The optimal perioperative treatment of patients taking warfarin (Coumadin) long-term for the prevention of thromboembolism has not been established. The current strategy for patients undergoing a procedure with a high risk for bleeding is to stop warfarin and bridge patients with intravenous heparin or subcutaneous low-molecular-weight heparin. No studies have evaluated the bleeding risk in patients taking oral anticoagulation for high-risk procedures. Two studies have shown that the risk of bleeding with hip surgery when the patient took warfarin was not significantly greater than in control groups. Studies have shown that moderate-control anticoagulation (i.e., maintaining an International Normalized Ratio [INR] between 1.5 and 2.0) does not increase bleeding risk in hip or knee replacement surgery but still protects against thromboembolism. Larson and associates evaluated the effects of continuing warfarin therapy at a lower temporary INR (1.5 to 2.0) before high-risk procedures.

The study included 100 patients from university and Veterans Administration hospitals who were taking long-term oral anticoagulants and who required an invasive or surgical procedure. Before the procedure, participants’ INRs were adjusted to between 1.5 and 2.0. The surgical and thromboembolic risks were assessed using published instruments. INR was reduced by decreasing the warfarin dose by one half for seven days before the procedure and then adjusting it again a few days before the procedure. Heparin was used if the INR dropped below 1.5. Bleeding complications were defined as major if they caused death, interruption of the anticoagulation, or a hemoglobin drop of 3 g per dL (30 g per L) or more, or if an unexpected blood transfusion was required. Minor bleeding was defined as unexpected bleeding that did not meet the major bleeding criteria.

Patients who had a thromboembolic event within the past six months (62 patients) or who had a thromboembolic event after past surgery (11 patients) were assigned to the high-risk group. Fifty-eight percent of the surgical procedures were considered significantly invasive by published criteria. The mean INR was 2.1 one day before surgery, and 1.8 on the day of surgery and on the first postoperative day. Of 100 patients enrolled in the study, two developed major bleeding and four had minor bleeding. One patient developed a deep venous thrombosis. The one patient who died had a prosthetic heart valve, and his INR was not increased to the preoperative level after surgery.

The authors conclude that moderate-intensity anticoagulation therapy (INR 1.5 to 2.0) with warfarin is a feasible method for preventing thromboembolic events in high-risk surgical patients who are on long-term warfarin therapy. They add that this was an observational study and does not prove equality with or superiority to other anticoagulation strategies.