Author and address for correspondence:

The Lister Hospital, Chelsea Bridge Road, London, SW1W 8RH

Tel: 020 7730 9563

Abstract

Background

Aim

Methods

Results

Conclusions

Keywords

Introduction

Patients and methods

Passengers were invited to the Stamford Hospital to undergo preliminary screening which involved an examination and completion of a medical questionnaire concerning previous medical problems and history of medication. Those volunteers who were eligible for inclusion in the study were investigated by duplex ultrasonography to detect evidence of previous venous thrombosis. The lower limbs were assessed by two qualified vascular technologists experienced in assessing venous problems. The examinations were conducted with volunteers standing using a General Electric LOGIQ 700. The competence of deep and superficial veins was assessed by manually compressing the calf and measuring the duration of reverse flow by colour or pulsed Doppler sonography. Venous reflux in a vein was defined as the duration of reverse flow exceeding 0.5 seconds. The presence of current or previous venous thrombosis was assessed from the B-mode image, colour flow mapping and compression assessment of veins during B-mode imaging. Passengers who had evidence of previous thrombosis were excluded from the study.

In the first 30 volunteers, ultrasound examination was undertaken two weeks before air travel and again within two days of commencing the first flight to provide a control interval in which the incidence of spontaneous DVT could be assessed in this population. No acute DVT was detected during this period. The logistics of the study made it difficult for passengers to attend the Stamford Hospital on two occasions prior to travel. This part of the investigation was abandoned in the remaining volunteers. All subsequent volunteers were screened once before commencing their air travel.

Blood was taken from all subjects prior to travel for a series of haemostatic tests. A full blood count and platelet count were performed on the routine cell counter (Sysmex XE-2100). D-dimer measurement was performed using the Dimertest® Gold EIA assay (Agen Biomedical Ltd, Acacia Ridge, Australia). The upper limit of normal was 120 ng/ml, using 95% confidence limits. The identification of the factor V Leiden (FVL) and the prothrombin G20210A (PGM) gene mutations were performed using our routine PCR techniques.

The volunteers were randomised according to a sealed envelope to one of two groups. The first group received no specific additional treatment. The second group were given a class I (German Hohenstein compression standard) below-knee elastic compression stocking (Mediven Travel supplied by Medi UK Ltd, Hereford, UK). Passengers were advised to put on the stockings before commencing their travel and to remove the stockings after arrival at their destination for each of the flights by which they travelled. Although the stockings were allocated on a random basis, the presence of the stockings made it impossible to conduct this as a blinded study. The passengers arranged their own air travel. There was no collaboration with the airlines, although two passengers were upgraded from economy class to business class.

Passengers re-attended the Stamford Hospital within 48 hours of their return flight. They were first interviewed by a research nurse and completed a questionnaire enquiring about the duration of air travel, wearing of stockings and of any symptoms in the lower limbs, they were asked if they suffered any illnesses or took any medication during their trip. The majority of passengers removed their stockings on completing their journey. The nurse removed the stockings in those passengers who had continued to wear them. A further duplex examination was then undertaken with the technologists unaware of the group to which the volunteer had been randomised. A further blood test was taken for repeat D-dimer assay. In passengers where significant abnormalities of the lower limb veins were detected on duplex ultrasonography, including calf vein thrombosis, the volunteers’ general practitioners were notified in writing so that treatment could be arranged.

In this study it was impossible to make any pre-calculation of sample size in view of the lack of available data. Since this investigation was intended as a pilot study, a total of 200 passengers was decided upon. Recruitment was continued until 100 volunteers had been investigated in each group. The finding of no case of venous thrombosis in this number of passengers would have resulted in a 95% confidence interval for the incidence of DVT of 0 - 2%. To measure the incidence of a thrombotic event occurring in 2% or fewer passengers would require a very large study and would have limited implications for air travellers. Data from the study were analysed by constructing contingency tables and calculating the differences in proportions and 95% confidence intervals using a computer program (CIA, BMA Publishers, London, UK). Summary statistics for haematological data used the median and interquartile range, since data was not of a Gaussian distribution. Haematological data was only included in the analysis when volunteers attended for examination before and after travel. All other analyses have been prepared on an ‘intention-to-treat’ basis which includes all randomised subjects.

