Background: The optimal duration of anticoagulation for a first episode of unprovoked venous thromboembolism (VTE) is uncertain. Methods for predicting risk for recurrence may identify low-risk patients who are less likely to benefit from prolonged anticoagulation.

Purpose: To synthesize evidence evaluating the value of D-dimer as a predictor of recurrent disease in patients who have stopped anticoagulant therapy after a first unprovoked VTE.

Data sources: The MEDLINE, EMBASE, CINAHL, and Cochrane databases were searched until March 2008 without language restrictions. The strategy was supplemented with manual review of reference lists and contact with content experts.

Study selection: Randomized, controlled trials or prospective cohort studies that measured D-dimer after anticoagulant therapy in patients who received at least 3 months of anticoagulant treatment of unprovoked VTE.

Data extraction: Two authors independently reviewed articles and extracted data.

Data synthesis: Seven studies, totaling 1888 patients with a first unprovoked VTE, were eligible for analysis. During 4500 person-years of follow up, annual rates of recurrent VTE differed statistically significantly: 8.9% (95% CI, 5.8% to 11.9%) in patients with positive D-dimer results and 3.5% (CI, 2.7% to 4.3%) in patients with negative D-dimer results.

Limitation: The duration of anticoagulation, timing of D-dimer testing, and D-dimer assay varied across studies.

Conclusion: In patients who have completed at least 3 months of anticoagulation for a first episode of unprovoked VTE and after approximately 2 years of follow-up, a negative D-dimer result was associated with a 3.5% annual risk for recurrent disease, whereas a positive D-dimer result was associated with an 8.9% annual risk for recurrence. These rates should inform decisions about the balance of risks and benefits of prolonging anticoagulation.