Outcomes in non-valvular atrial fibrillation patients with good and poor INR control after switching to direct oral anticoagulants

Document Type

Conference Proceeding

Publication Date


Publication Title

Res Pract Thromb Haemost


Background: Several international guidelines recommend direct oral anticoagulants (DOACs) over vitamin K antagonists (VKAs) for stroke prevention in non-valvular atrial fibrillation (NVAF), and warfarin patients with poor INR control are commonly viewed as ideal patients for switching to DOACs. Aims: To examine outcomes in NVAF patients with good and poor INR control before and after switching from warfarin to a DOAC. Methods: Since 2009, abstractors at 6 anticoagulation services have entered patient data into the Blue Cross Blue Shield of Michigan funded Michigan Anticoagulation Quality Improvement Initiative (MAQI2) registry. NVAF patients in the MAQI2 database were stratified based on time in therapeutic range (TTR): good control (TTR≥60%) and poor control (TTR<60%). Ischemic stroke and major bleeding rates were calculated for both groups before and after switching to a DOAC. Comparisons were made using Poisson and Chi-square tests and Cox's Proportional Hazard Modeling. Results: 413 NVAF patients were identified with 212 (51%) having good INR control and 201 (49%) having poor INR control. After switching to a DOAC, patients with good INR control had a significantly higher ischemic stroke rate (2.08 vs. 0.98 per 100 pt-yr, P<0.001, HR 2.33) and major bleed rate (4.2 vs. 2.4 per 100 pt-yr, P=0.04, HR 2.15) than while on warfarin. Patients with poor INR control had similar ischemic stroke rates (1.67 vs. 1.88 per 100 pt-yr, P=0.30, HR 0.79) and major bleed rates (6.7 vs. 5.0 per 100 pt-yr, P=0.14, HR 1.39) before and after switching. The composite endpoint of ischemic stroke and major bleeding showed a similar pattern (Table 1). Conclusions: In our MAQI2 patients with good INR control switched to DOACs, the increased convenience came at a price: double the risk of ischemic stroke and major bleeds. Our poorly controlled warfarin patients, who are frequently viewed as ideal candidates, showed no clinical benefit from switching.





First Page


This document is currently not available here.