HAS-BLED Predicts Warfarin Control in Australian Patients treated for Deep Vein Thrombosis

Abstract

The HAS-BLED model is widely utilized to assess patients' bleed risk prior to anticoagulant therapy including warfarin. Some of the variables assessed in the model are also known to influence warfarin control, commonly measured by time in therapeutic range (TTR). The aim of the study was to determine whether the HAS-BLED risk tool is a good predictor of bleed risk and warfarin control in deep vein thrombosis (DVT) patients. Retrospective data were collected for DVT warfarin care patients at Sullivan Nicolaides Pathology. Data included age, medical history and concurrent drug therapy to calculate HAS-BLED scores. INR results were used to calculate TTR with the Rosendaal method and mean TTR used for analysis and comparison. The eligible 533 patients had a mean TTR of 78.3%. Categorization according to HAS-BLED score resulted in 150 patients classified as low-risk, 331 as moderate-risk and 52 as high-risk with a haemorrhagic incidence per patient of 0.08, 0.53 and 0.54, respectively. Patients in the low-, moderate- and high-risk HAS-BLED categories had a mean TTR of 81%, 79% and 65%, respectively, with significant differences (p < 0.001) found in TTR between the low- and high-risk and moderate- and high-risk categories. In an Australian DVT population, the HAS-BLED score accurately predicts decreasing warfarin control with increasing risk category, and patients with scores ≥3 achieve poor control as indicated by a TTR <70%. In addition to predicting bleed risk, the HAS-BLED tool may also predict the potential benefit of warfarin treatment and hence influence choice of anticoagulant therapy.