than the change in CHA2DS2-VASc score (0.63, 95% CI = 0.62-0.64) with the DeLong test P < 0.001. Conclusions: Adapted DCSI change significantly outperforms the progression of CHA2DS2-VASc score with regard to AF prediction in type 2 diabetic patients.