A feasibility study of continuing dose-reduced warfarin for invasive procedures in patients with high thromboembolic risk.

BACKGROUND: The management of perioperative anticoagulation therapy for patients having a high risk of thromboembolism who are receiving long-term oral anticoagulant therapy is uncertain. The prevalent approach is to discontinue oral anticoagulation therapy and initiate heparin therapy. Another potential strategy is to continue oral anticoagulation therapy with a temporary adjustment of warfarin intensity to a preoperative international normalized ratio (INR) of 1.5 to 2.0. Such moderate-dose anticoagulation therapy with warfarin has been shown to be hemostatically safe yet effective in the prevention of thromboembolism after hip or knee replacement.
METHODS: Over an 11-year period (ie, 1993 to 2003), our hemostatic consultative service prospectively identified 100 consecutive patients for whom we continued warfarin therapy at adjusted doses during the perioperative period, targeting a goal for the INR of 1.5 to 2.0. Patients were assigned a score for venous thromboembolic risk as well as overall surgical risk using published instruments. Score assignment was based on what was deemed to be extremely high risk for thromboembolism in patients who were receiving long-term warfarin therapy. Although the patients were accrued prospectively, the final retrospective analysis was made after all patients were treated.
RESULTS: The most common indication (62%) for high-risk assignment was a thromboembolic event within the past 6 months. The second most prevalent reason was prior postoperative venous thromboembolism (VTE) [11%]. Indications for long-term anticoagulation therapy were recent VTE (62%), inherited thrombophilia (7%), antiphospholipid syndrome (13%), mechanical heart valves (18%), and prior cerebrovascular accident (4%). The prevalence of inherited thrombophilia probably has been grossly underestimated, as neither factor V Leiden mutation nor prothrombin 20210 mutation had been described during the bulk of the accrual time. Most surgical procedures (58%) were significantly invasive (Johns Hopkins category 3 to 5). The mean INR values were 2.1 on the day prior to surgery (SD, 0.9594; range, 1.2 to 6.5; n = 65), 1.8 on the day of surgery (SD, 0.4899; range, 1.2 to 4.9; n = 75), and 1.8 on the first postoperative day (SD, 0.4436; range, 1.1 to 3.3; n = 70). Two patients had major bleeding, and four patients had minor bleeding. One patient developed deep venous thrombosis. Several weeks after surgery, one patient with a prosthetic heart valve died from an embolic stroke, which was associated with a failure to increase his anticoagulation to therapeutic levels.
CONCLUSIONS: Moderate-intensity anticoagulant therapy with warfarin, targeting a goal INR of 1.5 to 2.0, appears to be a safe and feasible method for preventing thromboembolic complications in high-risk surgical patients who are receiving long-term oral anticoagulant therapy. This may be considered a reasonable method to afford thromboprophylaxis in highly selected patients who are occasionally seen in clinical practice. This observational study does not prove equality, let alone superiority, to other proposed methods of anticoagulation therapy.