Clinical Question: What is the optimal duration of anticoagulation therapy following venous thromboembolism (VTE)?

Setting: Various (meta-analysis)

Study Design: Meta-analysis (randomized controlled trials)

Synopsis: The optimal duration of anticoagulation therapy following VTE remains controversial. The investigators searched multiple databases, including the Cochrane database; reviewed clinical trial Web sites; and did hand searches of reference lists. They included English language randomized controlled trials that reported the risk of recurrent VTE compared with the duration of anticoagulation tharapy in patients with an initial VTE event. The investigators excluded studies that reported outcomes only on high-risk patients. Two reviewers independently assessed articles for inclusion and exclusion criteria and study quality. A third reviewer adjudicated in cases of disagreement. Fifteen of 67 studies (a total of 5,596 patients) met inclusion criteria, and the overall quality of the individual studies was high. A formal evaluation for publication bias found minimal, if any, effect, and the overall findings of the multiple trials were consistent (homogenous).

Long-term (median = six months) versus short-term (median = 1.75 months) therapy significantly reduced the risk of recurrent VTE events (number needed to treat = 50; 95% confidence interval, 25 to 1,000). Major bleeding events occurred most often in the first month of treatment, and the overall risk was similar with long-versus short-term therapy. Limited data suggested that extending anticoagulation beyond six months resulted in decreasing benefit.

Bottom Line: The optimal duration of anticoagulation therapy following an initial VTE event is six months or more. The risk of a major bleeding event is most pronounced in the first month of treatment, and the rate is similar to short-term (three months or less) treatment. Because the magnitude of benefit appears to lessen beyond six months, physicians and patients should reassess individual risk and benefit profiles beyond this time frame. (Level of Evidence: 1a)