Results

Table 1 compares the characteristics of the two groups of volunteers. Most factors are closely matched, but by chance a greater proportion of women were included in the stocking group.

The summary statistics from the haematological investigations are shown in table 2. Twelve (10%) of passengers not wearing elastic stockings developed an asymptomatic DVT in the calf on duplex ultrasound examination following air travel. None of the 115 passengers wearing compression stockings suffered a DVT. Wearing elastic compression stockings is associated with a reduced incidence of deep vein thrombosis (difference between proportions 10%, 95% CI 4.8 - 16%, table 3). A further four subjects developed superficial thrombophlebitis in varicose veins, all of whom were wearing compression stockings.

Four of the patients with asymptomatic DVT received treatment with low molecular weight heparin given subcutaneously for 5 days and were referred to their GPs for continuing treatment. The remaining 8 were asked to take aspirin, referred to their general practitioner and advised to undergo a further scan and receive treatment, if appropriate. The general practitioners were kept informed of these developments. The four passengers with superficial thrombophlebitis received treatment, one with aspirin and three with the non-steroidal anti-inflammatory drug diclofenac.

Fourteen volunteers (7%) of the 200 who completed both pre- and post travel examinations were found to have either the factor V Leiden or prothrombin gene mutation (PGM). There were eleven subjects with FVL and four with PGM, all in the heterozygous state. One subject had both FVL and PGM and suffered an episode of thrombophlebitis. Two of the passengers with DVT had FVL. The full blood count, platelet count and D-dimer assays provided no prognostic information. No passenger with a DVT was found to have elevated D-dimer levels in the post-travel sample.

The before and after travel questionnaires were examined to assess concomitant medication, including that commenced during the period of air travel. Only two passengers took drugs in addition to their usual medication during the study period. The results are summarised in table 4. Most drugs are evenly distributed in the two groups, although there is a trend towards more patients taking hormone replacement in the stockings group (percentage difference 8%, (95% C.I. -1 - 17%). Aspirin was taken by a number of volunteers as part of their regular medication. Similar numbers of passengers took this in each group.

Discussion

We were concerned that because of their interest in the problem some of the volunteers may have taken steps to reduce the incidence of venous thrombosis, i.e., by being more active on the aeroplane and drinking more fluids. It was not possible to assess the influence the study had on the behaviour of subject whilst aboard the aircraft on which they travelled. These factors would have applied equally to our study groups. In fact, no properly conducted study has ever been published to show that leg exercises, walking or drinking water prevent thrombotic events following airline travel. Nevertheless, we consider that we have represented a reasonable sample of passengers departing from London Airports, deliberately excluding volunteers with previous venous thrombosis. We only included passengers travelling in the economy cabin because evidence suggested these passengers have limited leg room and cramped conditions. ⁴ We would therefore expect these passengers to be more likely to develop a deep vein thrombosis.

