Keywords

Introduction

Worldwide, nearly 250 million children – one in every ten of the world’s children- live in countries and areas affected by armed conflict (UNICEF, 2016). The world is witnessing unprecedented numbers of displaced people and the highest numbers of war-related fatalities since 1989 (UNHCR, 2015). It is estimated that more than 2 million children have lost their lives in the past decade due to conflict (HIIK, 2016) and beyond that, 9.9 million children are refugees and a further 19 million children are internally displaced (ECHO, 2016). The 2016 annual report of the Secretary-General on Children and Armed Conflict reported ongoing serious concerns regarding both the protection of children affected by armed conflict, and the increased intensity of grave violations against children in many conflict-affected settings (including killing and maiming, child recruitment and use, sexual violence, abductions, attacks on schools and hospitals, and denial of humanitarian access) (UN, 2016).

The nature of modern conflict has seen an increasing shift towards wars of destabilization; that is, conflicts are increasingly fought within states, and are commonly not confined to distinct battlefields but rather specifically target civilian populations, along with essential infrastructure. Social networks, community structures and processes, service systems, and religious institutions can be disrupted or purposefully destroyed, and deep ethnic or political divides in society can be created or exacerbated (Barber, 2013; Betancourt & Khan, 2008). Thus, beyond the direct threat to life and individual impact of exposure to conflict-related violence and destruction, armed-conflicts affect the entire social ecology of children.

In this chapter, we will discuss how War Child Holland, an international non-governmental organisation (NGO) that aims to improve the wellbeing, and strengthen the resilience, of children affected by armed conflict, is working towards a comprehensive evidence-based system of care that can be implemented at scale. To put that work in context, we first discuss the impact that war and community violence might have on children and provide an overview of current research into the effectiveness of interventions for conflict-affected children. Based on this discussion, we formulate a number of recommendations that should be taken into account when developing a system of care.

Impact of War on Children

At the individual level the impact of armed conflict on the physical, mental and psychosocial wellbeing of children and youth has been well established (Barber, 1999; Barenbaum, Ruchkin, & Schwab-Stone, 2004; Panter-Brick, Goodman, Tol, & Eggerman, 2011). High rates of posttraumatic stress disorder (PTSD) symptoms, behavioral and emotional symptoms and disorders, sleep problems, disturbed play, and psychosomatic symptoms are found among conflict-affected children and youth (Attanayake et al., 2009; Betancourt et al., 2014; Fazel, Reed, Panter-Brick, & Stein, 2012; Miller & Jordans, 2016; Slone & Mann, 2016; Stichick, 2001).

It is widely recognised that a positive and nurturing family environment is essential for child development and wellbeing. Indeed, secure and consistent caregiving relationships can play a critical role in helping children to cope effectively with exposure to conflict and the many other ongoing stressors in these environments (Betancourt, Meyers-Ohki, Charrow, & Tol, 2013; Miller & Jordans, 2016; Tol, Song, & Jordans, 2013). Yet, in conflict-affected settings, caregivers exposed to conflict-related and other common daily stressors can have high rates of psychopathology and may have difficulty in providing responsive and effective parenting (Slone & Mann, 2016). There is evidence that conflict-affected parents often have difficulties interacting with children, become less sensitive and responsive to children’s needs, and may be less effective at maintaining rules and setting boundaries (Barenbaum et al., 2004; Khamis, 2014; Miller & Jordans, 2016). Furthermore, there is growing evidence across multiple settings that family violence increases significantly in settings of armed conflict (Catani, Schauer, & Neuner, 2008; Panter-Brick et al., 2011). Taken together, these findings suggest that the family environment, and parental wellbeing and parenting behaviour in particular, represent key mediators on the relationship between armed conflict and children’s mental health and psychosocial wellbeing (Miller & Jordans, 2016).

Community-level conditions, such as the prevalence of child labour and poverty, negatively impact upon children and increase a child’s risk of experiencing harm (McLeod & Shanahan, 1993; Srivastava, 2011). In times of crisis, child protection risks such as violence and exploitation increase and become exacerbated (Bartels & Hamill, 2014). Ongoing conflict may lead families and communities to resort to harmful coping mechanisms, such as early marriage—a strategy sometimes used to protect adolescent girls from sexual violence (Bartels & Hamill, 2014). In addition, ongoing conflict results in damage to the community fabric, including the generation of distrust among members of different religious or ethnic groups, and damage to structures and available services such as education and health facilities (CPWG, 2015).

