Abstract Background Stroke and bleeding risks in atrial fibrillation (AF) are often assessed at baseline, as a “one-off” evaluation. However, these baseline values are usually applied to predict outcomes many years later, and therefore lack the consideration that the risk is not static. Purpose Our aim was to investigate if dynamic changes of CHA2DS2-VASc and HAS-BLED over time have an effect on the prediction of stroke and bleeding risks. Methods We included AF patients who were stable while taking vitamin K antagonists (INR 2.0–3.0) for 6 months attending a tertiary hospital (May 2007-December 2007). During 6-years of follow-up, ischemic strokes/transient ischemic attacks (TIAs), major bleeds, and all-cause deaths were recorded. CHA2DS2-VASc and HAS-BLED were recalculated every 2-years, and their predictive abilities were tested for outcomes in periods of 2-years (from year 0 to 2, year 2 to 4 and year 4 to 6). Results 1361 patients (693 [50.9%] females, median age 76 [IQR 71–81] years, mean CHA2DS2-VASc and HAS-BLED of 4.0±1.7 and 2.9±1.2, respectively) were included. The predictive ability for ischemic stroke/TIA of the baseline CHA2DS2-VASc for 2-years events was 0.662 (0.637–0.688, p<0.001). Compared to the baseline CHA2DS2-VASc, the CHA2DS2-VASc re-calculated at 2-years presented significantly higher predictive ability for ischemic stroke/TIA during the period 2–4 years (c-indexes: 0.701 [0.675–0.727] vs. 0.604 [0.576–0.631], p<0.001). Integrated discrimination improvement (IDI) and net reclassification improvement (NRI) showed an improvement in sensitivity of 0.014 (p<0.001) and a better reclassification (0.677, p<0.001). Similarly, the CHA2DS2-VASc re-calculated at 4-years yielded significantly better predictive performance for ischemic stroke/TIA during the period 4–6 years in comparison to the baseline CHA2DS2-VASc (c-indexes: 0.761 [0.734–0.786] vs. 0.682 [0.653–0.710], p=0.026). Again, IDI reported an improvement (IDI = 0.030, p<0.001) and there was an important enhance of the reclassification ability (NRI = 0.757, p<0.001). The c-index of the baseline HAS-BLED for events at 2-years was 0.744 (0.720–0.767, p<0.001). At 2-years, the re-calculated HAS-BLED score showed higher predictive ability compared to the baseline HAS-BLED during the period 2–4 year (c-indexes: 0.709 [0.680–0.738] vs. 0.663 [0.632–0.693], p=0.003). Accordingly, IDI and NRI demonstrated significant improvements for the re-calculated HAS-BLED compared to baseline (IDI = 0.016, p=0.001; NRI = 0.444, p<0.001). For major bleeding during the period 4–6 years, the c-index of the HAS-BLED score re-calculated at 4-years was non-significantly different to baseline HAS-BLED at baseline (0.631 [0.601–0.660] vs. 0.623 [0.593–0.652], p=0.751), although showed a slight enhance in sensitivity (IDI = 0.009, p=0.018). Conclusions In AF patients, stroke and bleeding risks are dynamic and change over time. The CHA2DS2-VASc and HAS-BLED scores should be regularly reassessed. Funding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): This work was supported by the Spanish Ministry of Economy, Industry, and Competitiveness, through the Instituto de Salud Carlos III after independent peer review (research grant: PI17/01375 co-financed by the European Regional Development Fund)
Read full abstract