|Home | About | Journals | Submit | Contact Us | Français|
Despite multiple attempts to document and quantify the danger of venous thromboembolism (VTE) following prolonged travel, there is still uncertainty about the magnitude of risk and what can be done to lower it.
To review the methodologic strength of the literature, estimate the risk of travel-related VTE, evaluate the efficacy of preventive treatments, and develop evidence-based recommendations for practice.
Studies identified from MEDLINE from 1966 through December 2005, supplemented by a review of the Cochrane Central Registry of Controlled Trials, the Database of Abstracts of Reviews of Effects, and relevant bibliographies.
We included all clinical studies that either reported primary data concerning travel as a risk factor for VTE or tested preventive measures for travel-related VTE.
Two reviewers reviewed each study independently to assess inclusion criteria, classify research design, and rate methodologic features. The effect of methodologic differences, VTE risk, and travel duration on VTE rate was evaluated using a logistic regression model.
Twenty-four published reports, totaling 25 studies, met inclusion criteria (6 case-control studies, 10 cohort studies, and 9 randomized controlled trials). Method of screening for VTE [screening ultrasound compared to usual clinical care, odds ratio (OR) 390], outcome measure [all VTE compared to pulmonary embolism (PE) only, OR 21], duration of travel (<6 hours compared to 6–8 hours, OR 0.011), and clinical risk (“higher” risk travelers compared to “lower,” OR 3.6) were significantly related to VTE rate. Clinical VTE after prolonged travel is rare [27 PE per million flights diagnosed through usual clinical care, 0.05% symptomatic deep venous thrombosis (DVT) diagnosed through screening ultrasounds], but asymptomatic thrombi of uncertain clinical significance are more common. Graduated compression stockings prevented travel-related VTE (P<0.05 in 4 of 6 studies), aspirin did not, and low-molecular-weight heparin (LMWH) showed a trend toward efficacy in one study.
All travelers, regardless of VTE risk, should avoid dehydration and frequently exercise leg muscles. Travelers on a flight of less than 6 hours and those with no known risk factors for VTE, regardless of the duration of the flight, do not need DVT prophylaxis. Travelers with 1 or more risk factors for VTE should consider graduated compression stockings and/or LMWH for flights longer than 6 hours.
In 2003, 104 million passengers flew into or out of the United States on transoceanic flights.1 By 2015, this number will almost double. While in transit, some of these passengers will develop deep venous thrombosis (DVT) or pulmonary embolism (PE).2–4 Despite multiple attempts to quantify the danger, there is still uncertainty about the risk of venous thromboembolism (VTE) from prolonged travel, which travelers should receive VTE prophylaxis, and what prophylactic measures should be used. We performed a systematic review of this literature to address the following questions: (1) What are the methodologic strengths and weaknesses in this literature? (2) What is the risk for travel-related VTE? (3) Are there effective preventive measures? Using our findings, we propose evidence-based recommendations for practice.
Using the MEDLINE database from 1966 through December 2005, we searched for articles evaluating human subjects that either reported primary data concerning the risk of travel for VTE or tested preventive measures for travel-related VTE. Our search included the MESH headings thromboembolism, thrombosis, venous thrombosis, thrombophlebitis, and pulmonary embolism. We also searched under the MESH headings travel, aerospace medicine, aviation, and transportation, and the text words flight and flying. We then combined the 2 searches into an initial set of 465 articles. We sought additional articles by performing the same search strategy in the Cochrane Central Registry of Controlled Trials and the Database of Abstracts of Reviews of Effects.5 We then reviewed the references from relevant articles in the initial set for articles not already identified. We included case-control studies, cohort studies, and randomized controlled trials (RCTs), but excluded case reports, abstracts, and subgroup analyses of previously published studies.
Two of 3 authors (RS, DMB, and JTP) reviewed each study independently to determine whether it met inclusion criteria, classified it according to research design, and evaluated it for methodologic strength. The ratings of the 2 reviewers were compared and discrepancies were resolved by discussion to achieve consensus. To evaluate methodologic strength, we used 8 standards adapted from previous research on the natural history of disease.6,7 These standards are described in the Appendix.
