Health and Quality of Life Among Resettled Syrians in Norway

Participants:

A complete list of potential participants meeting Inclusion criteria (see relevant section in the submitted registration) was obtained in collaboration with the Norwegian National Registry administered by the Norwegian Tax Administration. The complete list consisted of 14350 individuals which constituted the study's sampling frame. A simple random probability sample of 10 000 individuals was selected from the sampling frame (sampling fraction = 0.7). The actual sampling was conducted without any involvement from members of the REFUGE-I research team.

Data collection:

All sampled participants will be sent a postal package to their registered residential address in Norway. The package will contain a prepaid return envelope, the survey questionnaire and a cover letter in Arabic describing the purpose of the study, its voluntary nature, what potential participants can expect if they do participate, and how issues surrounding confidentiality and anonymity will be handled. The cover letter also highlights that survey data will be linked to registry data as part of the longitudinal study. Potential participants will be enabled to give written informed consent to participation at the bottom of the cover letter and instructed to enclose the written consent with the completed survey questionnaire in the prepaid return envelope. Data from participants returning the questionnaire without also submitting written consent to participation will be deleted in accordance with agreements made with the Regional Committees for Medical and Health Research Ethics (REC) - Region South East which granted the ethical approval for the study.

The survey questionnaire is eight pages long and expected to take about an hour to complete. All instructions and questions in the questionnaire are in Arabic. Standard and pre-existing scales in Arabic were used whenever available and Arabic translations were done by professional translators in collaboration with an Arabic speaking reference group for scales/items that did not already exist in Arabic. Potential participants will have about 8 weeks to complete and return the questionnaire with written consent. One postal reminder, together with a blank copy of the survey questionnaire and a prepaid return envelope will be sent out to participants who have not returned the questionnaire by a preset deadline - about two weeks before the anticipated closing date for data collection (approximately January 15, 2019). This is in accordance with Norwegian rules and regulations governing research with voluntary subjects which prohibits more than one reminder being sent out.

Analysis:

Three main articles are stipulated to be submitted for publication from the data collected through the survey questionnaire of REFUGE-I. Article one will report descriptive statistics on the main outcome variables: depression; anxiety; and symptoms of posttraumatic stress disorder across key demographic and background variables in the sample, and describe the flow of participants and discuss potential selection bias issues. Article two will provide descriptive statistics on somatic pain, sleep difficulties and perceived general health, and report findings across the same demographic and background variables as article one and across the main outcomes of article one (depression, anxiety and symptoms of posttraumatic stress disorder). Article three will present statistics on reported quality of life across demographic and background variables, and across the outcome variables of article one and two.

Alternatively, one larger article will be submitted for publication incorporating all of the themes outlined in the three articles above.

Specific analysis plans for each article/theme are described below:

Article I:

Title: "Prevalence of common mental health disorders in adult Syrian refugees resettled in Norway between 2015 and 2017: a cross-sectional survey questionnaire study".

In order to compare results to a parallel study on newly resettled adult refugees in Sweden conducted by one of our collaborating institutions, Article I will use similar outcome and predictor variables as this study. The main outcome variables will be:

Prevalence of symptom-based posttraumatic stress disorder (PTSD) as measured through the first 16 items in the section on trauma symptoms (section IV) in the Harvard Trauma Questionnaire (HTQ). A mean-item score ≥ 2.06 will define a "checklist-positive" PTSD case. To be included in analyses, participants must have answered 14 or more of the 16 items on the HTQ scale.

Prevalence of symptom-based depression and anxiety as measured through the two subscales on depression and anxiety in the Hopkins Symptom Checklist (HSCL-25). A mean-item score ≥ 1.80 and ≥ 1.75 will define a "checklist-positive" depression and anxiety case, respectively. To be included in analyses, participants must have answered ≥ 23 items on the 25-item scale.

The outcome variables above will be summarized for the whole sample population and across the following demographic and background variables:

Gender Age; split into age groups: 18-29; 30-39; 40-49; 50-64; >64. Education; split into the categories: 0-9 years; 10-12 years; >12 years Civil status; split into the categories: Married; Unmarried; Divorced/separated/widow(er); Other Arrival status; split into the categories: Asylum seeker; Quota/resettlement refugee; Family immigration; Other Year of arrival/immigration to Norway; split into: 2015; 2016; 2017. Refugee trauma history measured through the Refugee Trauma History Checklist (RTHC). The two scales contained in this checklist will be combined - that is, the period will be "prior to arriving in Norway".

Post-migration stress measured through seven main items of the Post-migration stress scale thought to tap into the relevant domains of post-migration stress. Answers to each item will be dichotomized so that those answering "often" or "very often" are classified as having had this reaction often (i.e. positive for that post-migratory stress reaction)

Prevalence estimates will be presented with 95% confidence intervals (95% CIs). Odds ratios (ORs) with 95% CIs will be used when summarizing the main outcome variables across refugee trauma history and post-migration stress. More specifically, the odds of having a symptom-level above vs. below the preset cut-off value for the main outcomes will be calculated for each binary (Y/N) item on the trauma history and post-migration stress scales (eight and seven items for the two scales, respectively). The positive answer ("Yes") will define exposed for each item when reporting the ORs.

