Humanitarian practitioners and policy-makers increasingly recognise the importance of mental health and psychosocial support (MHPSS) for conflict-affected children and adolescents. The consequences of inaction are grave: profound stress can negatively affect decision-making, social behaviours, learning abilities and even earning capacities. In contexts of significant adversity, effective MHPSS programming responses can have positive impacts on young people’s lives in the short term, and help improve the chances of a flourishing life in the long term. Yet resources for psychosocial support as part of humanitarian responses remain scarce. Increasing investment for this priority hinges largely on providing credible, actionable evidence on what works in MHPSS programming. Researchers, practitioners and funders are searching for innovative, practical ways to rigorously evaluate existing programmes to produce both context-specific insights and transferable lessons. In this article, we reflect on ways to meet such demands for actionable evidence in humanitarian settings, drawing from a case study that evaluated the health impacts of Mercy Corps’ Advancing Adolescents programme, a youth-focused psychosocial intervention in Jordan. In the course of the 18-month project (2015–17), we came across many challenges, which sparked a process of deep reflection. Challenges to leveraging robust impact evaluations include:
- Short-term humanitarian funding cycles, which restrict timelines for scientific evaluation (e.g. randomised trials featuring wait-listed control groups to assess the added value of interventions; longer periods of follow up observation to test for sustained impacts).
- Slow academic timelines, as publishing research often involves extensive data analysis and lengthy peer review processes of critical evaluation.
- Deep engagement to create trust and build strong academic–humanitarian community partnerships (moving from processes of ‘cultural adaptation’ to ‘cultural engagement’).
- Identification of the ‘active ingredients’ or key elements of specific interventions, in order to understand what works, for whom, why and for how long, in specific contexts (moving beyond the basic question of ‘what works’ in mental health and psychosocial programming).
- Sustainability of partnerships over time, to foster new ways of working (funds are rarely geared to sustaining dialogue in order to achieve a more effective dissemination of findings).
The Advancing Adolescents progamme
Mercy Corps’ Advancing Adolescents programme, funded as part of the No Lost Generation initiative, was launched in 2013 in response to the Syria and Iraq crises. It is a psychosocial intervention of structured, group-based activities (technical, vocational and art skills) targeting 8–15-year-olds. This brief, scalable intervention (16 sessions across eight weeks) has been implemented in Jordan, Lebanon, Iraq and Syria. Its goals are to alleviate profound stress, build resilience, strengthen social cohesion and heal conflict. It is strategic in focusing on adolescence, a key time for protecting the next generation and building its future, and in serving both Syrian refugee and host communities. Key elements of the programme are common to other psychosocial interventions, including group-based skillsbuilding sessions (e.g. vocational skills, technical skills, fitness, arts and crafts) run by trained local community volunteers. The programme emphasises stress management, relationships and personal achievement (following the International Child Development Programme manual).+J. MacPhail et al., ‘Conflict, the Brain, and Community: A Neurobiology-Informed Approach to Resilience and Community Development’, in R. Phillips, S. Kenny and B. McGrath (eds), The Routledge Handbook of Community Development: Perspectives from Around the Globe (London: Taylor & Francis).
Our programme impact evaluation was independently funded by Elhra’s Research for Health in Humanitarian Crises (R2HC) programme, supported by the UK Department for International Development and the Wellcome Trust. Funding was contingent on a partnership between academics and humanitarian practitioners. We formed a re-search consortium+For details on the consortium, see: http://www.elrha.org/map-location/yale-psychosocial-call2. drawing on expertise from Mercy Corps, five universities and one in-country research partner to provide a robust impact evaluation. The research+C. Panter-Brick et al., ‘Insecurity, Distress, and Mental Health: Experimental and Randomized Controlled Trials of a Psychosocial Intervention for Youth Affected by the Syrian crisis’, Journal of Child Psychiatry and Psychology, October 2017 (http://onlinelibrary.wiley.com/doi/10.1111/jcpp.12832/full). had two main goals: to test programme impacts on mental health and psychosocial wellbeing, adopting a robust scientific approach; and to develop tools that were relevant to contexts experienced by refugee and war-affected youth. We introduced three major innovations: an experimental research design that enhanced both scientific rigour and cultural engagement; methods of assessment that tracked mental health and psychosocial wellbeing over time and corroborated self-reports with biological measures of stress; and activities to promote awareness and uptake of results and recommendations.
Innovation 1: research design and ownership
We sought to overcome common challenges of impact evaluations, including the lack of a control group, which can lead to unreliable evidence of impacts, and reliance on self-reports of programme participants, which can lead to response bias. We employed a waitlisted randomised control trial design, conducting the study at three time-points, with a gender-balanced sample of refugee and non refugee youth (n=817) who were registered and eligible for the Advancing Adolescents programme. We recruited a separate research team to ensure operational independence between the staff responsible for research evaluation and programme implementation.
