Steady time in therapeutic range is important for optimal care when treating atrial fibrillation patients using warfarin. However, research published online Sept. 2 in Circulation: Cardiovascular Quality and Outcomes insists that stability, seen through the international normalized ratio, not be overlooked.

In fact, researchers found that even in those patients where high time in therapeutic range (TTR) was noted, if coagulation control was low, outcomes were still poor.

To determine outcomes against TTR and variability in international normalized ratio (INR), Zayd Razouki, MD, of Bedford VA Medical Center in Bedford, Mass., and colleagues gathered data on atrial fibrillation patients receiving warfarin anticoagulation therapy from a combined Veterans Administration-Medicare database.

A total of 40,404 eligible atrial fibrillation patients were using warfarin for anticoagulation therapy between October 2006 and September 2008.

They then assessed the data using three different models adjusting how INR and TTR related to each other and to outcomes for major bleeding, fatal bleeding and ischemic stroke. They found most frequently that the patients with the worst anticoagulation control, including high INR instability and low TTR, had the poorest outcomes

Per 100 person-years, Razouki et al found 11.95 major hemorrhages, 6.28 ischemic strokes and 1.78 fatal hemorrhages in patients with low TTR. Patients with high variability in INR, 15.37 major hemorrhages, 7.98 ischemic strokes and 2.56 fatal hemorrhages per 100 person-years.

While individually, INR and TTR were indicators of poor outcomes, however when taken together, INR variability became the greater indicator of outcomes. Where variability of INR was high, patient TTR could be moderate to high and outcomes would still be poor. Patients with moderate TTR and high INR variability had a hazard ratio of 1.27 for major bleeding, 1.29 for fatal bleeding and 1.7 for ischemic stroke.

They noted no statistical difference in the risk of ischemic stroke, major or fatal bleeding between moderate and high TTR with high INR variability. They did note however, that when there was low TTR, stability of INR had no effect.

Razouki et al wrote that this data suggests that INR variability should be considered by clinicians when assessing anticoagulant effectiveness against risk. However, they did note that agreement of how INR would be assessed and calculated was needed.

Consensus and proper software, they wrote, were needed to bring these findings into clinical practice.