Interventions to Address the Impact of War on Children

Historically many explanatory and intervention models have focused on the direct impact of exposure to war-related violence on children’s mental health. However, increasing attention is now being paid to both the direct and indirect impact of armed conflict on children as well as understanding the multiple pathways by which armed conflict affects children’s ongoing development and psychosocial wellbeing more broadly (Miller & Jordans, 2016; Tol, Jordans, Kohrt, Betancourt, & Komproe, 2013; Tol, Song, et al., 2013). This more comprehensive model considers the “daily stressors” caused or exacerbated by exposure to armed conflict and draws attention to multiple risk factors at all levels of the social ecology i.e., the family, peers, school and wider-community. Protective factors have also been identified that may positively mediate the impact of exposure to violence on mental health and wellbeing and function as a buffer (Betancourt et al., 2013; Tol, Song, et al., 2013). Therefore, in attempting to understand and respond to the effects of war on children, it is important to consider each of these multiple pathways of impact.

Several recent literature reviews (including systematic reviews and meta-analyses) have synthesised mental health interventions and their evidence, and have specifically focused on interventions aimed at children and adolescents affected by armed conflict (Barry, Clarke, Jenkins, & Patel, 2013; Betancourt et al., 2013; Jordans, Pigott, & Tol, 2016; Jordans, Tol, Komproe, & de Jong, 2009; O’Sullivan, Bosqui, & Shannon, 2016; Tol, Barbui, et al., 2011). Altogether, these reviews have included 150 unique publications. In recent years, there has been a promising increase in the number and quality of research evaluations of the effectiveness of interventions targeting the mental health and psychosocial wellbeing of children in adversity in low and middle income countries (LMIC).

There is empirical support for the use of individual as well as group-based psychosocial interventions for children to improve mental health and promote psychosocial wellbeing, particularly among children experiencing clinical levels of distress (Jordans et al., 2016; Jordans et al., 2009; O’Sullivan et al., 2016; Tol, Barbui, et al., 2011). Evidence tends to be limited to sub-groups of children (e.g., only boys or girls, only older or younger children) and centred around a few treatment types (i.e., Trauma Focused-Cognitive Behavioural Therapy, Narrative Exposure Therapy), with evidence for CBT-based interventions generally showing larger effects than interventions based on other theoretical frameworks (O’Sullivan et al., 2016). The generalizability of these findings is limited, however, by an over-reliance on highly resourced efficacy studies, as the relevance of such studies for everyday practice remains largely unknown. The higher quality evidence available to date has also tended to focus mostly on trauma-focused interventions that are infrequently implemented outside research settings (Tol et al., 2012). Interventions that are more commonly implemented in the field, such as structured social activities and developing or strengthening community-based social supports, are rarely subjected to rigorous evaluation (Tol, Barbui, et al., 2011). The same conclusion can be drawn for interventions in child protection (Wessells, 2009). In order to bridge this divide between research and practice, there is a need for further robust evaluations using pragmatic designs that more closely approximate the real world of everyday practice in which NGOs and local institutions operate. There is also a compelling need to transcend the dominant focus on providing direct services to children, by including interventions that address other socio-ecological levels, such as families, schools, and other community organizations and structures (Jordans & Tol, 2015).

To date, only a handful of controlled evaluations have been published with parents and/or families in conflict-affected settings, with varying outcomes. A two-session parent group psychoeducation intervention delivered in rural Burundi led to reductions in aggressive behaviours in boys compared to a waitlist group; however, no treatment effect was seen for depression symptoms or family social support (Jordans, Tol, Ndayisaba, & Komproe, 2013). A study comparing a multi-component psychosocial intervention plus medical care to medical care only for parents in Bosnia found mixed results. The psychosocial intervention led to greater improvements in some measures of child psychosocial functioning and mental health as well as mother’s wellbeing, but no significant improvements on other measures of these outcomes (Dybdahl, 2001). Similarly, a family-focused group psychosocial intervention delivered in the Democratic Republic of Congo (DRC) for war-exposed youth at risk of attack or abduction, found significant intervention effects compared to a waitlist control group on PTSD symptoms, but no effect on depression and anxiety symptoms, conduct problems, or prosocial behaviour (O’Callaghan, McMullen, Shannon, Rafferty, & Black, 2013). A structured activities program combined with parent psychoeducation led to significant treatment effects (compared to a no-treatment comparison group) on behavioural and emotional difficulties for certain subgroups of children (Loughry et al., 2006). Finally, Puffer, Annan, Sim, Salhi, and Betancourt (2017), in their waitlist RCT of a 12 week family support and parenting intervention with Burmese refugees in Thailand, found improvements in family cohesion and parent-child relationships, and a parent-reported decrease in harsh parenting behavior (children did not report a decrease in harsh parenting). However, no effect was found for positive parenting and measures of child wellbeing were not included in the study.

