Occupational Medicine Advance Access originally published online on May 22, 2006
Occupational Medicine 2006 56(5):345-352; doi:10.1093/occmed/kql026
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Morbidity in expatriates—a prospective cohort study
1 Guy's, King's and St Thomas' School of Medicine and Dentistry, London, UK
2 Foreign and Commonwealth Office, London, UK
3 King's College, London, UK
Correspondence to: Dipti Patel, 201 Wood Lane, London W12 7TS, UK. e-mail: dipti_p_patel{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
Background Expatriates comprise an important, but rarely studied subset of international travellers. This study was performed to assess the incidence of health events in an expatriate group and to evaluate factors affecting this incidence.
Methods A cohort of 2020 Foreign and Commonwealth Office (FCO) staff and partners living abroad were followed-up over 1 year. The main outcome measure was incidence of illness or injury serious enough to require consultation with a doctor. Data collection was by means of a self-administered questionnaire. Poisson regression was used to estimate the rates of health events and to test for association between health events and a number of independent variables.
Results The incidence of health events was 21%. Trauma (incidence 5%), musculoskeletal disorders (incidence 4%) and infectious disease (incidence 3%) were the principal causes of morbidity. The incidence of psychological disorders was low (1%). Of significance, employees were at increased risk of morbidity when compared to partners, with a higher incidence of health events [incidence rate ratio (IRR) 1.4, 95% CI 1.1–1.9] and psychological disorders (IRR 5.9, 95% CI 1.0–34.1). Moreover, unaccompanied employees were at increased risk of health events (IRR 1.3, 95% CI 1.0–1.7), and of traumatic injury (IRR 2.3, 95% CI 1.3–4.3) when compared to accompanied employees.
Conclusion While the morbidity in FCO personnel is low in comparison to other expatriate groups, the higher risk of morbidity in employees and unaccompanied individuals merits further research, particularly to ascertain whether work demands, isolation or risk-taking behaviour are contributory factors.
Keywords Expatriate; morbidity; overseas workers; travellers
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
Expatriates are defined as individuals who take up residence in another country predominantly for occupational purposes, and return to their original country when their assignment is complete [1]. They are a heterogeneous group, consisting of government employees, aid workers, diplomats, military personnel, journalists or businessmen. They are frequently accompanied by their families.
While overseas, expatriates, unlike other travellers, have to adapt to the host culture, have a longer duration of exposure to country-related hazards, but also have the opportunity to modify risks in their immediate environment.
Although numerous studies have assessed the health problems experienced by short-term travellers [2–5], little information exists on expatriate morbidity. Most research has focused on aid workers in the tropics, military personnel or specific conditions [6–9]. In particular, there appears to be a lack of longitudinal data on worldwide expatriate health.
The Foreign and Commonwealth Office (FCO) is one of the largest employers of British people abroad, with just over 4000 staff and dependants posted to 190 countries for an average of 3 years. To help fill some of the gaps in current knowledge, and to optimize preventive strategies for organizations expatriating their staff, a 1-year longitudinal study in the FCO expatriate population was undertaken.
The overall aim of the study was to assess the morbidity in this population. Specific objectives were to determine the incidence of health events worldwide and to examine risk factors associated with this incidence.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
This was a 1-year cohort study. The study population consisted of all FCO employees and partners (defined by the FCO as partners of officers who are legally married, or who have been living together for >1 year) who were expected to be overseas for the duration of the study. The cohort, comprising 2020 adult individuals was identified from employment records. Ethical approval for this study was obtained from Guy's and St Thomas' Hospitals National Health Service Trust ethics committee.
The main outcome measure was doctor-reported health events, and all new illnesses or injuries experienced by the study population that required a doctor consultation were included in the study. Trivial self-limiting medical conditions, check-ups, follow-ups or routine pregnancy-related appointments were not included, but acute exacerbations of previously stable conditions were.
Individuals consulting doctors for health problems were identified from medical expense records or by FCO medical staff. Individuals making a medical claim for new events were sent a structured questionnaire for completion. The questionnaire was sent to non-responders after 2 months.