In addition, descriptive statistics on the prevalence of refugee-related possible traumatic events will be presented. Descriptive statistics (distribution, mean and variance) on the seven domains of the complete 24-item post-migration scale will be presented and compared to the study in Sweden. The prevalence of comorbidity between checklist-positive anxiety, depression and PTSD will also be reported with 95% CIs.

The number of participants contributing data to a specific statistics will be reported. The study will report, compare and discuss statistics on the participants vs. the sample frame on all available parameters for both groups in order to evaluate selection bias issues (anticipated available parameters include: age, gender, civil/marital status, permanent vs temporary residence status, number of registered mobile phone numbers).

Article II:

Title: "Subjective health and pain, and sleep difficulties in adult Syrian refugees resettled in Norway between 2015 and 2017: a cross-sectional survey questionnaire study".

Main outcomes:

Somatic pain (including pain in muscles and joints and general pain), measured as described under Outcome Measures. Statistics on pain in muscles and joints will be aggregated (i.e. all items will be combined) and reported as the percentage of participants in each of three categories: Not troubled with pain in muscles and joints (Not troubled reported for all items); Somewhat troubled with pain in muscles and joints (Somewhat troubled reported for at least one item); Very troubled with pain in muscles and joints (Very troubled for at least one item). Participants must have answered at least four of the five items (if all answers are Not troubled), or Somewhat troubled/Very troubled on at least one item to be included in analyses. Statistics on general pain will be presented in an identical manner (i.e. the percentage in each of three categories, aggregated across all items on general pain) Sleep difficulties, measured as described under Outcome Measures. The overall mean-item score will be reported, as well as the estimated prevalence of insomnia, defined by the following criteria: a score of 3 or above on at least one of the first 4 items + a score of 3 or higher on at least one of the last 2 items. Participants must have answered at least five of the six items to be included in analyses when analyzing the mean-item score of the scale. When analyzing the prevalence of insomnia, participants must have answered at least three of the first four items and one of the last two items; or, they must have a score of 3 or above on at least one of the first four items and a score of 3 or above on at least one of the last two items to be included in analyses.

Perceived general health, measured as described under Outcome Measures. Statistics on perceived general health will be reported as the percentage of participants in each answer category (with 95% CIs). Question one on overall health will be dichotomized with Very good/Good as one category and the remaining three as one category (Don't know will not be included). Question two on functional impairment will use all three answer categories (Don't know will not be included).

The outcome variables above (with 95% CIs) will be presented for the whole sample combined, and for the sample split on demographic and background variables 1-6 as outlined under the description of article I above. In addition, the outcome variables will be presented across the main outcome variables of article I: depression; anxiety; and PTSD (all binary). Furthermore, perceived general health will be reported across outcome variable one: pain in muscles and joints (3 categories) and general pain (3 categories); and outcome variable two: sleep difficulties (binary - insomnia vs. not). The number of participants contributing data to a specific statistics/analysis will be reported.

Article III:

Title: "Quality of life in adult Syrian refugees resettled in Norway between 2015 and 2017: a cross-sectional survey questionnaire study".

The main outcome variable of Article III will be quality of life (QoL) measured through The World Health Organization Quality of Life assessment (WHOQOL-BREF). All four domain-scores in the scale (physical health; psychological health; social relationships; and environment) will be used as outcome in analyses. In order to compare with relevant prior studies, the domain scores will be converted as needed according to the method described by the WHO. That is, results on quality of life will be presented in three different ways: as raw scores, converted to a 4-20 scale, and converted to a 0-100 scale. In order to be included in analysis for a certain domain, participants must have answered a minimum of six, five, two, and six items for the domains: physical health, psychological health, social relationships, and environment, respectively.

QoL scores (with standard deviation, SD, and 95% CIs) for the four domains will be presented for the whole sample combined, and for the sample split on demographic and background variables 1-6 as outlined under the description of article I above. In addition, QoL scores (with SD and 95% CIs) will be presented across the main outcome variables of article I: depression; anxiety; and PTSD (all binary); as well as the main outcome variables of article II: muscle/joint pain and general pain, both variables split into three categories: Not troubled (for all 5 items); Somewhat troubled (at least one item); Very troubled (at least one item); sleep difficulties (binary - insomnia vs. not); and perceived general health (binary for overall health, and three categories for functional impairment).

Furthermore QoL scores from REFUGE-I will be compared with scores from other refugee populations, and with QoL scores in the general population around the world. The number of participants contributing data to a specific statistics/analysis will be reported.

Ethical issues:

In the event that participation in the study triggers substantial adverse reactions in any of the participants, a help-line will be set up by the research team in collaboration with clinical personnel with experience on refugee mental health. Participants will be informed that they can send a text message to the help-line's phone number, and that the clinical personnel will call them back with an Arabic interpreter as soon as possible. Arrangements will be made with a professional Norwegian interpreter firm to ensure the availability of an interpreter within a reasonable time frame (hours). This service will be available during regular work hours. During off-hours and weekends, participants will be instructed to contact Norway's National out-of-hours medical service for immediate medical assistance and given the necessary information on how to do so.