Meta-analysis: travel and risk for venous thromboembolism

Background: The potential risk for travel-related venous thromboembolism (VTE) has become an important public health concern because of rapid increases in long-distance travel; however, previous studies on this relationship are surprisingly contradictory. Purpose: To estimate the risk for VTE in travelers, determine whether a dose-response relationship exists, and identify reasons for the contradictory results of previous studies. Data sources: MEDLINE, EMBASE, BIOSIS, CINAHL, grey-literature sources, contact with investigators, and reference lists of studies, without language restrictions. Study selection: Reports were selected if they investigated the association between travel and VTE for persons who used any mode of transportation and if nontraveling persons were included for comparison. Data extraction: Data on study and patient characteristics, risk estimates, and quality were independently extracted by 2 investigators. Pooled effect estimates were obtained by using random-effect meta-analysis. Data synthesis: Of 1560 identified abstracts, 14 studies (11 case-control, 2 cohort, and 1 case-crossover) met inclusion criteria, including 4055 cases of VTE. Compared with nontravelers, the overall pooled relative risk for VTE in travelers was 2.0 (95% CI, 1.5 to 2.7). Significant heterogeneity was present because of the method for selecting control participants (P = 0.008): whether the studies used control participants who had been referred for VTE evaluation or nonreferred control participants. When the studies that used referred control participants were excluded, the pooled relative risk for VTE in travelers was 2.8 (CI, 2.2 to 3.7), without significant heterogeneity. A dose-response relationship was identified, with an 18% higher risk for VTE for each 2-hour increase in duration of travel by any mode (P = 0.010) and a 26% higher risk for every 2 hours of air travel (P = 0.005). Limitation: All available studies were from Western countries; generalizability to non-Western populations is expected but needs confirmation. Conclusion: Travel is associated with a nearly 3-fold higher risk for VTE, with a dose-response relationship of 18% higher risk for each 2-hour increase in travel duration. Heterogeneity in results of previous studies was due to selection bias toward the null from use of referred control participants.