Abstract 14012: Opportunities to Improve the Efficacy and Safety of Oral Anticoagulant Therapy in Atrial Fibrillation—Insights From a Multistate Healthcare System

Document Type


Publication Date


Publication Title





Introduction: Vitamin K antagonists (VKAs) effectively reduce thromboembolic risk in atrial fibrillation (AF), but are limited by a narrow therapeutic window. Patients with reduced time in the therapeutic range (TTR) also face an increased risk of bleeding and ischemic events. Based in part on this, current guidelines give preference to direct-acting oral anticoagulants (DOACs) over VKAs in AF.

Hypothesis: We hypothesized that DOACs are underutilized among those on oral anticoagulant therapy and that TTR remains suboptimal for large numbers of individuals on VKAs in real-world settings.

Methods: We sought to evaluate a) the breakdown of OAC type and b) TTR for those on VKAs in an ambulatory population of at-risk AF patients within a large multistate healthcare system. EHR and coding (ICD-10) data were used to identify patients with AF, calculate their CHA2DS2-VASc score, and define their current antithrombotic regimen. For those on a VKA, TTR was assessed with the Rosendaal method and reported as mean values. Demographics were assessed to allow for comparison between those receiving a DOAC and a VKA, as well as, those with high (>70%) vs. low (<70%) TTR. Chi square or Fisher exact tests were used to examine differences between groups.

Results: Data was pulled from our EHR on 8/1/18, identifying 147,455 unique patients with AF, of which 102,728 (76.3%) had a CHA2DS2-VASc score >2 (excluding female gender). Among these at-risk patients, 61,698 (60.1%) were receiving an OAC, of which 47.8% were on a VKA and 52.2% were on a DOAC. The mean TTR was 56.3%, with 37.1%, 49.9% and 60.8% with TTRs >70%, >60%, and >50%, respectively. Patients on a DOAC were more likely to be female and less likely to have heart failure, coronary artery disease, peripheral vascular disease, diabetes and renal disease (p<0.0001 for all). Patients with a TTR >70% were more likely to be male and less likely to have heart failure, diabetes, and renal disease (p<0.05 for all).

Conclusions: In a contemporary, non-registry setting, VKAs continue to be used in nearly half of at-risk patients on an OAC for AF, with a suboptimal TTR in nearly two thirds. Further investigation is needed into tools that facilitate interchange from a VKA to a DOAC, particularly among those with a suboptimal TTR.

Clinical Institute

Cardiovascular (Heart)




Center for Cardiovascular Analytics, Research + Data Science (CARDS)