The authors ensured, as far as was reasonable within the scope of this pilot study, that risk factors for DVT were evenly distributed between the two study groups. One of the most important factors predisposing to DVT is a previous history of a DVT or other objective evidence of venous thrombosis. ¹⁴ Any volunteer with a history of DVT or post-thrombotic damage to their deep veins shown on duplex ultrasonography was excluded from randomisation. Malignant disease, severe pulmonary or cardiac disease predispose to venous thrombosis¹⁴ so volunteers with these diagnoses were also excluded. This limited the likelihood of an uneven distribution of risk factors for DVT confounding the study as well as ensuring that our study reflected the risk of DVT in an otherwise healthy population of passengers over the age of 50 years. Previous studies show that age is an important factor determining the risk of venous thrombosis.¹⁴ The data in table 1 confirms that this was closely matched in our two study groups. Varicose veins have been implicated in the development of DVT¹⁴ and occur in about 30 - 40% adults in the UK.¹⁵ These were equally distributed in our two study groups. The authors suspect that the duration of air travel, especially long haul air travel, is important in causing venous thrombosis. Again the two study groups are closely matched for this factor. Prothrombotic gene mutations were evenly distributed between the study groups. However, by chance the stocking group contained more women than men by a significant proportion (table 1). The randomisation procedure was not stratified or minimised for any factor, since the authors regarded this study as a pilot investigation for the much larger studies that would be required definitively to investigate the incidence of DVT following long haul air travel. There is little evidence to show that women are more or less susceptible than men to venous thrombosis in the age group we investigated.¹⁶ Following airline travel, DVT occurred approximately evenly distributed between men and women (5 of 55 men and 7 of 61 women, table 3) in the non-stocking group. Drugs may also influence the development of DVT. The medication taken by our volunteers is summarised in table 4. This shows that volunteers more frequently took hormone replacement in the stocking group. This probably reflects the larger numbers of women in this group, but is a medication which may increase the risk of DVT.¹⁷ However, no DVT occurred in this group of passengers. Aspirin is advocated for primary and secondary prevention of arterial thrombotic events¹⁸ as well as prevention of DVT.¹⁹ This drug was taken by a number of volunteers in our study but was evenly distributed between the stocking and non-stocking groups. A number of other drugs including anti-hypertensives, anti-peptic ulcer drugs and thyroxine were taken by our volunteers. The authors can find no evidence that these influence the development of venous thrombosis and in any case this medication was taken evenly in the two groups of volunteers. In summary, our simple randomisation procedure failed to produce equal distribution of men and women in the stocking and non-stocking groups, but succeeded in all other respects. Our stocking group included more women who took more HRT that the non-stocking group, a factor which might have led to an increased risk of DVT. No DVT was detected in this group.

The method of detecting a deep vein thrombosis was duplex ultrasonography. Venography could not have been undertaken in this study since it was conducted on asymptomatic volunteers. We consider that it would have been unethical to use venography in this subject group. Recent studies have shown that duplex ultrasonography is a reliable method of detecting calf vein thrombosis, as well as proximal vein thrombosis, in asymptomatic patients ²⁰, ²¹. A series of studies has been conducted in which the reliability of duplex ultrasonography in the diagnosis of calf vein thrombosis has been compared to venography. ²², ²³, ²⁴, ²⁵, ²⁶ The main failing of duplex ultrasonography is that it may underestimate the true incidence of calf vein thrombosis. In the studies mentioned here the sensitivity of duplex ultrasonography compared to venography ranged from 67% - 100%. Therefore, our data may have underestimated the true incidence of calf vein thrombosis by as much as 30%. The same studies show that the specificity of duplex ultrasonography ranges from 79% to 99%. Therefore, it is unusual for this investigation to result in false positives. The two studies in which duplex ultrasonography had been used as a screening measure ²², ²⁴, showed sensitivities of 83% & 74% and specificities of 85% & 99%, confirming that it is an effective diagnostic tool when used in this context.

The fact that some subjects wore compression stocking until shortly before the post-travel examination is unlikely to have influenced the sensitivity of the test and is not the reason why no thrombosis was detected in this group. The authors have conducted a number of investigations in which the calf veins were imaged by ultrasound during the wearing of compression stockings ²⁷, ²⁸ and discovered no factor that would preclude the diagnosis of deep vein thrombosis after stocking wear. Since compression stockings are widely used in hospital practice to prevent deep vein thrombosis, many patients in the studies mentioned above ²², ²³, ²⁴, ²⁵ were probably wearing these prior to investigation. Stocking wear was not reported as a factor which might influence the sensitivity or specificity of duplex ultrasonography in the diagnosis of deep vein thrombosis. The most important factors are whether it is technically possible to image the deep veins and the presence of post-thrombotic vein damage.²⁵ All volunteers with post-thrombotic appearances on ultrasonography were excluded from this investigation and none of our subjects was affected by severe calf swelling which prevented adequate images of the calf veins being obtained. We consider that neither of these factors resulted in the failure to find a single calf vein thrombosis in the stocking group. In those passengers where a diagnosis of calf vein thrombosis was made, we considered it unnecessary and unethical to perform any further confirmatory test (such as venography) since it is our normal clinical practice to use the results of duplex ultrasonography in managing our patients with calf vein thrombosis. The specificity of duplex ultrasonography when used as a screening measure is in the range 85 - 99% ²², ²⁴ so there is no need for additional investigation. Venography is not as precise as some would assume. In one study the agreement between different radiologists resulted in a value of kappa for agreement between observers of only 0.568 to 0.669 with a difference of up to 16% between radiologists. ²⁹ Much of the variation was attributable to differences in opinion concerning the state of calf veins. Values of kappa between 0.65 and 0.92 for agreement between pairs of observers have been reported in another venographic study. ³⁰