Our main challenge was to randomise the trial (namely, randomly allocate youth to ‘intervention’ or ‘control’ groups). The short and unpredictable funding timeline for humanitarian programming made it more difficult to establish waitlisted controls, and we could not ethically randomise participants until Mercy Corps was certain it could continue programming in Jordan, with renewal of donor funding. The trial also had to be fair from the standpoint of local participants. We talked with the community to establish an open and fair process, and settled on a lollipop draw. With twice as many young people recruited to start the next round of programme sessions, all of whom completed baseline assessments, we asked youth to draw a lollipop from an opaque bag (containing lollipops of two different colors, randomised with a coin flip to either treatment or control). This process of randomisation was transparent to families, who agreed to a chance allocation for accessing the programme at that point and two months later. In this way, we were able to put in place one of the very few randomised control trials undertaken in a humanitarian setting. Our experience confirmed that bottom-up approaches to community engagement are key. It also showed that in-country scholars could be powerful champions of the scientific process, helping to develop a sense of ownership through a deeper engagement with local communities.
Innovation 2: a multi-level toolkit
Our main research questions were: what are the psychosocial, physiological and cognitive impacts of this eight-week intervention, and which tools best capture these multiple dimensions of wellbeing? Figure 1 illustrates these three main health outcomes: psychosocial measures through face-to-face surveys with youth, with paper and pen, in private locations; physiological outcomes through stress biomarkers, by tracking the biological signature of stress hair cortisol concentrations; and cognitive impacts through tablet-based tests of general attention and executive functioning. We reflect here on the value of surveys and the challenges they can present.
Surveys: ensuring context specificity and cultural relevance
Insecurity, distress, mental health difficulties (anxiety/depression), traumatic stress and resilience are distinct dimensions of human wellbeing, but are often conflated in everyday speech and scientific measurement. We took steps to capture all these different dimensions of mental health and psychosocial support.+Ibid. For instance, we used the Human Insecurity scale, developed for conflict-affected regions of the Middle East. This scale asks questions such as: to what extent do you fear for your family in daily life; worry about losing your source of income or your family’s source of income; fear displacement or being uprooted? Responses capture levels of fear or insecurity, which in conflict zones is qualitatively different from common understandings of ‘stress’. In assessing lifetime trauma, we used a checklist developed for conflict-affected populations, which includes questions such as: have you directly witnessed bombardment as a result of war; seen someone else severely beaten, shot or killed; had your life put in danger; been expelled from your home?
We also developed a culturally relevant, Arabic-language measure of resilience,+C. Panter-Brick et al., ‘Resilience in Context: A Brief and Culturally-grounded Measure in Syrian Refugees and Jordanian Host-community Adolescents’, Child Development, June 2017 (http://onlinelibrary.wiley.com/doi/10.1111/cdev.12868/epdf). manifested in the ways people overcome adversity and sustain wellbeing. Appraising resilience was an explicit request from the youth themselves, who disliked a survey exclusively focused on trauma and risk, and asked why we did not look at their strengths and the positive aspects of their lives. The challenge was to develop a brief but valid metric that captured the individual, relational and cultural understandings of resilience – a word locally translated to muruuna (literally ‘flexibility’). Through qualitative interviews, we listened attentively to the vocabulary that expresses lived experience, and in surveys piloted several measures. We found the Child Youth Resilience Measure (CYRM, 12 items) to be a brief and reliable measure for use in population surveys. This measures the extent to which youth feel strong as a person, in their relationships with others and in their community. It thus covers three culturally-relevant dimensions of resilience: individual, inter-personal, and social dimensions.
Stress biomarkers: a biological stress diary
Stress biomarkers help provide a biological signature of adverse experiences. Cortisol is a hormone secreted in the body associated with the regulation of energy and psychosocial stress. It can be measured in saliva and hair, to capture an individual’s stress profiles over time to evaluate health and development. Measuring cortisol in human hair reveals a ‘stress diary’ that tracks the cumulative effects of stress in the body. By measuring hair cortisol concentration levels in samples of hair, we were able to test, beyond self-reports, the effectiveness of the intervention in reducing stress over time.+We measured cortisol in hair, rather than saliva; the former measures chronic stress levels. This required cutting around 100 strands of hair, covering a scalp area smaller than a pencil head, three times for each participant (pre-/post-intervention and 11-month follow-up). R. Dajani et al., ‘Hair Cortisol Concentrations in War-affected Adolescents: A Prospective Intervention Trial’, Psychoneuroendocrinology, 89, 2018 (http://www.psyneuen-journal.com/article/S0306-4530(17)31207-6/fulltext).
Incorporating this major methodological advance into our study was no mean feat. Some local staff and parents had serious questions as to why this was necessary. However, the youth themselves were open to this type of scientific measurement. Again, community engagement was key to getting staff and families on board. We explained the science behind measuring stress and its impacts in the body, which capitalised on the youths’ eagerness to learn like scientists. We also hired male and female local hairdressers as part of our field team to give young people a professional haircut when taking their hair samples, which was greatly appreciated.
