Building low intensity psychosocial support for Syrian refuge families in Istanbul

Background: The study is located in Istanbul, Turkey, where more than 750,000 Syrian refugees reside, largely in urban settings. It develops and pilot tests a novel model for helping urban refugee families in settings with limited to no access to evidence-based mental health services for refugees, by delivering a transdiagnostic family intervention for common mental disorders in health and non-health sector settings using a task-sharing approach. This case study addresses the following question: How can we address the common mental disorders of both children and parents, and support protective family resilience processes, through a low intensity trans-diagnostic family support intervention? Discussion: The rapidly growing scale of humanitarian crises requires new response capabilities geared towards addressing populations with prolonged high vulnerability to mental health consequences and limited to no access to mental health, health, and social resources. We faced multiple challenges in conducting this research including: 1) identifying local academic partners with research capacity, including in implementation science; 2) lack of culture of partnership between academics and humanitarian organizations; 3) getting local clinicians to embrace on a task-sharing model; 4) cultural competency of local and U.S. partners for refugee population; 5) getting local academics to focus on humanitarian emergency; 6) planning for a family intervention that would work with families with rigid gender role perspectives; 7) multiple social and economic problems that could not be solved, such as children working; 8) engagement challenges due to high demands on families. Through the research process, the research team learned lessons concerning: 1) building a coalition of academic and humanitarian organization partners; 2) investing in research capacity building of local partners; 3) working in a community-collaborative and multi-disciplinary approach to best understand and address socio-cultural, contextual, practical and scientic challenges needed to develop and implement the new family support model. Conclusion: Conducting research in humanitarian emergency settings calls for signicant attention to building a coalition of academic and humanitarian organization partners, investing in research capacity building of local partners, and working in a community-collaborative and multi-disciplinary approach.

The war in Syria has displaced over 5.6 million persons to neighboring low and middle-income countries (LMIC). 3.6 million Syrian refugees live in Turkey, now the world's largest refugee hosting country, and only about 2.85% live in refugee camps [1]. With limited access to housing, work, information, food, education, and health and mental health care, they are especially vulnerable [2]. Differences in language and culture make integration a challenge [3]. Syrians do low wage work with no bene ts, and are accused by some locals of stealing their jobs and causing crime [4][5]. Forty percent of school-aged children are not enrolled in Turkish schools [6] with child labor being a driving factor [7].
Syrian refugees are at high risk for CMD due to war trauma compounded by displacement stressors. They experienced high rates of con ict-related violence and family loss and separation, followed by the daily stressors of displacement, such as lack of resources, discrimination, and loss of social networks, limited livelihood options, and uncertainty about their future [8][9]. Several prior studies assessed the mental health of adult Syrian refugees in Turkey and found rates of common mental disorders in adults between 23% and 42% [10][11][12]. Syrian refugee children have high exposure to severe traumatic events and nearly half have PTSD symptoms [13]. Most Syrian refugees show little knowledge and awareness of mental illness, self-care strategies, or clinical treatment, and express high stigma towards mental health problems [14].
The scienti c and professional mental health literature on Syrian refugees has largely focused on PTSD through an individual lens. Despite the family orientation of Syrian culture, there was little evidence that family interventions were being developed or deployed for Syrians. We initially conducted longitudinal qualitative research with 30 Syrian refugee families so as to better characterize the family stressors and coping mechanisms [15]. We concluded that many families may be able to bene t from a family intervention approach and this called for developing family focused interventions conducted by community-based lay providers.