The authors accept that asymptomatic calf vein thrombosis is probably not a major risk to health. Published clinical series concern DVTs detected following investigation for calf symptoms. These show that 10 - 20% of isolated calf vein thromboses undergo extension to more proximal veins. ³¹, ³² Pulmonary embolism may occur in about 10% of patients presenting with isolated calf vein thrombosis. ³¹, ³³ In the longer term as many as half of patients with calf vein thrombosis may experience symptoms in the lower limb attributable to post-thrombotic vein damage. ³³, ³⁴ However, patients presenting with symptomatic calf vein thrombosis often have recognised predisposing factors such as malignant disease or thrombophilias. ³¹, ³² Since we excluded patients with a history of serious illness or previous thrombotic episodes and all those with post-thrombotic vein damage on duplex ultrasonography, we consider that the thrombi detected in our study were attributable to long haul air travel rather than to any other factor.

We were surprised at the number of passengers in whom an asymptomatic calf vein thrombosis developed in the non-stocking group. This was presumably because we studied an elderly group of volunteers who undertook lengthy airline travel, both factors which would be likely to increase the risk of thrombosis. In comparison, studies conducted during the 1970s showed that patients over the age of 40 years undergoing general surgical operations developed asymptomatic deep vein thrombosis (detected by radiofibrinogen scanning) in about 30% of cases.³⁵ Therefore the incidence of 12% found in our highly selected passenger group should not be surprising. Presumably the majority of isolated calf DVTs which develop following air travel resolve completely, otherwise the number of symptomatic DVTs treated by hospital practitioners in passengers following long-haul air travel would be much greater. The frequency of asymptomatic calf vein thrombosis in the general population does not appear to have been investigated and therefore neither has the outcome. We suggest that long haul airline travel acts as the factor which provokes calf vein thrombosis. Once the journey has been completed this factor is removed. In the majority of cases this allows spontaneous resolution of calf vein thromboses without complication. In a minority of instances extension of the thrombus occurs resulting in symptomatic deep vein thrombosis and occasionally to pulmonary embolism. This is in contrast to patients presenting to hospital with symptomatic calf vein thrombosis in whom the factors which caused the calf vein thrombosis remain leading to a less satisfactory outcome, ³¹, ³² as discussed above.

D-dimer, a specific degradation product of cross-linked fibrin, measured by a sensitive EIA procedure is a useful diagnostic tool for the detection of venous thromboembolism. ³⁶ Failure to detect raised levels of D-dimers in passengers with positive ultrasound scans may reflect the short half life of D-dimer levels (about 6 hours) combined with the delayed time to blood sampling on return from travel. The interval of up to 48 between completion of the final leg of air travel may have been sufficient to prevent this test being useful. In addition, in all volunteers who developed a DVT, the thrombus was limited to the calf veins and this would also result in a modest rise in plasma D-dimer levels. The authors have personal knowledge of patients outside this study with moderately extensive DVT in whom the D-dimer levels were normal. D-dimer levels detects developing thrombosis and unless this is present at the time of the blood test a DVT may not be detected.