Increasing attention is also being paid to interventions that strengthen social networks and reinforce traditional support mechanisms to help promote children’s wellbeing. However, there is a scarcity of well-designed studies focusing on community support (Betancourt et al., 2013; Jordans et al., 2016; Wessells, 2009). A number of studies suggest that community support may play a protective role for children affected by armed conflict (Betancourt et al., 2013). Community level support is likely to reach large groups of children, tends to be low cost and therefore more sustainable, and community-level actors are often well-positioned to support families and children when problems come to light. When working towards the improved wellbeing of children, community-level action to prevent harm is required (Wessells, 2009). Moreover, there is agreement within the humanitarian sector that education, crucial as in intervention in itself, is significant for psychosocial wellbeing as it restores a sense of normalcy, dignity and hope (INEE, 2012).

Recommendations and Future Directions

The following section discusses a number of recommendations that we believe should be taken into account when developing services for children affected by armed conflict.

More Attention for Multi-level Interventions

Multi-level interventions focusing on the mental health and psychosocial wellbeing of children in adversity, including complex emergencies, are commonly advocated. The Inter-Agency Standing Committee (IASC) Guidelines for Mental Health and Psychosocial Support (MHPSS) in Emergency Settings specifically promote this approach via a four-tiered model; Level 1: basic security and service to enhance wellbeing of the general population; Level 2: responses to a smaller group that is able to maintain their wellbeing through the support of family and community networks; Level 3: focused, non-specialised supports for the still smaller number of people requiring more focused individual, family or group interventions to recover from their distress; and Level 4: specialised services delivered by qualified professionals to severely distressed and/or impaired individuals (IASC, 2007). Multi-level interventions have demonstrated feasibility and promising preliminary findings. However, they are rarely reported and evaluated (Betancourt et al., 2013; Jordans et al., 2016). A systematic review in 2016 found that 52% of the publications reviewed recommended that interventions should apply multi-level approaches (Jordans et al., 2016). Text Box 1 provides a promising example of a multi-level intervention.

Box 1: Multi-levelled Child Mental Health Program

A multi-levelled care system, where different levels of support interventions are available to address the psychosocial needs, for children in five conflict affected countries (Indonesia, Sri Lanka, Nepal, South Sudan and Burundi) was positively evaluated, with high levels of satisfaction and considerable levels of perceived post-treatment outcomes among children and parents (Jordans, Tol, et al., 2010). The program included organizing recreational activities, a 15-session group intervention that incorporated various components, such as structured expressive behavioral activities for children with elevated distress, non-specialized individual (or group) counselling for children with more severe problems in combination with family support, as well as specific referrals to specialized support (Jordans, Tol, et al., 2010), corresponding to levels 2 and 3 as described above.

Boosting Multi-sectoral Interventions

In the IASC MHPSS guidelines the importance of working inter-sectorally is articulated, reflecting growing consensus that the various sectors impact on psychosocial wellbeing (IASC, 2007). Protection and education guidelines and standards further emphasise the effect of both sectors on psychosocial wellbeing (CPWG, 2012; INEE, 2012). An inter-agency review of child protection interventions cited multiple studies related to the mental health and psychosocial wellbeing of children but also noted a lack of rigorous study designs, which limits any conclusions about the effectiveness of interventions. Moreover, these evaluations focused heavily on process and output indicators, while paying insufficient attention to psychosocial and mental health outcomes for children and their families. A recent review on health interventions argues for the need for more evidence of inter-sectoral approaches (nutrition, education, protection) in relation to mental health and psychosocial support (Blanchet et al., 2015). Text Box 2 provides an example of an education programme with psychosocial benefits.

Box 2: Healing Classrooms

The International Rescue Committee conducted a program that emphasizes education’s role towards enhanced mental health and psychosocial wellbeing. Classrooms were converted into places of not only academic learning but also of healing. The program aimed to improve teachers’ wellbeing, with a subsequent effect on students’ wellbeing and academic performance. The research indicated significant psychological benefit for the children as they could picture themselves being able to provide support to their families in the future (Winthrop & Kirk, 2005). The importance of each teacher’s identity and competences was identified, and further research into teacher professional development was recommended.