Events fulfilling the study inclusion criteria were categorized according to the International Classification of Diseases 10th Revision (ICD-10) [10]. Data from the completed questionnaires, together with basic demographic information about the study population (obtained from personnel records) were anonymized and entered onto a Microsoft Access database for further analysis using Stata statistical package.
The incidence of medical claims, health events, hospital admission and medical repatriation was calculated. The incidence of illness or injury (according to ICD-10 classification) was calculated, and the proportional contribution of each ICD-10 category to total health events, major medical events, hospital admission and medical repatriation was also determined.
Poisson regression was used to test for association between health events and age, sex, marital status, employment status (employee or partner), grade (senior or junior), region of posting and previous overseas experience.
Estimates of association for each independent variable were then corrected for confounding by adjusting for age, sex, marital status, employment status, grade, region of posting, employer (FCO or another Government department) and previous overseas experience.
Poisson regression was also carried out to test for association between the main ICD-10 categories and the independent variables in the main analysis (taking into account the possible confounders already specified).
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
The general characteristics of the study population are illustrated in Table 1.
|
The study response rate was 89%. Unaccompanied individuals, and those in their 40s were significantly overrepresented in the respondent group.
The rate of new medical claims over 1 year was 41%, with 733 individuals making 834 claims.
A total of 422 (21%) new medical claims qualified for inclusion in the study. Of these, 274 (14%) were considered to be major medical events (defined as medical problems which needed specialist or hospital attention).
The mean age of individuals with doctor-reported health events was 42 years (range 23–59 years, SD 9.6).
There were no deaths or premature departures from post in the cohort during the study period, but the hospital admission rate was 4% (81), and the medical repatriation rate was 2% (43).
Table 2 illustrates the incidence of illness according to major ICD-10 category, and the proportional contribution of each ICD-10 category to total health events, major medical events, hospital admission and medical repatriation.
|
‘Injuries, poisonings and certain other consequences of external causes’ were the most common cause of doctor-reported health events (21%), and the most common cause of major medical events (29%).
Table 3 illustrates the incidence rate ratio (IRR) of doctor-reported health events according to employment status, sex, marital status, grade, age, region of posting and previous overseas experience.
|
Being an employee (adjusted IRR 1.4, 95% CI 1.1–1.9), and being unaccompanied at post (adjusted IRR 1.3, 95% CI 1.03–1.7) was associated with a significantly increased risk of doctor-reported health events.
The incidence of infectious disease was 3%. Diarrhoeal illness accounted for 73% of cases of infectious disease (incidence 2%), and only four cases of malaria were reported (incidence in malaria endemic countries was 1%).
Being posted to the Americas (adjusted IRR 4.1, 95% CI 1.7–10.3) or Africa (adjusted IRR 3.0, 95% CI 1.2–7.6) was associated with significantly increased risk of infectious disease. Employees and unaccompanied individuals had an increased, but non-significant risk of infectious disease (Table 4).
|
Individuals posted to the Americas had a significantly higher risk of diarrhoeal illness (adjusted IRR of 5.7, 95% CI 2.1–15.9), with most cases occurring in Latin America or the Caribbean. Unaccompanied individuals had an increased (but non-significant) risk of diarrhoeal illness. The incidence of diarrhoeal illness decreased with increasing age.
Of the reported health events, 20% were due to accidents (incidence 4%). Of the reported injuries, 39 (46%) were sprains or strains of joints and ligaments and 26 (31%) were fractures.
Recreational activities were responsible for the majority of accidents, with 43 (51%) injuries being as a direct consequence of a sports activity. Of the injuries, 32 (38%) were due to falls and 10 (12%) were the result of road traffic accidents.
The mean ages for injury due to road traffic accidents, sports and falls were 33 years (range 27–51 years, SD 7.6), 38 years (range 25–56 years, SD 9.1) and 42 years (range 26–56 years, SD 8.3), respectively.