A wide range of specific haemostatic abnormalities is now recognised to predispose to venous thrombosis. The genetic defects studied in this investigation (FVL and PGM) are together present in about 10% of European populations and might be considered as predisposing to airline DVT. In our investigation, the combined prevalence of these abnormalities was 7% in all volunteers, but 19% in those developing superficial or deep venous thrombosis. This data must be regarded with caution, as the total number of subjects studied is too small for statistical analysis. In this study we found that only those passengers who did not wear stockings developed deep vein thrombosis. Those volunteers wearing elastic stockings did not develop a deep vein thrombosis, but four passengers developed superficial thrombophlebitis behind the knee in pre-existing varicose veins. Airlines are now providing passenger leaflets about exercises to be performed during flights. These exercises comprise ankle and foot movements. They have not been found to be effective in influencing leg swelling, which commonly accompanies long haul air travel ³⁷. The effect of these exercises on the incidence of deep vein thrombosis has not yet been studied. Compression stockings have been widely used in the prevention of venous thrombosis in hospital practice ³⁸, and are commonly advocated by phlebologists for patients with chronic venous disorders of the limb. Their role in prevention of venous thrombosis following airline travel had not been investigated. We used class I compression stockings to ensure maximum compliance of our volunteers. These are easily applied and the below knee variety are well tolerated by passengers. In four passengers with large varicose veins the stocking may have caused superficial thrombophlebitis. The total number of cases is too small for reliable statistical analysis.

In all, we detected superficial venous incompetence in 43 (22%) of passengers. Superficial venous incompetence (varicose veins) is common and these figures are consistent with recent estimates for prevalence of venous disease in the general population. Deep venous incompetence was present in 29 volunteers (15%). This was considerably higher than we would have expected, but may simply reflect the age of the subjects that we were studying.

The recent study published by Kraaijenhagen et al found no association of DVT with travel. However, this study included few subjects who had actually undertaken any substantial amount of travel. Our volunteers undertook a median of 24 hours of long haul economy class travel during the study interval. Kraaijenhagen et al included subjects with a number of potential confounding factors such as previous venous thrombosis, malignant disease and recent surgery which were excluded in our study. Bendz et al simulated long haul flights in a hypobaric chamber and demonstrated substantially increased plasma markers of thrombosis in volunteers exposed to reduced ambient pressure. ³⁹ This study did not include a control group and needs to be repeated following actual airline travel. However, this paper suggests a possible additional mechanism resulting in thrombosis after following air travel.

Conclusion

Acknowledgements

References

2     Milne R. Venous thromboembolism and travel: is there an association? J R Coll Physicians Lond 1992;26:47-9.

3     Sahiar F, Mohler SR. Economy class syndrome. Aviat Space Environ Med 1994;65:957-60.

4     Nissen P. The so-called "economy class" syndrome or travel thrombosis Vasa 1997;26:239-46.

5     Ribier G, Zizka V, Cysique J, Donatien Y, Glaudon G, Ramialison C. Venous thromboembolic complications following air travel. Retrospective study of 40 cases recorded in Martinique. Rev Med Interne 1997;18:601-4.

6     Eklof B, Kistner RL, Masuda EM, Sonntag BV, Wong HP Venous thromboembolism in association with prolonged air travel. Dermatol Surg 1996;22:637-41

7     Ferrari E, Chevallier T, Chapelier A, Baudouy M. Travel as a risk factor for venous thromboembolic disease: a case-control study. Chest. 1999;115:440-4.

8     Kraaijenhagen RA, Haverkamp D, Koopman MMW, Prandoni P, Piovella, Buller H. Travel and the risk of venous thrombosis. Lancet; 2000: 356: 1492-3.

9     Kakkar VV, Howe CT, Flanc C, Clarke MB. Natural history of postoperative deep-vein thrombosis. Lancet. 1969; 2:230-2.

10     Negus D, Pinto DJ. Natural history of postoperative deep-vein thrombosis. Lancet. 1969; 2:645.

11     Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog. Cardiovasc. Dis. 1975; 17:259-270.

12     Coon WW. Epidemiology of venous thromboembolism. Ann. Surg. 1977; 186:149-164.

13     THRIFT Consensus Group. Risk of and prophylaxis for venous thromboembolism in hospital patients. Thromboembolic Risk Factors. BMJ. 1992; 305:567-74.

14     Nicolaides AN, Irving D. Clinical factors and the risk of deep vein thrombosis. In: Thromboembolism: aetiology, advances in prevention and management, ed. Nicolaides AN, MTP Co Ltd, Lancaster, 1975. pp 193-203.

15     Evans CJ, Fowkes FG, Ruckley CV, Lee AJ. Prevalence of varicose veins and chronic venous insufficiency in men and women in the general population: Edinburgh Vein Study. J Epidemiol Community Health 1999; 53:149-53.