For each cohort study group and RCT study arm, we recorded number of subjects, number of subjects with VTE, VTE outcome (PE, DVT, or both), travel duration (mean of less than 6 hours, mean of 6 to 8 hours, and mean greater than 8 hours), treatment (yes/no), method of screening for VTE (usual clinical care or screening ultrasound after travel) and clinical VTE risk of travelers (lower or higher). We then examined associations between these variables and VTE outcomes in untreated travelers using the 26 study arms where there was both travel and no intervention (treatment=no). Because the dependent variable was dichotomous (whether or not each participant had a VTE), we concluded that the natural distributional assumption was the binomial distribution. Because we only had summary data on groups rather than data on individual subjects, we used the total number of subjects in each group and the number of those with VTE to define a grouped binary data structure. We then fitted a logistic regression model to the data, where the coefficients of regression were log odds ratios (LOR). These coefficients were related to the dependent variable, VTE rate (r), by 2 transformations. The first was a transformation to an odds ratio (OR) and the second to a logarithm [LOR=log(r/(1−r)].8,9 This method allows for a more realistic distributional assumption for the dependent variable than the normal distribution, accounts for different sample sizes of the studies, has more statistical power, allows independent variables to be incorporated into the model, estimates effect of those variables on the dependent variable, and decreases heterogeneity. We used SAS/STAT software, version 9.1, to fit the models and estimate the parameters (PROC GENMOD).
The search strategy identified 24 publications including 25 studies that met criteria for inclusion in this review.10–33 There were 6 case-control studies, 10 cohort studies, and 9 RCTs, with 1 article including the results of both a cohort study and a RCT.18
The 6 case-control studies, described in Table 1, ranged in size from 20714 to 988 subjects12 and were from 4 different countries. Five of the 6 investigated risk from any form of travel,11–15 while 1 limited its focus to air travel.16 The 10 cohort studies, described in Table 2, ranged in size from 32019 to more than 135 million subjects17 and were performed in 9 countries. All investigated air travel risk only. The 9 RCTs, described in Table 3, ranged in size from 18632 to 83318 and were performed in 3 countries. All limited investigation to air travel. Eight of the 9 trials18,27–33 came from the same group of investigators.
Compliance with the methodologic standards is noted in Tables 1 through through3.3. The reviewed studies used strong research methods to establish VTE diagnosis and to evaluate the efficacy of preventive interventions. Each of the included studies used objective diagnostic tests to confirm the VTE (standard 7, Appendix). Therefore, within the limits of the test operating characteristics, any reported episode of VTE was in fact VTE. Each of the preventive interventions was evaluated with RCTs so that conclusions about treatment efficacy are less prone to bias.
There were 3 major methodologic weaknesses of the reviewed studies. Because none of the studies avoided all 3, it was impossible to precisely determine the risk of travel-related VTE.
First, the 4 largest cohort studies did not undertake routine VTE surveillance (standard 6, Appendix) but, rather, identified cases from retrospective reviews of medical records.10,17,21,23 This strategy leads to a low count of VTE cases, particularly when PEs but not DVTs are counted and when cases are sought for a limited time period after travel in a limited number of medical settings.10,17,21 As we expected, when cases are identified from medical records but without limits on time or setting, a higher count is reported.23 When routine surveillance with lower extremity ultrasound is used, as occurred with most of the cohort studies and all of the controlled trials, there are many more VTE cases identified, mostly asymptomatic DVTs. Our multivariate analysis, controlling for travel duration and clinical risk, found that the OR for screening ultrasound–diagnosed VTE was 390 compared to VTE diagnosed through usual clinical care (Table 4).
Second, many of the studies were prone to volunteer bias. While 4 studies enrolled consecutive series or sampled the entire population,10,17,21,23 most studies enrolled volunteers who tend to be healthier than nonvolunteers.34 In addition, volunteer bias works to compound the “healthy traveler effect,” providing an explanation for why the baseline risk of VTE in travelers is less than that of the general population.23
Third, outcome measures varied among studies. Some studies looked only for DVTs, others only for PEs, and others for all VTEs (i.e., both DVTs and PEs). As we expected, there were more frequently reported outcomes when DVT was included than when the outcome was PE alone [OR 23 and 21, respectively, for DVT and all VTE compared to PE alone (Table 4)].
