Air travel thrombosis 2001

68IntroductionOn June 8th, 2001 an expert workshop on air travel throm-bosis was held in Berlin.The aim of the workshop was to accumulate presentknowledge on this phenomenon and to assess the availableevidence. This is particularly important from a scientificpoint of view, especially with respect to further studies. Ifsound scientific evidence should be available, the membersof the workshop advise the development of consensus re-commendations for daily clinical practice.DefinitionThe term “traveller’s thrombosis” comprises deep veinthrombosis (DVT), with or without pulmonary embolism(PE), that occurs during or within four weeks after travel-ling, predominantly in sitting posture. The sub-group of“air travel thrombosis” is defined as travel thrombosis thatoccurs when the main part of the travel was done by plane.Current knowledgeUntil recently, apart from some experimental studies [6],only case reports or retrospective analyses were availablefor the epidemiology of air travel DVT (previously alsotermed economy class syndrome) [8–10, 12, 14]. In 1999and 2000, three case control studies were published thatfocus on the association of travel with the incidence of deepvein thrombosis, however, only one study explicitly ex-amined the risk of air travel. In addition, the results are notuniform: In a study by Ferrari [3], who examined patientsthat were admitted to a hospital in the vicinity of the Parisairport, an odd’s ratio of 4.0 was found for long distancetravel (both air travel and ground transportation) with aconfidence interval of 1.9–8.4. This study included 160cases and 160 controls with an average travel duration of5.7 hours. The percentage of patients who developed deepvein thrombosis after long distance travel was 24.4% andthe percentage of air travellers 5.6%, while the percentageof long distance travellers in the control group was 7.5%,without further details provided on air travel. All patientswho were admitted for venous thromboembolism, as wellas the age-matched controls were examined twice with theuse of a questionnaire. The control group consisted of con-secutive patients who were admitted to the same depart-ment during the same time period for other reasons thanDVT. In the SIRIUS-Study [11], a multicenter case-controlstudy with 624 general practitioners, consecutive patientswere admitted with DVT of the lower extremities, the con-trols were age- and sex-matched. For each DVT patient,the first patient who presented with viral flu infection or anaso-pharyngeal syndrome was matched as a control. Thepercentage of long distance travel was 12.6% in patientswith DVT compared to 6.3% in the control group, result-ing in an odd’s ratio of 2.35 with a confidence interval of1.45 to 3.8.The third case control study by Kraaijenhagen et al. [5]appears to avoid recruting bias from its design: In this studypatients were included who were examined for suspectedDVT, controls comprised patients with excluded DVT. Ac-cordingly, in this study 186 patients were included with acontrol group of 602 patients. The percentage of a long dis-tance travel was 4.8% in the cases and 7.1% in the con-trols, respectively. The percentage of air travel in bothgroups was 2.2%, the absolute number was very low with4 and 13 long distance air travels, respectively. The odd’sratio for all travels in this study was 0.7 (confidence inter-val 0.3–1.4) and 1.0 for air travel (confidence interval0.3–3.0). The upper limit of the 95% confidence intervalfor the relative risk of thromboembolic events after air trav-el of 3.0 is not in sharp contrast to the study of Kraaijen-hagen and the other mentioned case-control studies [3, 11].As emphasised by Hirsh [4] an association between longdistance travels and thrombosis is possible but probablynot strong.In another study with surrogate parameters (ultrasoundexamination of the lower extremities with investigation foran apparent and inapparent deep vein thrombosis) Scurr etal. [13] examined air travellers with and without com-pression stockings and found an incidence of 12% DVT intravellers without the use compression stockings. All DVTwere clinically asymptomatic. In this study no details areprovided on diagnostic criteria, the localisation and exten-sion of DVT. Notable is the fact that none of the travellerswith compression stockings had DVT, implicating a 100%protection by compression stockings, which does not seemplausible. Another point of critic is, as mentioned by Hirsh(4), the high percentage of travellers without compressionstockings that developed DVT, which is a magnitude higherthan observed in the previously mentioned case-controlstudies. In addition, the employed design carries a poten-tial for bias during the ultrasonographic examination.A recently published study retrospectively analysed theincidence of severe PE requiring immediate medical careon arrival at Charles de Gaulle Airport, France [7]. The fre-quency of PE was clearly related to the distance travelled,with the highest incidence of 4.77 per million for those pas-sengers who had travelled more than 10000 km. The over-all incidence of severe PE was 0.4 per million passengers.