Research Study
This project uses a low intensity intervention approach to develop and test for feasibility of Family Support (FS), a novel trans-diagnostic intervention for refugee families with common mental disorders.
FS is delivered in a multiple family group format by community workers at NGOs including family and community health workers. Our family model is informed by the theoretical frameworks underlying its two major intervention components. First, the conceptual framework for Family Support (FS) is based upon cognitive-behavioral theory as applied in the Problem Management Plus (PM+) intervention (de ned later) via evidence-based techniques focused on stress management and behavioral activation [16]. Second, the conceptual framework for Family Support (FS) is also based on family resilience theory, which explains family protective processes that can ameliorate the negative consequences of hardships and challenges and enable healing and growth in families [17][18].
This research was designed to ask and answer several research questions, which are needed to advance the eld (see Table 1). To answer these questions, we organized the research around three speci c aims. Aim 1 form a Family Support Design Team (FSDT) to develop the family support (FS) intervention for implementation in community sites using a four-session multiple family group format. Aim 2 pilots FS with families in community and clinical sites, and then through observations and qualitative interviews, assesses FS's feasibility, delity, the impact of context and local capacity, the experiences of intervention delivery, and practitioner and organizational perspectives on scale up. Aim 3 conducts pre, immediate post, and 3-month post assessments of the refugee families who received FS in all sites, to demonstrate the kind of pre-post changes that have been reported for comparable interventions and to determine key parameters of interest with su cient accuracy and precision. Table 1 Domain Research Questions Intervention How can we address the CMDs of both children and parents of displaced families and support protective family processes through a low intensity trans-diagnostic family support (FS) intervention?
Uptake How can the facilitators and community organizations best-overcome obstacles to promote engagement and retention in the FS groups?
Intervention Acceptability Were the family members satis ed with the FS intervention content, delivery, and length? How did it t with the sociocultural, environmental, and organizational contexts?
Intervention Feasibility Was FS feasible as indicated by recruitment, attendance, retention, and delity, safety for participants and providers, and completion of outcome assessments?

Implementation
What do the families and facilitators think about offering FS to more refugee families? What are the opportunities or obstacles and how can they be best addressed to facilitate scale up?
CMD Outcomes Were there less depression, anxiety, and traumatic stress symptoms as a function of FS at 3-month assessments? Were these symptoms less after 3-months-post compared with immediate-post assessment?

Family Outcomes
Was there improvement in family members' knowledge and attitudes regarding responses to adversity, family support, and family problem solving, and accessing external resources as a function of FS at 3-month post assessments?

Constructs & Measures
What constructs and measures are the best ts for evaluating a family support intervention in an LMIC?
The academic medical centers which we visited in Istanbul, were not experienced either in conducting NIH sponsored research, research with Syrian refugees, or working with Syrian refugees. We initially chose one academic medical center partner, but due to obstacles we encountered it was necessary to nd another partner who was more experienced in working with Syrian refugees and more open to conducting research concerning refugees. We looked for professionals with both scienti c and practical experience in working with refugees and were able to identify another academic medical center partner. This partnership also led us to engage a new partner, the Turkish Red Crescent, in addition to our academic medical partner and NGO partners. This existing combination of mentioned institutions is currently enabling us to successfully implement the project.
This research should result in an adapted and pilot tested low intensity family support model and preliminary pilot data that can inform the development of a follow-up larger scale study in low resource settings. The ndings will also build knowledge on implementation science for refugees in low resource setting and LMIC's. NOTE: At the time of writing, the FS intervention has been developed, the measures chosen, the facilitators trained, and we conducted run-through and practice sessions with 42 families. We are currently conducting the pilot testing and follow-up assessments with 72 families.

