Traveler's Thrombosis: A Systematic Review

Mohammed T. Ansari; Bernard M.Y. Cheung; Jia Qing Huang; Bo Eklof; Johan P.E. Karlberg


J Travel Med. 2005;12(3):142-154. 

In This Article


Of 16 observational studies reviewed, nine concluded an association between travel and VTE supported by statistically significant but imprecise risk estimates in eight.[1,2,8,15,16,17,20,22,23] Further, one of the two prospective studies that were not designed to estimate risk of prolonged travel also pointed to such an association.[9,11] Eight studies were restricted to air travel.[8,9,11,15,16,20,22,23] Five studies, all case-control, demonstrated no risk of VTE attributable to travel.[7,13,14,18,19] Only one of these studies exclusively examined business air travel.[7] Since most case-control studies were fraught with methodologic shortcomings, underpowered with a low prevalence of exposure to travel, and not restricted to investigating specific modes of prolonged travel, their evidence was considered of lower quality than the other studies.

We conclude that prolonged air travel qualifies as thrombogenic exposure, which can be of serious concern especially in those with preexisting risk factors for VTE. Since travelers are likely to be healthier than the general population,[22] the incidence of VTE in passengers remains low. No conclusions could be drawn about other modes of transport.

Case-control studies have inherent weaknesses. Problematic selection of controls with comparable characteristics other than the risk factor of interest, the inability of direct risk estimation, a small sample size relative to exposed total population, recall bias, and the restriction of cases to disease survivors are all examples of such limitations. Cohort or prospectively controlled studies overcome some of these weaknesses but still have limitations, for example, large sample size requirement, a bias owing to the loss of subjects, and biased assessment of the disease. These studies have reached contrasting conclusions. Differences in opinions are likely to be the results of differences in subjects/patients and controls, and overall prevalence of travel exposure; inclusion of symptomatic versus subclinical disease, venous thrombosis versus VTE, and proximal versus distal DVT; duration and type of travel; and diagnostic techniques employed.

Negative case-control studies that employed controls initially suspected for VTE—but ruled out later—would have underestimated the risk of VTE associated with travel because of referral bias owing to heightened concerns for TAVTE such that history of travel would have been overrepresented in controls.[13,14,18,19] Further, as travel histories were obtained before diagnosis, risk overestimation on account of recall bias is unlikely. On the other hand, risk could have been overestimated in studies that included sick people unlikely to travel as controls.[1,2] Further, given that the baseline age and sex standardized incidence of VTE is 1 in 1,000 in the general population,[25] case-control studies were largely underpowered because of limited sample size to give precise, significant, and adjusted risk estimates. Although the use of ultrasonography along with compressibility of veins in detecting proximal DVT is well established,[24,26] its role in diagnosing calf DVT is still controversial; sensitivity as low as 75% has been reported.[26] Therefore, significant distal DVTs that were more frequent in the traveling population could have gone undetected. It is also noteworthy that most case-control studies, with the exception of one, did not specifically examine prolonged travel. But crude odds ratio estimates from case-control studies should be interpreted with caution. In most, prevalence of established VTE risk factors was either higher in cases than in controls or not compared; only one case-control study adjusted for common VTE risk factors.[13]

The two prospectively controlled studies, primarily investigating lower limb venous thrombosis were, to a lesser extent, also limited in validity and generalizability by few differences in travel exposed and unexposed groups, a high percentage of participants with personal and/or family history of VTE in the larger study, a significant number of excluded participants, and by TAVTE-cautioned passengers who performed stretching exercises of lower limbs, walked in the cabin, and drank plenty of fluids—biases for both under- and overestimation of risk.[15,16] Further, anticoagulant therapy of calf vein thrombosis was likely to have underestimated the overall incidence of DVT.

Retrospective reviews and record linkage studies provide better evidence within limitations of their designs, which is likely to be conservative. Temporal and dose-response relationships were established as risk was estimated.[8,22,23] Importantly, these studies with significant but imprecise risk estimates are likelier to have underestimated the real risk of symptomatic VTE in association with prolonged air travel. The increase in risk with duration of air travel and in the 2-week hazard period following air travel undoubtedly suggests the thrombogenic nature of this exposure. The recent NZATT study that is, at best, an underestimation of incident VTE following prolonged air travel further strengthens our conclusion.

Although a meta-analysis of observational studies has been discouraged because of implications of confounding and selection bias,[27] the concern became obvious from this review. Gross heterogeneity in the studies precluded the possibility of statistical pooling.

The significance of travel-associated venous thrombosis remains unknown owing to lack in understanding of its natural history. Most of what is known about the natural history of deep venous thrombosis is from postoperative hospitalized patients. In the absence of treatment, about one-quarter of calf vein thrombi have been shown to extend more proximally to popliteal and iliofemoral veins,[28] and up to 40% of patients with proximal DVT may develop pulmonary embolism.[29] The significance of isolated calf vein thrombosis is even more uncertain. A prospective observation found ascending propagation up to the popliteal vein of symptomatic, isolated calf vein thrombosis in 10 days in 13% of limbs ( n = 7) in a group of patients, some of whom were on anticoagulants;[30] another observation up to 3 months could find thrombus propagation to the level of popliteal vein in only 3% of limbs ( n = 4) with baseline isolated calf vein thrombosis in patients who completed the follow-up.[31]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.