The cohort studies and the control groups of the RCTs provide information about the incidence of VTE after travel. These studies looked only at air travel and report a wide range of VTE risk, from none19,24,29 to 12% of travelers.26 As expected, studies of symptomatic VTE patients reported lower rates of VTE, about 0.5 PEs per million travelers presenting in the airport on the day of arrival10,17,21 and 27 VTEs (both PE and DVT) per million travelers presenting within 14 days of travel.23
The studies performing ultrasound surveillance found much higher VTE rates in travelers, 1.2% DVT (44/3,820, ranging from 0 to 12%). Of note, all subjects were proven by ultrasound before embarkation to be DVT-free. One third of the reported DVTs were in proximal veins, and two thirds were in calf veins. However, only 2 of the 44 ultrasound-positive patients had symptoms. Therefore, the overall rate of symptomatic DVT was 0.05%, but with a wide 95% confidence interval (0.01% to 0.19%). One study screened travelers with the D-dimer test followed by 3 months of surveillance for VTE symptoms and reported a cumulative VTE rate of 1.0%, with 5 of 9 cases symptomatic.22
Two cohort studies included nontraveling control groups.19,20 Although no statistically significant difference in VTE rates between travelers and nontravelers was found, in one of these studies there was a trend toward a higher DVT rate in the travelers (0.7% vs 0.2% in nontravelers, P=NS).20
A travel dose–response curve is demonstrated in individual studies17,21 and confirmed by our multivariate model (Table 4). Symptomatic PE is almost unheard of for flights of less than 6 hours17,21 [OR of 0.011 compared to flights of 6–8 hours (Table 4)], but flights of more than 8 hours have increased risks (OR for VTE of 2.3 compared to flights of 6–8 hours).
In addition to flight duration, our multivariate analysis found that a higher clinical VTE risk was associated with higher VTE rates (Table 4). Nine studies17,20–22,26,27,30–32 reported the following risk factors in 126 VTE patients: age over 40, 45%; female hormone use, 31%; varicose veins, 19%; obesity, 17%; inherited clotting disorder, 6%; and other factors, 7%. None of these studies compared rates of DVT risk factors in VTE subjects with those of non-VTE subjects. Despite the publicity surrounding “economy class syndrome,” no study compared the seating of VTE subjects (first class or not, aisle seating or not) with that of non-VTE subjects.
Table 3 lists the studies that evaluated preventive interventions, all RCTs. All subjects, whether in a control arm or an intervention arm, were encouraged to exercise and maintain good hydration during flight. The interventions tested were low-molecular-weight heparin (LMWH), aspirin, graduated compression stockings, and herbal remedies with putative antithrombotic properties. Only 1 study30 evaluated LMWH (enoxaparin) and found a trend favoring effective DVT prevention that did not reach statistical significance. The same study found no effect from aspirin. No studies tested warfarin or unfractionated heparin. Graduated compression stockings, ranging from 12 mmHg to 30 mmHg at the ankle, were evaluated in 6 studies, and 4 of these studies reported a statistically significant benefit.18,26,27,31 Of the 1,237 subjects using compression stockings, only 2 DVTs were found (0.2%), compared to 46 in the 1,245 controls (3.7%). Two studies32,33 evaluated preparations containing pycnogenol, an extract of pine bark with antioxidant and possible antithrombotic effects,35 with 1 study reporting fewer DVTs among treated passengers.32
We conclude that there is a risk of symptomatic VTE from prolonged air travel. Our estimate of the magnitude of this risk varies according to study methodology. If we depend on VTE rates identified through usual clinical care, the risk may be as low as 27 cases per million travelers [number needed to harm (NNH)=38,000].23 This estimate, equivalent to 1.1 VTE per million person-days and incorporating the healthy traveler effect while avoiding volunteer bias, is close to but less than the baseline risk in general populations, estimated to be 1.9 to 5.2 VTE per million person-days.23,36–40
If we depend on ultrasound screening studies, the overall VTE risk may be as high as 1.2%. However, in contrast to symptomatic travel-related PE, little is known about the natural history of asymptomatic travel-related DVT. We believe that most asymptomatic thrombi resolve without treatment but that some progress to symptomatic DVTs and PEs during the several weeks after travel.7,41,42 Therefore, the rate of clinically important travel-related VTEs must be less than the rate found by screening ultrasound studies. Counting only symptomatic DVTs, 0.05% is a more likely estimate of the incidence of clinically important DVTs from prolonged air travel. Assuming none of these travelers would have developed DVTs if they had stayed home, the NNH would be 2,000.
