Abstract
Background: The potential association between venous thromboembolism (VTE) and travel, particularly air travel (“economy class syndrome”), has been the subject of extensive media coverage. While it is biologically plausible that prolonged travel is an independent risk factor for VTE, epidemiological data to date are conflicting, and confounders have rarely been accounted for.
Aim: To determine whether there is a greater risk of exposure to travel in patients with confirmed DVT compared with patients in whom DVT is ruled out. To examine the influence of confounding variables on the relation between DVT and travel.
Methods: This was a Canadian multi-center case control study. Consecutive patients presenting to the vascular laboratory with clinically suspected DVT were eligible to participate. Cases were patients with objectively confirmed DVT on venous ultrasound; controls were patients in whom DVT was ruled out. Detailed recent travel history, medications and clinical characteristics were obtained via standardized, interviewer-administered questionnaires. Genetic testing for Factor V Leiden and Prothrombin gene mutations was performed. Unconditional multivariate logistic regression analyses with adjustment for confounders and testing for interactions were performed to examine the relation between DVT and (1) any travel, and (2) duration of travel. Plane and car travel were also analyzed separately.
Results: There were 359 cases and 359 controls. Mean age among cases was 56 years and 50% were male. Among controls, mean age was 64 years and 35% were male. Body mass index, smoking status and patient location (inpatient vs. out-patient) were comparable between the two groups. The crude and adjusted odds ratio (OR) for exposure to travel in cases was 1.15 (95% confidence interval (CI): 0.78, 1.69) and 1.44 (95%CI: 0.86, 2.40), respectively. Travel of ≥ 12 hours’ duration was associated with a higher OR (adjusted OR 2.92, 95%CI: 0.54, 15.73) than shorter travel durations (adjusted OR 1.29; 95%CI: 0.62, 2.66). Analyzing plane and car travel separately showed that the adjusted OR for plane travel was 2.28 (95%CI: 0.94, 5.50) but for car travel was 1.00 (95%CI: 0.54, 1.83). Increasing durations of plane travel, but not car travel, resulted in higher ORs. For plane travel ≥ 12 hours, the crude OR was 8.22 (95%CI: 1.02, 66.05) and the adjusted OR was 7.10 (95% CI: 0.70, 72.35). No statistical interactions were detected between travel and thrombophilia, hormonal therapy, or clinical VTE risk factors.
Interpretation: This is the largest case control study to date of the relation between DVT and travel that takes into account concurrent DVT risk factors. Plane travel but not car travel appears to be a mild independent risk factor for DVT. However, flights of 12 hours or longer were associated with a 7-fold increased risk of DVT. Our findings may have future implications regarding the use of thromboprophylaxis during long-haul travel.
Risk factors for post-bronchoscopy pneumonia: a case–control study