Discussion
Scienti c Importance of the Research There is an urgent need for research to advance evidence-based brief psychosocial interventions that can be scaled up during humanitarian crises. The World Health Organization's Department of Mental Health and Substance Abuse, led by Dr. Mark van Ommeren, has been developing low intensity cognitive behavior therapy (LICBT) to help address refugees' mental health needs. They developed Problem Management Plus (PM+), which is a brief, basic, one-on-one, paraprofessional-delivered version of LICBT for adults in communities affected by adversity [19][20].
PM + incorporates stress management and individual problem solving [19]. In con ict-affected Pakistan, a pilot comparing PM + to enhanced treatment as usual had high uptake, with 73% completing all sessions, and showed improvement in traumatic stress and functioning [19]. When WHO conducted a larger scale randomized controlled trial (RCT) of PM + in Pakistan, it demonstrated clinically signi cant reductions in anxiety and depressive symptoms at 3 months among adults impaired by psychological distress [20].
Psychosocial interventions are needed for refugee families and humanitarian responses, both because parents and children are experiencing common mental disorders (CMD), and because supporting family protective processes and learning self-care strategies can ameliorate the impact of CMD symptoms and stress caused by trauma and displacement. Syrian refugees, like many others, are strongly shaped by the family context. In a receiving country with limited resources for refugees, positive psychosocial outcomes for children and adults depend to a great extent on their families, yet refugee families nd few empirically based services geared toward them [21][22]. Refugee families often demonstrate resilience and strengths, but in highly adverse circumstances these processes can be strained or overwhelmed [23], FS interventions can bolster family protective processes.
We previously developed and evaluated a multiple family support and education group model for Bosnian refugees called Coffee and Families Education and Support (CAFES), which in a US National Institute of Mental Health (NIMH)-funded RCT was shown to have high uptake and to improve mental health help seeking, with depression and family comfort talking about trauma mediating the intervention effect [24]. Based on our community engagement and current study of Syrians, we concluded that the multiple family approaches also ts well with Syrian culture. Of note, this type of family intervention is designed to include adolescents age 12 and above, with time allotted in each session for adolescents and adults to meet in separate breakout groups [25].
The science of implementation and dissemination (SID) can help address the burden of mental health problems in refugees in LMICs by moving evidence-based interventions from more controlled to naturalistic settings. This is a pressing challenge for studying mental health interventions during humanitarian crises [26][27]. In such settings, SID research is needed to collect information on potential barriers and facilitators to real world implementation, as well as on speci c implementation strategies [28]. Regarding PM+, formative research is recommended to help adapt the intervention to the local sociocultural context [29] and to develop a novel family intervention.
Our plan is to share the research ndings through publication in peer-reviewed journals so as to contribute to developing the science in this area. But given the urgency and demands of the humanitarian situation, we also intend to share the research ndings through several two-page practically-focused research briefs that accessible and engaging for NGO's, governments, trans-governmental organizations, and practitioners. Overall, our aim is to increase the awareness of refugee family issues, and to promote interventions that are at the very least family friendly, but also family focused which might be a more feasible approach in speci c socio-cultural contexts. We believe that there is a signi cant shortage in this area, and we would like NGO's and governments to make this more of a focus.
Family interventions for refugees and migrants are also very relevant in low and middle-income countries that are not facing a humanitarian crisis, and in high-income countries, especially in low resource settings. Thus, we believe that developing the science in the humanitarian emergency will have signi cance beyond the crisis settings. For example, in the U.S. there are 10.5 million illegal migrants [30] a large number of migrant families from Mexico and from Latin America who are facing high stress due to the threat of deportation on top of the ordinary exposure to adversity. A family support intervention delivered in community-based organizations, including religious facilities or churches, could be a very good t to address their capacities to address various stressors and risk for CMD's. The interventions are held on weekly bases and on weekends when most of the participants especially fathers don't work. The data collection started in December 2019 and will end by October 2020.
Research Methods Aim 1 activities included convening the family support design team (FSDT) and drafting the FS manual. Aim 2 involve partnering with the 6 community and clinical sites, training the workers, having each worker administer the intervention to 72 families, and then conducting observations and qualitative interviews to assess implementation issues. Aim 3 involved conducting pre, immediate post, and 3-month post assessments of the 72 families who received the intervention. Regarding Aim 3, our primary hypothesis is that FS primary participants will report fewer symptoms of depression, anxiety, and traumatic stress from baseline to immediate post and 3-month post. Our secondary hypothesis is that FS family members will report improved family members' knowledge and attitudes regarding responses to adversity, family support, family problem solving, and accessing external resources, and that these will explain improvements in CMD symptoms. In addition, we will also study engagement and retention of families in FS, both as outcomes and as possible determinants of the hypothesized changes. We will conduct analyses: 1) to demonstrate the feasibility of the implementation and evaluation methods; 2) to explore patterns of attendance and retention to FS groups, to inform the researchers in making modi cations; 3) to demonstrate the kind of pre-post changes that have been reported for comparable interventions, for both primary participants and family members; 4) to determine key parameters of interest with su cient accuracy and precision.