As illustrated in Table 4 being unaccompanied at post was associated with a significantly increased risk of traumatic injury (adjusted IRR 2.3, 95% CI 1.3–4.3). In addition, the risk of traumatic injury progressively decreased with age.
The incidence of psychological disorders (mental or behavioural problems, ICD category F00-F99) was 1%. Cases in this category were due to affective disorders (five) or neurotic, stress-related and somatoform disorders (13).
Being an employee was associated with a significantly higher risk of psychological illness (adjusted IRR 5.9, 95% CI 1.01–34.1). Unaccompanied individuals were also at increased risk, but this was non-significant once adjusted for potential confounders (Table 4).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
This study is the first to look at all cause morbidity in a worldwide expatriate population. Over 1 year, 21% of the study population experienced new health problems, 4% required hospital admission and 2% required medical repatriation. The majority of reported health events were due to ‘injury, poisoning and certain other consequences of external causes’. The incidence of psychological disorders was low when compared to the findings in other expatriate studies [11,12].
Employment status, gender, marital status and region of posting were found to be associated with development of illness or injury, but the most significant finding was that employees were at increased risk of morbidity when compared to their partners, with a higher incidence of health events and psychological disorders. Moreover, unaccompanied employees were at increased risk of illness and traumatic injury when compared to accompanied employees.
While different methodologies make comparison difficult, the incidence of doctor-reported illness or injury and hospital admission in the FCO population was high when compared to short-term travellers [13]. A possible explanation for this includes the longer duration of exposure to country-related hazards in expatriates. The disparity could also be explained by a difference in illness behaviour in the two groups; short-term travellers can delay seeking medical treatment until they return home, and unfamiliarity with local medical facilities may make them less likely to seek medical attention while overseas.
The incidence of reported illness or injury, hospital admission and medical repatriation was lower than that reported in previous studies on expatriates. This could be due to a difference in case ascertainment and definition in the previous studies, or it could reflect differences in environmental and occupational risks; most expatriate studies have focused on aid workers, or military personnel. By virtue of their work, such groups tend to live in rural areas in the developing world. By contrast, FCO employees are posted to urban areas. Better living conditions and increased control over their immediate surroundings, mean that FCO personnel are less likely to be exposed to many of the disease and environmental hazards encountered by other expatriate groups. The reduced morbidity could also be due to the selection and pre-assignment preparation of FCO personnel.
The overall incidence of infectious diseases was lower than expected when compared to other expatriates and short-term travellers [2,5,6,14]. Reasons for this include the different case definition and inclusion criteria used in this study, more effective pre-travel advice, the greater control FCO employees have over their environment and a lower overall exposure to infectious disease due to their urban residence.
As demonstrated in other studies on expatriates and short-term travellers, trauma accounted for the majority of reported problems, and the peak incidence for trauma was in the 20- to 29-year-old age group, with a steady decrease in risk with increasing age [6,8,15]. Unaccompanied individuals were at particular risk.
Trauma only accounted for 5% of medical repatriations. This low rate could reflect a lower incidence of accidents in FCO personnel due to effective pre-travel advice and avoidance of risk-taking behaviour. It could also represent a lower exposure to occupational hazards, a decreased incidence of serious accidents, a decreased incidence of trauma due to their urban residence or the occurrence of accidents in places where local medical facilities are capable of managing the injuries sustained (urban medical care generally being more sophisticated than rural medical care).
The incidence of psychological disorders was low, and is at odds with other studies, where in excess of 10% of expatriates were found to be suffering from psychological disorders, and this category of illness accounted for the majority of repatriations or premature terminations of assignment [1,12,16]. This low incidence could represent a true lower incidence, or could be due to a selection effect, as greater care is taken when assessing fitness for posting in those with a previous history of psychological disorder. Alternatively, it could reflect a lower incidence of psychological disorders requiring medical intervention, good support networks within the FCO or a lower tendency for individuals to consult overseas. A more likely reason however is an under-reporting of psychological disorders due to perceived associative stigma and possible career implications [17,18].