We sent three hair samples per participant (each around 100 strands, pre-/post-intervention and 11-month follow-up) for laboratory analysis. We assessed changes in cortisol levels in response to programme participation, lifetime trauma events, perceived insecurity and mental health difficulties. In tracking cortisol over time, we identified individuals whose cortisol profiles were chronically high, and also chronically low.[footnote]Ibid. Both hyper-production and hypo-production of cortisol can raise health concerns, indicating a heightened versus blunted sensitivity to environmental challenges. We thus saw that a biological ‘sensitivity’ to life experiences could lead to blunted cortisol responsiveness, as well as put the body on high alert, such that the positive impacts of an intervention might lie in cortisol regulation, raising chronically low levels as well as reducing chronically high levels. One cannot assume a one-to-one relationship between exposure to extreme stressors in the environment and the biological stress response.
Tablet-based tasks: measuring cognitive performance
Lastly, we tested whether our measures of risk and resilience impacted cognitive performance, as measured by long-term memory, working memory and inhibitory control. We used the Rapid Assessment of Cognitive and Emotional Regulation (RACER) tasks, a set of cognitive tasks designed for use in low and middle-income countries to measure general attention and executive functioning, as games run on tablet computers. For Syrian refugees, we found that feelings of high insecurity were associated with shortfalls in general cognitive resources: the more fear and worry the adolescents felt, the poorer they performed in trials.+A. Chen et al. ‘Minds under Siege: Cognitive functioning in Syrian refugee adolescents impacted by aremed conflict and displacement’, Child Development (under review). This highlights the importance of current insecurity – rather than past trauma – in shaping self-regulatory skills and cognitive performance, and by extension learning abilities. We also found that the Advancing Adolescents intervention did not impact cognitive function, suggesting that brief interventions that can improve mental health and reduce physiological stress do not necessarily improve cognitive function.+K. Hadfield et al. ‘How Malleable Is Cognitive Function in War-affected Youth? A Randomized Controlled Trial with Refugee and Host Populations’, Development Science (under review).
Innovation 3: translating research into practice
Our ultimate goal was to promote the uptake of rigorous academic research to improve health-related programming in humanitarian crises. To bridge the traditional gaps between academics, humanitarian practitioners and donor decisionmakers, we organised and co-presented at multiple conferences and thoroughly discussed the findings with funders, scholars, humanitarian stakeholders and local communities. Regular, in-person discussions have been essential in translating the research findings and recommendations into practice. This process informed programmatic decisions on targeting and scaling up the Advancing Adolescence programme: for example, Mercy Corps integrated stress-attunement elements into livelihood interventions in the region, and included measures of resilience and human insecurity in evaluation surveys. The research has also prompted learning from negative findings, where we observed no direct or sustained impacts for certain outcomes. For example, Mercy Corps was led to consider how to engage whole families in resilience-building efforts.+Panter-Brick et al., ‘Resilience in Context’. We also shared insights regarding the fundamental roles that donors can play in supporting strategic collaborations to enable rigorous and longer-term impact evaluations. At national and local levels, we shared results with Syrian and Jordanian research staff and community youth, which created palpable excitement about the goals and achievements of ‘science’.
This research partnership holds valuable lessons with important implications for psychosocial programming. First, young people – even those who live in very difficult conditions – are willing and curious to participate in a scientific study, but ask that the scientific community focus on their resilience and human dignity, not just their trauma and vulnerability. In crisis settings, researchers and practitioners need to better understand how people experience and communicate psychosocial distress or support for wellbeing as a prerequisite to effective responses. They need to take ‘culture’ seriously, and when choosing assessment tools or implementing key programme components, move from surface cultural adaptation to deep cultural engagement.
Second, the signatures of stress are malleable and changeable. This tells us that evidence-based interventions can influence stress physiology and benefit young people living in adversity. The Mercy Corps programme evaluated was effective in regulating stress physiology and reducing the time that youth spent experiencing high levels of insecurity. Modest but demonstrable impacts (Box 1) are important for meeting the humanitarian imperative to improve lives and alleviate suffering in the short term. How this can translate to improving young people’s learning abilities, social relationships and economic potential in the longer term remains an important question.
Third, establishing lasting partnerships between academics, humanitarians, funders and local communities can make major contributions to generating credible evidence and improving its uptake by frontline agencies. Such partnerships are crucial to help us strengthen the evidence for strategic investments that put children and families at the heart of humanitarian efforts to boost health and dignity. However, we need to structure partnerships with sustained funding: it requires time and effort to establish productive dialogue between academics and humanitarians, and then communicate insights to wider audiences, including donors and the media.
Catherine Panter-Brick is Professor of Anthropology, Health and Global Affairs at Yale University, where she directs the Program on Conflict, Resilience and Health. Jon A. Kurtz is Senior Director for Research and Learning at Mercy Corps. Rana Dajani is Associate Professor of Molecular Cell Biology at the Hashemite University in Jordan.