Challenges to Research
The researchers faced multiple program development challenges in conducting this research in the humanitarian crisis setting, even beyond those challenges generally encountered in LMICs or lowresource settings.

Non-Existent or Weak Partnerships
We were able to identify both academic centers and humanitarian organizations, but it was di cult to nd those academic centers that had partnerships with humanitarian organizations, or vice versa. Syrian refugees were generally not a focus of interest for service or research by academic medical centers in Istanbul. It was di cult to identify local academic partners who had research capacity in mental health, including implementation science, which would be necessary for investigating family support interventions. However, we found that the Turkish Red Crescent (TRC) was interested in building mental health research and implementation science capacity in the Turkish Red Crescent, so we focused the supplement award on the TRC.

Lack of Focus on Task-Sharing
It was also challenging to nd a psychiatrist who would embrace the task-sharing model, which aims to deliver interventions that do not depend upon psychiatrists as the service providers. Over 6 months, the U.S. and Kosovar investigators were able to identify a child and adolescent psychiatrist, Dr. Vahdet Görmez, who is also a certi ed trainer in cognitive behavioral therapy (CBT), conducted research with Syrian adolescent refugees in school setting in Istanbul [31][32] who became our primary partner. The main argument in deciding to collaborate with Dr. Görmez was his practical experience in working with Syrian refugees in school setting and his interest in refugee mental health. He was identi ed by conducting literature review on refugee mental health in Turkey, following which he was contacted and included in the project.

Insu cient Language and Cultural Competency
Another important challenge was that both the U.S. investigators and the local Turkish investigators faced limitations of language and cultural competency for the refugee population. Therefore, we hired two Syrian persons as the project manager and assistant who were able to help us to overcome these hurdles due to language and cultural competency, familiarity with the community and the local NGOs.
The Syrian project personnel were identi ed with help of Dr. Görmez who had experience in working with the staff in previous psychosocial projects. During the implementation process of the project we met many other capable persons in the Syrian community within collaborating organizations which further enabled us to overcome the language and cultural gaps.
Fit with Families One of the major challenges that we faced was planning for a family intervention that would work with families who had cultural values and practices around gender relations that were very strict in terms of women and men interacting and sharing the same space. Another challenge was that these families faced many burdens and priorities and it was di cult to engage them due to high demands on their time.
Another major challenge is that the families were still exposed to major social and economic problems that could not be solved, such as child labor. Another challenge was involving fathers in all sessions as they were either working or having other obligations. In addition, most of the families involved in the run through of FS intervention had a child working.

Refugee Policies
Just prior to the time of implementation in October 2019, the Turkish government began to more strictly enforce policies regarding the residential permits of Syrian refugees. When refugees came to Turkey, they were allowed to live in a particular region where they were registered, but could not move to another, such as Istanbul to get work. Stricter enforcement of policies includes deportation to the city where they are registered or deportation to Syria if they lack documentation, causing fear among refugees.
We tried to develop a shared understanding with the partner organizations and the refugees they served through regular consultations to come up with problem-solving strategies or solutions in the FSDT meetings and in consultation with the partner organizations. This demanded a great deal of exibility and patience from the research team members. The research process was slowed down at several junctures in order to deal with these sorts of issues. For example, regarding settling on a format for convening both male and female family members which would be acceptable to Syrians with a range of gender-related values and practices.

Research Strategies
Addressing the unique challenges faced by the research team, required additional strategies.

Coalition Building
First, we built a coalition of academic and humanitarian organization partners and worked on strengthening relations between them. This coalition included an academic medical partner, six small NGO's working with Syrian refugees and the Turkish Red Crescent, one of the major actors in Turkey in the provision of various services to refugees.

