Recurrence after unprovoked venous thromboembolism

In the linked meta-analysis (doi:10.1136/bmj.d813), Douketis and colleagues provide additional evidence that the risk of recurrent thrombosis is higher in men with unprovoked venous thromboembolism (VTE) than in women.1 They conducted a patient level meta-analysis using data from seven prospective studies of 2554 patients with first time VTE, which aimed to investigate an association between D-dimer concentration and long term recurrence of VTE. The three most important findings were that men with unprovoked VTE had a higher risk of recurrence than women with unprovoked VTE (hazard ratio 2.2, 95% confidence interval 1.7 to 2.8). This association persisted even after excluding women with hormone associated initial VTE (1.8, 1.4 to 2.5). No difference was found in recurrence of VTE between men and women with an initial provoked event (1.2, 0.6 to 2.4), and women with initial hormone associated VTE had a lower risk of recurrence than women with unprovoked VTE (0.5, 0.3 to 0.8)⇓.

The above findings raise several important questions. Firstly, is the association between male sex and recurrent VTE “real”? Male sex was not identified as a predictor of recurrent VTE in early epidemiological studies, but these retrospective studies did not adequately control for important confounders, such as type of VTE (provoked or unprovoked) and previous hormonal therapy. Later prospective studies provided …