Oral Vitamin K Versus Placebo to Correct Excessive Anticoagulation in Patients Receiving Warfarin: A Randomized Trial

Abstract

Conclusion: Bleeding is not reduced by low-dose vitamin K therapy in warfarin recipients with international normalized ratios (INRs) of 4.5 to 10.0. Summary: Bleeding risk with warfarin therapy increases when the INR is >4.5 (Chest 2008;133:160-98s). In addition, INRs outside the therapeutic range can occur up to one-third to one-half of the time (Chest 2005;127:1515-22). When patients present with an INR >4.5, clinicians may elect to treat with vitamin K therapy or just withhold warfarin treatment and allow the INR to drift down over time. The authors hypothesized patients with INRs >4.5 treated with vitamin K therapy would have less bleeding complications than those in which warfarin therapy was merely withheld until the INR returned to therapeutic levels. They tested their hypothesis by performing a multicenter randomized trial in nonbleeding patients with INRs of 4.5 to 10.0. Patients were randomized to receive oral vitamin K (1.25 mg) or placebo. The primary outcome was the frequency of all forms of bleeding during the first 90 days. Patients' primary care givers and those assessing outcomes were blinded to treatment assignment. Thromboembolism and death were also monitored. There were 335 patients assigned to oral vitamin K therapy, and 347 were analyzed. There were 369 patients randomly assigned to placebo, and 365 were analyzed. At least one bleeding complication occurred in 56 patients (15.8%) in the vitamin K group and in 60 patients (16.3%) of the patients in the placebo group, for an absolute difference of −0.5 percentage points (95% confidence interval [CI], −6.1 to 5.1 percentage points). Major bleeding events occurred in nine patients (2.5%) of the vitamin K group and in four patients (1.1%) of the placebo group (absolute difference, 1.5 percentage points; 95% CI, −0.1 to 3.7 percentage points). Four episodes of thromboembolism (1.1%) occurred in the vitamin K group and three episodes (0.8%) in the placebo group (absolute difference, 0.3 percentage points; 95% CI, −1.4 to 2.0 percentage points). One day after treatment, INR had decreased by 2.8 in the vitamin K group and by 1.4 in the placebo group (P < .001). Conclusion: The data indicate that active reduction of the INR by administration of vitamin K does not reduce bleeding episodes. The data also indicate administration of vitamin K does not result in increased thromboembolism. The results have significant implications for clinical practice. In patients with INRs between 4.5 and 10, warfarin therapy should be withheld and reinstituted once the INR has returned to the desired range. The data cannot be extrapolated to patients who are bleeding or who present with INRs >10. Overall, the results of the trial should help simplify the management of most patients treated with warfarin who present with supratherapeutic INR levels.