Research Capacity Building of Local Partners
Two, we invested in building the research capacity building of the local partners, both academic and NGOs. Through the research project, we built the research capacity of several faculty members at Medeniyet University in Istanbul. In addition, we developed and were awarded a supplement with TRC to focus on improving their research capacity.

Engaged Multiple Perspectives
Three, we utilized a family support design team to draw upon multiple perspectives and adapt existing interventions and design new components. The FSDT convened twice for three-day face-to-face meetings in Istanbul. The FSDT was co-led by Weine, Arenliu, and Gormez, and included other LMIC researchers as well as 8 additional community advocates, nurses, and physicians. Flexible Intervention Design We designed a exible structure for the intervention to accommodate gender considerations including use of the Internet and mobile technology. For example, to accommodate families who preferred to sit all together rather than to mix with other families we provided sitting arrangements in which families were grouped as in a restaurant. In addition, each group had one male and female facilitator where female facilitators reinforce women participation as we were concerned that women would not speak out in mixed groups or in presence of husbands. Furthermore, we produced easily accessible videos such as stress reduction exercises to be used by members who might have not attended the session and as reminder of exercises to be done at home.
The overall aim was to write a model that was brief and simple enough to be delivered by community workers and nurses. The FSDT rst met face-to-face for three days in month 2 in Istanbul, followed by regular Skype calls and ongoing e-mails, and a follow-up three-day meeting in month 6. Draft materials were prepared and distributed in advance, in English and Arabic and decisions were made by consensus after deliberation.

Expand Communication and Collaboration Between Researchers and End Users
Regarding the TRC, to address the needs we jointly identi ed, we collaboratively developed a plan to foster evidence-based policy and program development in mental health for refugees by expanding communication and collaboration between the researchers and end users, including the TRC and other NGOs, including practitioners, managers, and policy makers. The speci c aims are to: 1) Convene a refugee mental health implementation research group composed of researchers and end users in the TRC that will collaboratively: a) conduct trainings on the state-of-the-science in implementation science research (e.g. strategies for implementation, dissemination, and evaluation of effective mental health interventions); b) disseminate research ndings on proven mental health assessments and interventions; and c) build a model for best impacting the TRC organizational structure, climate, culture, and processes regarding the implementation, dissemination, and evaluation of effective and evidence based mental health interventions.

Conclusions
Through the research process, the team learned several key lessons regarding conducting research in a humanitarian crisis setting. One, it is helpful to build a coalition of academic and humanitarian organization partners who can ensure that the research is focused on issues that matter for service providers and recipients. Two, it is helpful to invest in research capacity building of the local partners to strengthen research implementation and facilitate dissemination and scaling up. Three, utilizing a community-collaborative and multi-disciplinary approach can help to best understand and address multiple key socio-cultural, contextual, and scienti c challenges needed to develop, adapt and implement the family support model. Although the pressing needs of Syrian refugees called for urgent action, it was bene cial to invest in and to allow su cient time for these three activities.
List Of Abbreviations Dr. Scott Langenecker -Dr. Langenecker is a clinical neuropsychologist, and neuroimaging specialist who focuses on the translational neuroscience of mood disorders. His goal is to contribute to work de ning the neural circuits leading to the development and perseveration of depression and related mood disorders, as well as prediction of treatment response.
Current work by Dr. Langenecker, funded by two grants from the NIMH, focuses on mood disorders in the remitted state. The rst study is designed to predict recurrence of depression using performance and neural circuit markers. The second study is assisting in de ning key neural and performance markers of mood disorders and how they may have shared and unique features across many mood disorders.Dr. Langenecker directs an adolescent mood disorders neuropsychology clinic, specializing in adaptive transitions from adolescence to adulthood, treatment optimization, and identi cation of strengths and weaknesses to assist in care for mood disorders and comorbid conditions.

Dr. Hakan Demirtas -Hakan Demirtas is an Associate Professor in Division of Epidemiology and
Biostatistics University of Illinois at Chicago, USA. His research interests includes Stochastic simulation, statistical computing, statistical distribution theory, random number generation, missing data, multiple imputation, incomplete multivariate data, longitudinal data, Bayesian computation.