Employees were at increased risk of developing illness when compared to partners, with a significantly higher incidence of doctor-reported health events and psychological disorders. The latter correlates with Anderzen's study on Swedish expatriates where employees reported a greater reduction in well-being [19]. The difference between employees and partners is likely to be even greater due to the healthy worker effect. Possible reasons for these findings may include staff having a lower threshold for consulting doctors due to a need for medical certification to cover sickness absence. Another reason for the differences in risk of illness development could be work-related issues, and would be an area worth investigating further.
Women also had higher reported morbidity when compared to men; with a higher incidence of reported illness, major medical events and psychological disorders. These findings are consistent with publications on sex differences and reported morbidity [20]. In the case of female employees, this could be a direct consequence of their position within the FCO (47% of the female employees in this study were employed in secretarial and office support grades compared to 7% of men).
Being unaccompanied at post was also associated with increased risk of morbidity; with unaccompanied individuals having a significantly higher incidence of both doctor-reported health events and traumatic injury. The results conform with findings in other studies, suggesting that being unaccompanied overseas may result in increased risk-taking behaviour, and that being accompanied is protective [21–23].
This study is well placed to add to current knowledge of expatriate health. Particular strengths include the prospective study design, the high response rate and the limited loss to follow-up. Additionally, the utilization of an internal comparison group allowed the assessment of health events in relation to a variety of risk factors, and avoided bias due to the healthy cohort effect.
One possible limitation was that information on potential risk factors for morbidity was obtained from personnel records. While this meant that information was available for the whole cohort, as the original collection of this data was for personnel purposes, the investigator had no control over the information collected, and no way of ascertaining whether the information had changed. This potential bias was minimized by validating all the data against the ‘diplomatic service list’ and medical records. Therefore, any misclassification of exposure was likely to be small and non-differential, resulting in an ‘underestimation’ of the effect of the exposure.
Another potential limitation was that the main outcome measure was doctor-reported health events. This meant that only illness or injury serious enough to require consultation with a doctor was analysed, with the consequence that the results could have been affected by medical culture and care-seeking behaviour.
Further information on the outcome of interest was obtained by means of a self-administered questionnaire, which was piloted prior to the commencement of the study. The use of a self-administered questionnaire meant that the potential to yield accurate data on sensitive information was greater, and the possibility of recall bias was minimized by use of a longitudinal study design. Information bias of the outcomes of interest was reduced further by use of objective and closed-ended questions and, where possible, data obtained from the questionnaire were validated against FCO medical records. Finally, observer bias during the transcription and coding of the questionnaire was reduced by blinding the investigator to exposure data (the investigator's decision to include and categorize a reported health event was made on the basis of information on the health event only, the investigator did not have any demographic information when making this decision).
Age, sex, employment status, marital status, main employer, grade, region of posting and previous overseas experience were considered to be potential confounders in this study, and they were controlled for by multivariate analysis. One potential flaw was that information on lifestyle factors was not obtained. However, by using an internal control group there were likely to be few important lifestyle differences between the various exposure groups; therefore, any confounding due to these factors was likely to be relatively weak [24].
This study has determined the incidence of health events in an expatriate population, analysed factors associated with morbidity and identified certain vulnerable groups. The overall findings suggest that the FCO has effective selection and pre-travel procedures, and good support arrangements for their employees and dependants. However, the higher risk of morbidity in employees and unaccompanied individuals warrants further research, particularly to ascertain whether work demands could be a contributory factor. For unaccompanied individuals, the role of factors such as isolation, lifestyle and risk-taking behaviour also needs to be assessed.
