Risks Associated with Temporarily Stopping Warfarin Therapy

From McMaster University and Hamilton Civic Hospitals Research Centre, 711 Concession St., Hamilton, ON L8V 1C3, Canada, where reprint requests should be addressed to Dr. Kearon. ©1997, Massachusetts Medical Society. HE most common indications for warfarin therapy are atrial fibrillation, the presence of a mechanical heart valve, and venous thromboembolism. 1,2 Treatment with warfarin presents a problem if patients with these indications need surgery, because the interruption of anticoagulant therapy increases the risk of thromboembolism. After warfarin therapy is discontinued, it takes several days for its antithrombotic effect to recede, and when it is resumed, several days are needed to reestablish therapeutic anticoagulation. There is no consensus on the appropriate perioperative management of anticoagulation for patients who have been receiving long-term warfarin therapy. Rational decisions about the treatment of such patients can be made only if one can quantify the risks of thrombosis and bleeding associated with the various alternatives. In this review, we will consider the expected risks and benefits of different approaches to anticoagulation in patients who require warfarin because of atrial fibrillation, a mechanical heart valve, or a history of venous thromboembolism. Our assessment of the consequences of arterial and venous thromboembolism and postoperative bleeding is then used as the basis for an approach to management designed to maximize the patient’s safety and the efficient use of health care resources. We quantified the estimated risks and benefits of two different strategies: an aggressive approach, in which intravenous heparin is given for two days before and two days after surgery; and a minimalist strategy, under which patients receive no heparin immediately before or after surgery. These two approaches were chosen because they are widely used in clinical practice, conceptually clear, and likely to T be associated with the most divergent levels of risk of thromboembolism and bleeding. Whenever possible, our estimates of risk and benefit are based on data from randomized trials or prospective studies.