Optimal International Normalized Ratio Range for Asian Patients on Anticoagulation Therapy

Introduction: Despite the recent emergence of direct oral anticoagulants, the use of warfarin will continue in patients on warfarin with stable control, those with mechanical heart valves, antiphospholipid syndrome and significant renal impairment. Warfarin therapy is challenging because of its narrow therapeutic window. The aim is to administer the lowest effective dose to maintain the target International Normalized Ratio (INR) in order to minimize bleeding and thromboembolic complications. The optimal INR of 2.0-3.0 has been validated in Western populations but not in Asians. It is common belief that Asians on warfarin having more bleeding complications and results of retrospective small scale studies conducted in Asian populations suggested a lower therapeutic INR range. Hence a lower target INRs had been advocated.¹

Objective: To determine the optimal INR range associated with the lowest thromboembolism and bleeding rates for a large multi-ethnic Asian population receiving warfarin therapy in Singapore.

Methods:A retrospective cohort study was conducted among adult patients receiving warfarin from Singapore General Hospital and National Heart Centre Singapore, between 1 January 2010 and 31 December 2014. These patients were closely followed up with tight reporting and recording of adverse events. Medical records from clinics, hospitalisation and emergency room visits in both centres were analysed for warfarin-related thromboembolism and bleeding events and INR values at time of events. Thromboembolic events included ischemic stroke, deep vein thrombosis (DVT) and pulmonary embolism (PE). Bleeding events were differentiated into major and minor bleeds according to the International Society on Thrombosis and Haemostasis criteria.

Results: A total of 1216 patients (650 males, 566 females) with 12,784 patient-years were included in the study. Racial distribution was 79.2 % Chinese, 12.6% Malays, 6.6% Indians and 1.6% others. 89.2% of the patients were above 50 years old, with 48.8% aged between 51 to 70 years and 40.3 % aged between 71 to 100 years. Indications for warfarin were atrial fibrillation (57.3%), prosthetic heart valve replacement (16.9%), intracardiac thrombus (5.4%), venous thromboembolism (16.6%), ischaemic strokes (1.6%), rheumatic heart disease (1.4%) and others (0.8%).

There were 1776 thromboembolic and bleeding events. The event rates were 2.7 per 100 patient-years for major bleeds, 3.8 per 100 patient-years for thromboembolic events and 7.4 per 100 patient-years for minor bleeds. Using the Poisson Regression analysis, the INR range of 2.01-2.60 was found to correspond with a lowest combined major bleeding and thromboembolism risk of 2.9% (figure 1). When analysed to obtained a bootstrapped (n=1,000) estimate of the rate of events by 1-unit INR ranges, given a usual "good" time in target range of 60% achieved by our anticoagulation clinic, we found that the 2.0-3.0 INR range had the minimum rate of major bleed and thromboembolism events (figure 2), accounting for uncertainty in actual INR values at adverse events. This finding is likely to be further reinforced if one accounts for uncertainty in the actual rate of events by INR range.

Conclusion:The major bleeding rate of 2.7 per 100 patient years in this Asian population is consistent and not higher than reports in recent western population cohort studies^2,3.The optimal INR range for Asian patients on anticoagulation therapy in Singapore was 2.01-2.60 by the Poisson Regression analysis. However, such a narrow target INR range will be challenging to implement in practice. Except for those at extreme risks of bleeding or thrombosis, the data from our bootstrapped analysis supports the use of the recommended INR range of 2-3 for Asians. Thus while Asians may have an increased sensitivity to warfarin in requiring lower dosages, the target therapeutic INR range of 2-3 remains.

References:

1.Circulation Journal 2011; 75: 2719 - 2721

2.British Medical Journal 2016;353:i3189

3.JAMA Cardiology 2016;1(2):172-180

View largeDownload slide

View largeDownload slide

Close modal