|
Conflicts of interest |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
None declared.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConflicts of interestReferences |
|---|
- Foyle MF, Beer MD, Watson JP. Expatriate mental health. Acta Psychiatr Scand 1998;97:278–283.[Web of Science][Medline]
- Cossar JH, Reid D, Fallon RJ et al. A cumulative review of studies on travellers, their experience of illness and the implications of these findings. J Infect 1990;21:27–42.[CrossRef][Web of Science][Medline]
- Ahlm C, Ludberg S, Fesse K, Wistrom J. Health problems and self-medication among Swedish travellers. Scand J Infect Dis 1994;26:711–717.[Web of Science][Medline]
- Hill DR. Health problems in a large cohort of Americans travelling to developing countries. J Travel Med 2000;7:259–266.[Web of Science][Medline]
- Steffen R, Rickenbach M, Wilhelm U, Helminger A, Schar M. Health problems after travel to developing countries. J Infect Dis 1987;156:84–91.[Web of Science][Medline]
- Bernard KW, Graitcer PL, van der Vlugt TB, Moran JS, Pulley KM. Epidemiologic surveillance in Peace Corps volunteers: a model for monitoring health in temporary residents of developing counties. Int J Epidemiol 1989;18:220–226.
[Abstract/Free Full Text] - Stenbeck JL. Health hazards in Swedish field personnel in the tropics. Travel Med Int 1991;9:51–59.
- Steffen R, Desaules M, Nagel J et al. Epidemiological experience in the mission of the United Nations Transition Assistance Group (UNTAG) in Namibia. Bull World Health Organ 1992;70:129–133.[Web of Science][Medline]
- Gambel JM, Drabick JJ, Martinez-Lopez L. Medical surveillance of multinational peacekeepers deployed in support of the United Nations Mission in Haiti, June–October 1995. Int J Epidemiol 1999;28:312–318.
[Abstract/Free Full Text] - World Health Organisation. International Statistical Classification of Diseases and Related Health Problems—10th Revision, Volume 1. Geneva, IL: World Health Organisation, 1992.
- Dupont RL, Valk TH, Heltberg J. Psychiatric illness and stress. In: DuPont HL, Steffen R, eds. Textbook of Travel Medicine and Health, 2nd edn. Ontario, CA: BC Decker Inc., 2001; 365–370.
- Peppiat R, Byass P. A survey of the health of British missionaries. Br J Gen Pract 1991;41:159–162.[Web of Science][Medline]
- Getz L, Larssen KE, Dahl B, Westin S. Health problems in Norwegians travelling to distant countries. Scand J Prim Health Care 1990;8:95–100.[Medline]
- Steffen R, van der Linde F, Gyr K, Schar M. Epidemiology of diarrhoea in travellers. J Am Med Assoc 1983;249:1176–1180.
[Abstract/Free Full Text] - Fairhurst RJ. Accidents and the traveller. Travel Med Int 1996;1:33–44.
- De Haan BB. Stress and psychological issues in humanitarian activities: the experience of the International Committee of the Red Cross. In: Proceedings of the 5th International Conference of Travel Medicine, 1997; Geneva. Atlanta: International Society of Travel Medicine, 1997.
- Glozier N. Workplace Effects of the stigmatization of depression. J Occup Environ Med 1998;40:793–800.[CrossRef][Web of Science][Medline]
- Stuart H. Stigma and work. Healthc Pap 2004;5:100–111.[Medline]
- Anderzen I. The internalisation of work. Psychophysiological predictors of adjustment to foreign assignments [theses]. Stockholm: Karolinska instituter, 1998.
- Verbrugge LM. Gender and health; an update on hypotheses and evidence. J Health Soc Behav 1985;26:156–182.[CrossRef][Web of Science][Medline]
- De Graaf R, van Zessen G, Houweling H, Ligthelm RJ, van den Akker. Sexual risks of HIV infection among expatriates posted in AIDS endemic areas. AIDS 1997;11:1173–1181.[CrossRef][Web of Science][Medline]
- Striker J, Luippold R, Nagy L, Liese B, Bigelow C, Mundt K. Risk factors for psychological stress among international business travellers. Occup Environ Med 1999;56:245–252.
[Abstract/Free Full Text] - Anderzen I, Arnetz B. Psychophysiological reactions to international adjustment. Psychother Psychosom 1999;68:67–75.[CrossRef][Web of Science][Medline]
- Siemiatycki J, Wacholder S, Dewar R et al. Smoking and degree of occupational exposure: are internal analyses in cohort studies likely to be confounded by smoking status? Am J Ind Med 1988;13:305–308.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||