VTE risk is very low when flight time is less than 6 hours, but is progressively greater with longer-duration flights. However, the overall VTE risk is still much lower than what is considered the “low-risk” category in surgical patients (2% calf DVTs, 0.4% proximal DVTs, and 0.2% clinical PEs).43 The odds of VTE after long flights (>8 hours) and for higher-risk travelers (Table 4) are slightly greater than those reported for “weak” VTE risk factors (e.g., bed rest more than 3 days, increasing age, laparoscopic surgery, obesity, antepartem pregnancy, and varicose veins) and similar to those reported for “moderate” risk factors (e.g., arthroscopic knee surgery, congestive heart failure, hormone replacement therapy, stroke, postpartum pregnancy, previous thromboembolism, antithrombin deficiency, proteins C and S deficiencies, prothrombin G20210A, and heterozygous factor V Leiden).44,45 They are much less than those reported for “strong” risk factors (e.g., hip fracture, joint replacement, major trauma, spinal cord injury, oral contraceptives plus factor V Leiden, and homozygous factor V Leiden).44,45 Healthy travelers without VTE risk factors other than travel do not need to worry about VTE from prolonged air travel. However, travelers with VTE risk factors are at increased risk directly related to the importance of their risk factor(s) and the duration of their flight.
In addition to exercise and hydration for all travelers, compression stockings are helpful for long-haul travelers. LMWH does not have established efficacy, and none of the reviewed studies evaluated warfarin and unfractionated heparin. However, the efficacy of LMWH, unfractionated heparin, and warfarin for VTE prophylaxis has been established in other settings.43 If compression stockings could not be used or if additional preventive measures were thought to be needed, we believe that various proven anticoagulant regimens (e.g., low-dose unfractionated heparin bid or tid; LMWH qd or bid) could be adapted to prolonged air travel, including the one used by Cesarone et al.30 We favor the use of compression stockings over anticoagulants because the number needed to treat is likely more than 2,000 and there is some risk, although small, from anticoagulant use. Because aspirin is ineffective in other DVT prophylaxis settings, we were not surprised that the evidence available in this review did not suggest an aspirin benefit. We do not recommend pycnogenol preparations at this time, due to the limited data on its properties.
The decision regarding VTE prophylactic measures in long-distance air travelers should be made in a manner analogous to other decisions regarding VTE prophylaxis—according to the degree of risk. Unfortunately, in contrast to surgery, risk levels for air travel are not well defined for the clinical conditions associated with increased VTE risk. We make the following specific recommendations:
As always, further research is needed; for example: evaluation of the VTE risk of travel (including the healthy traveler effect) using study methodology that avoids volunteer bias, screens for both PE and DVT, includes an assessment of VTE symptoms, and follows travelers for a sufficient time after arrival; evaluation of the clinical importance of asymptomatic DVT after prolonged travel; further investigation of the additional risk conferred by air travel on travelers with 1 or more VTE risk factors, including whether seating location is a risk factor; derivation and validation of a decision rule for VTE prophylaxis for air travel; further investigation of the VTE risk of other travel modalities; and further investigation of the effectiveness of other preventive measures including LMWH and unfractionated heparin.
Potential Financial Conflicts of Interest None disclosed.
Standard 1: Adequate description of the subject assembly process (a) Methods for patient selection should be described in enough detail that the study could be replicated with a similar group of patients. (b) The number of eligible but not enrolled subjects, as well as any reasons for exclusion, should also be reported.
Standard 2: Adequate description of subjects Summary demographic information (gender, age), as well as type and duration of travel, should be reported.
Standard 3: Equality of comparison groups RCT should not only employ random allocation to treatment modality but also demonstrate that the treatment groups were similar in terms of demographics (age and sex) and duration of travel. In cohort studies with control groups, both demographic characteristics and VTE risk should be similar between groups; if not, statistical adjustment for differences should have been performed. In case-control studies, the cases and controls should be matched for age, sex, and at least 1 VTE risk factor; otherwise, statistical adjustment should take these variables into account. With the exception of travel history, cases and controls should be enrolled using the same selection criteria.
Standard 4: Adequate reporting of subject follow-up The number of patients unavailable for outcome or exposure ascertainment should be reported, as well as the reasons why this information was not available.
Standard 5: Adequate description of treatment Treatment should be described in enough detail so that other subjects could be treated in a similar fashion.
Standard 6: Unbiased surveillance for adverse outcomes and determination of exposure For cohort and RCT studies, either objective VTE test results or systematic survey for travel-related VTE symptoms should be reported. For case-control studies, travel exposure should be ascertained in the same fashion for cases and controls, including use of measures that would limit recall bias.
Standard 7: Adequate VTE diagnostic evaluation Any diagnoses of DVT or PE should be based on objective test results.
Standard 8: Analysis that takes type and duration of travel into account VTE risk should be adjusted by travel type and duration.