Abstract 223: Impact of Differences in Once- vs Twice-Daily Medications Adherence on the Risk of Bleed and Stroke in Non-Valvular Atrial Fibrillation: Analysis of Randomized Trials and Claims Data Sources

Background: In non-valvular atrial fibrillation (NVAF) patients, those receiving once-daily (QD) versus twice-daily (BID) non vitamin-K antagonist oral anticoagulants (NOACs) may have better medication adherence. The impact on stroke and bleed risk is not known. Objective: To estimate the impact of adherence differences between QD vs BID therapies on bleed and stroke risks in NVAF patients. Methods: The relation between adherence (proportion of days covered [PDC]) for QD vs BID NOACs and one year bleed risk was modeled using claims data from Truven Health Analytics MarketScan databases (7/2012-10/2015). Next, the relation between adherence and bleeding was calibrated to match that seen in the placebo and NOAC arms of previous randomized controlled trials (RCTs). Finally, we used adherence rates for QD (PDC=0.849) and BID (PDC=0.738) cardiovascular medications from a meta-analysis (Coleman et al.). These rates were used in the calibrated model to estimate bleeds. An analogous method was applied to evaluate the impact of QD vs BID adherence on stroke risk. Results: The relation between PDC and risks of bleed and stroke was modeled using claims data (N=65,022) and calibrated using RCTs. In the calibrated model, compared with BID dosing, QD dosing was associated with 81 fewer strokes (34% reduction) and 14 more bleeds (6% more) per 10,000 patients/year (Figure). Conclusion: Among NVAF patients, better adherence to QD dosing was associated with a significantly lower stroke risk of QD but similar risk of bleed.