16     Kniffin WD Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994;154:861-6.

17     Grady D, Wenger NK, Herrington D, Khan S, Furberg C, Hunninghake D, Vittinghoff E, Hulley S. Postmenopausal hormone therapy increases risk for venous thromboembolic disease. The Heart and Estrogen/progestin Replacement Study. Ann Intern Med 2000; 132: 689-96.

18     Sarin S, Shami SK, Cheatle TR, Bearn P, Scurr JH, Coleridge Smith PD. When do vascular surgeons prescribe antiplatelet therapy? Current attitudes. Eur J Vasc Surg. 1993; 7:6-13.

19     PEP Trials Group. Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: Pulmonary Embolism Prevention (PEP) trial. Lancet 2000 355:1295-302.

20     Robinson KS, Anderson DR, Gross M, Petrie D, Leighton R, Stanish W, Alexander D, Mitchell M, Mason W, Flemming B, Fairhurst-Vaughan M, Gent M. Accuracy of screening compression ultrasonography and clinical examination for the diagnosis of deep vein thrombosis after total hip or knee arthroplasty. Can J Surg 1998;41:368-73.

21     Cornuz J, Pearson SD, Polak JF. Deep venous thrombosis: complete lower extremity venous US evaluation in patients without known risk factors--outcome study. Radiology 1999;211:637-41.

22     Westrich GH, Allen ML, Tarantino SJ, Ghelman B, Schneider R, Laskin RS, Haas SB, Sculco TP Ultrasound screening for deep venous thrombosis after total knee arthroplasty. 2-year reassessment. Clin Orthop 1998; 356:125-33.

23     Forbes K, Stevenson AJ The use of power Doppler ultrasound in the diagnosis of isolated deep venous thrombosis of the calf. Clin Radiol 1998;53:752-4.

24     Mantoni M, Strandberg C, Neergaard K, Sloth C, Jorgensen PS, Thamsen H, Torholm C, Paaske BP, Rasmussen SW, Christensen SW, Wille-Jorgensen P Triplex US in the diagnosis of asymptomatic deep venous thrombosis. Acta Radiol 1997;38:327-31.

25     Robertson PL, Goergen SK, Waugh JR, Fabiny RP Colour-assisted compression ultrasound in the diagnosis of calf deep venous thrombosis.Med J Aust 1995; 163:515-8.

26     Krunes U, Teubner K, Knipp H, Holzapfel R. Thrombosis of the muscular calf veins--reference to a syndrome which receives little attention. Vasa 1998; 27:172-5.

27     PD Coleridge Smith, JH Scurr, JH Hasty. Venous stasis and lumen changes during surgery. Br J Surg 1990; 77: 1055-9.

28     PD Coleridge Smith, JH Hasty and JH Scurr. Deep vein thrombosis: effect of graduated compression stockings on the deep veins of the calf. Br J Surg 1991; 78:724-6.

29     Kalodiki E, Nicolaides AN, Al-Kutoubi A, Cunningham DA, Mandalia S. How "gold" is the standard? Interobservers' variation on venograms. Int Angiol 1998;17:83-8.

30     Lensing AW, Buller HR, Prandoni P, Batchelor D, Molenaar AH, Cogo A, Vigo M, Huisman PM, ten Cate JW Contrast venography, the gold standard for the diagnosis of deep-vein thrombosis: improvement in observer agreement.Thromb Haemost 1992; 67:8-12.

31     Kazmers A, Groehn H, Meeker C. Acute calf vein thrombosis: outcomes and implications.Am Surg 1999; 65:1124-7.

32     O'Shaughnessy AM, Fitzgerald DE. The value of duplex ultrasound in the follow-up of acute calf vein thrombosis. Int Angiol 1997; 16:142-6.

33     Meissner MH, Caps MT, Bergelin RO, Manzo RA, Strandness DE Jr. Early outcome after isolated calf vein thrombosis. J Vasc Surg 1997; 26:749-56.

34     McLafferty RB, Moneta GL, Passman MA, Brant BM, Taylor LM Jr, Porter JM. Late clinical and hemodynamic sequelae of isolated calf vein thrombosis. J Vasc Surg 1998; 27:50-6.

