Does the CHA2DS2-VASc score reliably predict atrial arrhythmias? Analysis of a nationwide database of remote monitoring data transmitted daily from cardiac implantable electronic devices

Abstract

CHA2DS2-VASc is a validated score for predicting stroke in patients with atrial fibrillation (AF). The purpose of this study was to assess whether the CHA2DS2-VASc score can predict new-onset AF in a cohort of patients with a cardiac implantable electronic device (CIED) followed with remote monitoring. Using the database of the Home Monitoring Expert Alliance project, we selected 2410 patients with no documented AF who had received a CIED with diagnostics on atrial high rate episodes (AHREs). The primary endpoint was time to first day with cumulative AHRE burden ≥15 minutes, 5 hours, 24 hours, and ≥7 consecutive days. During a median duration of 24.1(11.5-42.9) months, the incidence of AHRE increased with increasing CHA2DS2-VASc. At 6 years, occurrence of ≥15-minute AHRE was 80.2% (CHA2DS2-VASc ≤1) vs 93.7% (CHA2DS2-VASc ≥5), whereas ≥5-hour AHRE incidence was 68.4% (CHA2DS2-VASc ≤1) vs 92.5% (CHA2DS2-VASc ≥5). Occurrence of ≥24-hour and ≥7-day AHREs also increased with increasing CHA2DS2-VASc: 9.1% and 3.9% (CHA2DS2-VASc ≤1) vs 40.4% and 28.7% (CHA2DS2-VASc ≥5), respectively. Adjusted hazard ratio for unitary CHA2DS2-VASc increase ranged from 1.09 (confidence interval 1.04-1.14; P<.001) with AHRE burden ≥15 minutes to 1.26 (confidence interval 1.11-1.42; P <.001) with AHRE burden ≥7 days. At receiver operating curve analysis, CHA2DS2-VASc ≥2 was estimated to predict persistent forms of AHREs with 95.8% sensitivity but 11.7% specificity at 3 years. CHA2DS2-VASc ≥5 had 77.0% specificity but 34.6% sensitivity. In a CIED population with no previous diagnosis of clinical AF, AHRE incidence increased with increasing CHA2DS2-VASc score. The association was stronger with longer AHREs, but the accuracy of CHA2DS2-VASc as AHRE predictor was moderate.