Excessive prolongation of the international normalized ratio (INR) occurs frequently in patients taking warfarin; in fact, about one in six INR values is above the desired range. Excessive prolongation of the INR is clinically important because the risk of bleeding approximately doubles for each one point increase in the INR beyond the usual therapeutic range. Thus, treatment strategies which rapidly and reliably lower an excessively prolonged INR into the desired range have the potential to reduce bleeding. When taken by patients with INR values between 4.5 and 10, a small dose of oral vitamin K (1 mg to 2.5mg) reduces the INR into the desired INR range in about 75% of cases within 24 hours of its administration. If warfarin is simply withheld, and no vitamin K is given, about 25% of patients will have an INR in the desired range at 24 hours. However, vitamin K is rarely given to such patients. In a recent survey carried out by our group, less than 20% of such patients would have been given low dose oral vitamin K by a group of physicians who regularly supervise warfarin therapy.

The most common treatment for excessive prolongation of the INR is to simply withhold warfarin and allow the INR to fall into the therapeutic range. Although this strategy is effective its safety has never been adequately examined. In fact, recent evidence suggests that patients with INR values of more than 6.0 who are treated with simple warfarin withdrawal have a risk of major bleeding of 4% in the two weeks after they develop their prolonged INR.

When asked why they did not give oral vitamin K to a non-bleeding patient who has an excessively prolonged INR, physicians generally give one of three reasons: (1)They are not convinced that oral vitamin K reduces bleeding. (2) They are concerned that oral vitamin K may cause thrombosis. (3) In contrast with simply withholding warfarin, giving oral vitamin K requires a patient to visit the physician, and the physician must have a supply of vitamin K.

The investigators hypothesize that the routine practice of not administering oral vitamin K to patients with excessively prolonged INR values is causing patients to have major, life-threatening and fatal bleeds. To convince physicians that oral vitamin K should be administered to all non-bleeding patients with INR values of more than 4.5, the investigators propose a study which the investigators anticipate will demonstrate that oral vitamin K reduces bleeding, does not cause thrombosis, and can be administered at home without direct physician supervision.

To accomplish these goals, the investigators propose a multinational, double-blind, placebo-controlled trial. The investigators will randomize patients with INR values between 4.5 and 10.0 to receive 1.25 mg of oral vitamin K or placebo and follow them for bleeding and thrombosis. Patients with INR values of more than 10.0 will receive a single 1.25 mg dose of oral vitamin K.

Successful completion of this study will establish a treatment standard supported by clinical data which will, in turn, change the way that patients taking warfarin who present with an excessively prolonged INR are treated.