35     Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. Overview of results of randomized trials in general, orthopedic, and urologic surgery. N Engl J Med 1988; 318:1162-73.

36     Bounameaux H, de Moerloose P, Perrier A, Reber G. Plasma measurement of D-dimer as a diagnostic aid in suspected venous thromboembolism: an overview. Thromb Haemost 1994; 71:1-6.

37     Landgraf H, Vanselow B, Schulte-Huermann D, Mulmann MV, Bergau L. Economy class syndrome: rheology, fluid balance, and lower leg edema during a simulated 12-hour long distance flight. Aviat Space Environ Med 1994;65:930-5.

38     Colditz GA, Tuden RL, Oster G. Rates of venous thrombosis after general surgery: combined results of randomised clinical trials. Lancet 1986;2:143-6.

39     Bendz B, Rostrup M, Sevre K, Andersen T, Sandset PM. Association between acute hypobaric hypoxia and activation of coagulation in human beings. Lancet 2000; 356:1657-8.

Table 1

^*Difference between proportions 18% (95% CI 5.5 - 31%).

Table 2

	Number	Age	Length of stay (days)	Total flying hours	Hb	WBC	Haematocrit	Platelets	FVL (no. +ve)	PGM (no. +ve)	DD pre-flight (ng/ml)	DD post-flight (ng/ml)	Stock-ings
DVT	12 5M, 7F	67 (58 - 68)	18 (8 - 21)	21 (17 - 25)	14.2 (13.2-14.6)	6.1 (5.7 - 7.0)	0.44 (0.43-0.47)	240 (206 - 272)	2	0	<32 (<32 - 44)	31 (<32 - 55)	0
No DVT	188 71M, 117F	62 (55 - 68)	16 (13 - 27)	24 (18 - 27)	14.0 (13.3-14.8)	6.0 (5.0 - 7.1)	0.44 (0.42-0.47)	244 (216 - 285)	9	4	<32 (<32 - <32)	<32 (<32 - 32)	100
SVT	4 0M, 4F	67 (64 - 70)	18 (16 - 21)	28 (25 - 33)	13.0 (12.5-13.3)	6.3 (5.6 - 6.8)	0.40 (0.39-0.40)	264 (237 - 236)	1	1	32 (<32 - 39)	33 (<32 - 50)	4
No SVT	196 76M, 120F	62 (55 - 68)	16 (13 - 26)	24 (18 - 35)	14.0 (13.3-14.8)	6.0 (5.0 - 7.2)	0.44 (0.42-0.47)	241 (214 - 286)	10	3	<32 (<32 - <32)	<32 (<32 - 32)	96

Table 3

		DVT		No DVT		Total
No stockings		12 (10%)* (5M, 7F) age: 67 (IQR: 58-68)		104 (50M, 54F) age: 62 (IQR: 56-68)		116
Stockings		0 (0%)*		115 (34M, 81F) age: 61 (IQR: 56-66)		115
* Difference between proportions 10% (95% CI 4.8 - 16%)
		Superficial thrombophlebitis		No superficial thrombophlebitis
No stockings		0 (0%)*		116 (55M, 61F) age: 62 (IQR: 56-68)		116
Stockings		4 (3%)* 0M, 4F age: 67 (IQR: 64-70)		111 (34M, 77F) age: 61 (IQR: 55-66)		115
* Difference between proportions 3% (95% CI 0.1% - 7%)
Confidence interval for events:
	No stockings				Stockings
Event	Frequency of event		Percentage		Frequency of event		Percentage
DVT	12 of 116		10% (95% CI 4.8 - 16%)		0 of 115		0% (95% CI 0 - 3.2%)
Superficial thrombophlebitis	0 of 116		0% (95% CI 0 - 3.1%)		4 of 115		3% (95% CI 1.0 - 8.7%)

Table 4

Drug	No stockings Number of volunteers	Stockings Number of volunteers
Aspirin	9	11
Hormone replacement	8	16
Thyroxine	6	6
Antihypertensive drugs, including diuretics	10	12
Anti-peptic ulcer drugs	8	3

Figure 1

Trial profile.