Design, Setting and Participants
We performed a cross sectional study using validated questionnaires of self-reported health status. The study sample consisted of adult asylum seekers and refugees from thirty countries including Iran (4), Iraq (7), Afghanistan (4), Pakistan (6), Vietnam (1), Bosnia (1), Bhutan (2), Georgia (2), Kazakhstan (1), Kuwait (2), Lebanon (1), Russia (1) and a number of African countries (56). The target population of this study was asylum seekers living in two direct provision centres in Sligo and Leitrim, Ireland and refugees who were living in the same community services area. The questionnaire included questions on physical health, mental health, utilisation of health care services, pre and post migratory traumatic experiences and each person's socio-demographic background.
Procedures
Participants in this study were sent a letter and consent form, informing them about the study, the questionnaire, confidentiality, and the voluntary nature of participation. Asylum seekers were assured that participation in the study would have no influence on their asylum request. The study protocol was approved by the Research Ethical Committee of Royal College of Surgeons in Ireland.
With permission from the authors, a modified version of a questionnaire developed in the Netherlands was used in this study [6]. The questionnaire was modified by the removal of sections on life-style, acculturation and social support. To minimize the risk of misunderstanding or miscommunication and to address difficulties in relation to translation or clarification on items, one of the researchers (MT) administered the questionnaire to all participants in a face to face interview and used an interpreter when necessary to overcome any language difficulties.
Sampling and sample size calculation
All asylum seekers and refugees included in this study were 18 years of age or above. As a family tend to have similar values for some of the outcomes studied, only one member per family was randomly selected on the basis of a registration code.
Offshore refugees were excluded from the study as they are granted refugee status prior to arrival in Ireland and have never been through the asylum process.
The sample size was calculated based on the results of a previous study that reported the proportion of asylum seekers and refugees with PTSD symptoms using the Harvard Trauma Questionnaire (HTQ) scoring system [7]. We planned to recruit between 53–66 asylum seekers and 27–33 refugees, assuming a difference in the proportion of patients with PTSD symptoms to be 32.4% (43.4% for asylum seekers and 11% for refugees) [7]. This would give the study a power of between 80–90% at a significance level of 0.05.
The recruitment was carried out over a three month period (February to April 2007). A random sample of asylum seekers (66 people) were selected from the register of the two direct provision centres and asked to participate in the study. For the refugee sample, Community Welfare Officers have a register of the refugees living in their catchment area, and a random sample from this register (33 people) were asked to participate in the study.
A) Health Status
Outcome Measures
We assessed three principal outcomes in relation to health: general health status; symptoms of PTSD; and symptoms of anxiety and depression. For analyses, the health outcome variables were dichotomised as follows.
1. General health status – current health status of respondents was measured according to the general health questions on the 36-item Short-Form (SF-36) [8]. The response options ranged from '5 = excellent' to '1 = poor' and were subsequently dichotomised into: 'good' (excellent, very good, good) and 'poor' (fair, poor). Although not originally developed for use in migrant populations, the SF-36 has been translated into numerous languages for use in research. However, it should be noted that its cross cultural validity has been questioned [5].
2. PTSD status – part IV of the Harvard Trauma Questionnaire (HTQ) was used [9]. It includes 30 symptoms, of which 16 were derived from the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R) criteria for PTSD and 14 describe symptoms related to the traumatic events in the refugees' lives. The format of the response options is comparable to that of the HSCL-25 (item 3 below). Individuals with a mean score on the 16 PTSD symptoms and/or on the total list of 30 symptoms = 2.5 were considered to be symptomatic for PTSD. This scale was developed for use in refugee populations and has been used in numerous studies on refugees [1, 5]. The 2.5 cut-point was chosen to make our study comparable to others, however, it should be noted that the validity of this cut point has only been demonstrated in outpatient psychiatric patients and not in a community based setting [5].
3. Anxiety and depression status – the Hopkins Symptom Check List-25 (HSCL-25) [10, 11], was used to measure symptoms of anxiety (10 items) and depression (15 items). Respondents were asked to indicate the extent to which they were bothered by each symptom in the previous week, ranging from '1 = not at all' to '4 = extremely'. Individuals with a mean score for anxiety and/or depression and/or the total list of symptoms > 1.75 were considered to be symptomatic. This instrument wasn't originally designed for use in refugee populations but has been adapted [11], and subsequently used in a number of refugee studies [1, 5]. The cut point of 1.75 was chosen so our results would be comparable with other studies. However, the validity of this cut point in community based refugees and asylum seekers of multiple nationalities is unclear.
Explanatory variables
To explore the influence of potential explanatory variables on general health, PTSD and depression/anxiety, we assessed the following:
1. Residence status – refugee or asylum seeker.
2. Pre migration stressors – Pre migration stressors (traumas) were assessed with part 1 of the HTQ, which includes 17 events (e.g. lack of food and water, being close to death) with three possible responses (experienced, witnessed, no) [9]. An additional 15 traumatic events (e.g. rocket attacks) were added to the questionnaire by Gerritsen and colleagues based on Amnesty International Annual Reports (1975–2002) [6] as well as items from HTQ part III that focus on traumatic events. This variable was dichotomised into '0 to 7 pre migration traumas' and '≥ 8 pre migration traumas'.
3. Post migration stressors – Respondents were asked about possible stressful experiences they have experienced in Ireland. The checklist included 18 problems often reported by refugees in research on post migration stressors (e.g. uncertainty about residence status, loneliness, discrimination and communication problems) [6, 12–14]. They were asked to indicate the extent to which any of the items had bothered them in the previous month ('1 = not at all' to '4 = extremely'). A mean score was calculated and dichotomised into < 2.5 and ≥ 2.5 events [15].
4. Numbers of chronic conditions – Respondents were asked to indicate for 28 chronic conditions (e.g. severe headache, diabetes) whether or not they had had this condition in the last 12 months. As done elsewhere, this variable was dichotomised into '0 or 1' and '>1' chronic condition [15].
B) Health care Utilisation
Utilisation of Health Care Services was assessed by recording the following self reported data: (i) frequency of contact with a general practitioner, outpatient medical specialist, nurse or area medical officer in the previous two months; (ii) hospital admissions or contact with mental health service in previous 12 months; (iii) use and type of prescribed and over-the-counter medication in the previous 14 days.
Analysis of Results
SPSS v.15 was used for analysis. A comparison was made between asylum seekers and refugees using two-tailed Pearson chi-square and Student's t-tests to examine differences in socio-demographic variables and other risk factors. Differences in health outcomes and the use of health care services between the two groups were examined by calculating the odds ratios (ORs), 95% confidence intervals and P values using univariable analyses.
We examined a number of potential confounding factors including socio-demographic factors (e.g. gender, age, marital status) as well as period of residence in Ireland and highest level of education completed. These variables were explored in univariable analysis with general health, PTSD and depression/anxiety as the health outcomes.
To identify factors that were independently associated with health outcomes and with use of health care services; multivariable logistic regression analyses was performed by the inclusion of explanatory variables that were significant (P < 0.05) on univariable analysis. Adjusted odds ratios (OR) were calculated using a backwards conditional multivariable logistic regression models to control for the presence of other variables (removal probability limit of P > 0.10).