Bleeding risk factors in chronic oral anticoagulation with acenocoumarol

We studied major bleeding complications, death related to hemorrhage, and t ried to identify predisposing factors for bleeding in outpatients treated w ith acenocoumarol. We evaluated 811 outpatients attending a specialized ant icoagulant therapy unit. The intended INR range was 3.5-4.5 for mechanical heart valve replacement (N = 384) and 2.0-3.0 for other indications (N = 42 7). The variability of INR for the total follow-up and the 2 months before the hemorrhage was calculated. The total follow-up was 1,963.26 years with 27,321 control tests. We observed 47 major bleeding episodes, including 2 f atal (central nervous system hemorrhages), in 37 patients. 49.5% of the pat ients had underlying diseases. The rate of major and fatal hemorrhage was 2 .39 and 0.10 episodes per 100 patients year, respectively. Hemorrhagic comp lications were more frequently observed in patients with a more intense int ended range (8.2% in the INR 3.5-4.5 group vs. 1.5% in the 2.0-3.0 INR grou p), The risk of major bleeding increased in patients with an achieved INR h igher than 6 and in those with higher INR variability during follow-up. The estimated probability of bleeding also increased with time: it was 0.102% st 78 months, and at the beginning of therapy it was 0.006% and 0.007% at 1 and 4 months, respectively. The intensity of anticoagulation and the devia tion of the INR from the target are the most important risk factors for ble eding in patients taking acenocoumarol. Monitoring the variability of INR c an help identifying patients predisposed to bleeding. However, the screenin g for underlying disease should always be performed, Am, J, Hematol. 63:192 -196, 2000, (C) 2000 Wiley-Liss, Inc.