Interventions from a Socio-ecological Perspective

As we discussed earlier, it is imperative that interventions go beyond focusing solely on direct work with children, and focus on the critical factors in the environment that impact on children’s wellbeing: family, school, peers, community, and a larger macro or societal context. The various ecological levels include protective factors that influence the psychosocial wellbeing and resilience of an individual child. Research suggests that interventions targeting risk and protective factors beyond an individual child are essential for helping children deal effectively with adversities (Betancourt & Khan, 2008). Indeed, interventions targeting the family and linking child protection and mental health have shown promise in improving child psychosocial outcomes (Sim, Annan, Puffer, Salhi, & Betancourt, 2014). Text Box 3 describes an intervention that simultaneously focuses directly on children as well as their families.

Box 3: Building Happy Families

Sim et al. (2014) conducted an RCT on parenting and family skills in a low-resourced displacement setting to evaluate feasibility and effectiveness in reducing behavioral problems, as well as promoting psychosocial wellbeing and resilience. Caregivers and children attended sessions separately, which were succeeded by joint family activities afterwards. Both caregivers and children expressed high interest and satisfaction. The intervention led to improvement on many family factors such as positive parenting and improved caregiver-child interaction. Though no impact was reported on children’s emotional problems, improvement in resilience and reduction in negative behavioral issues were demonstrated. This intervention unintendedly also showed potential in improving caregiver mental health.

Need for Increased Scientific Rigour

The evidence-base for interventions for children in areas of armed conflict is mixed, and significant gaps in knowledge persist as outlined above (Betancourt et al., 2013; Brown, Graaff, Annan, & Betancourt, 2017; Jordans et al., 2016; O’Sullivan et al., 2016). Furthermore, there are a number of lessons that can be drawn from recent literature reviews regarding the future development and improvement of services. Although the intervention research base has increased significantly in recent years, many evaluations continue to lack rigor (e.g., frequent use of un-controlled studies, use of waitlist control designs rather than active comparison groups) (Jordans et al., 2016). O’Sullivan et al. (2016) stress that future research needs to consider more appropriate (culturally adapted) outcomes. There is a call for diversification as intervention studies seem to be skewed both geographically (some conflict-affected settings are over-represented) and by type of intervention (a disproportionate number of trauma-focused interventions). Programs are commonly implemented as one-off initiatives and with little conclusive longitudinal data available or ongoing implementation of services within the setting (O’Sullivan et al., 2016). Unfortunately, research results are often not used nor translated for the improvement of programs in humanitarian settings (Tol et al., 2012). Further, there is a lack of empirical knowledge in this field about active therapeutic ingredients and clinical processes driving change in effective interventions (i.e., through dismantling studies and mediational analyses) (Brown et al., 2017). In order to adequately inform the development and implementation of optimally effective, targeted, efficient, and sustainable interventions, it is vital that future work consider these issues.

Importance of Attending to Stigmatisation and Discrimination

Worldwide, people are excluded from participation in their society, including making optimum use of services, because of aspects of their identity such as gender, race, health status, sexual orientation, age and ability (Thornicroft, 2008). Although research specifically concerning mental health-related (experienced) stigma has increased over the last decades, intervention studies have only recently been conducted, mostly in high-income countries (HICs) and mostly focused on adult populations. Recent research indicates that globally over 70% of young people and adults with mental illnesses receive no treatment, with a higher percentage in LMICs where potential barriers to accessing treatment include prejudice against people with a mental illness as well as expectations of discrimination (Thornicroft, 2008). Social contact-based interventions with people with mental illness seem promising for short-term attitudinal changes but caution is warranted in generalizing these findings to other stigmatized groups and populations (Thornicroft et al., 2015). Beyond stigma and discrimination, issues of inclusion can extend to other groups – for example, out-of-school children who are unable to participate in school-based interventions (Fazel & Tol, 2014).

Cultural Adaptation of Interventions

Cultural adaptation of existing evidence-based interventions when delivering them in different linguistic and cultural contexts is important not only to ‘do no harm’ and build upon local strengths and contexts, but also to increase the potential of positive outcomes and promote sustainability (Bernal & Sáez-Santiago, 2006; Castro, Barrera Jr, & Holleran Steiker, 2010). For an example of a cultural adapted intervention, see Box 4. Studies of culturally adapted psychotherapeutic interventions have demonstrated increased effect sizes compared to studies of non-adapted interventions (Benish, Quintana, & Wampold, 2011; Griner & Smith, 2006), with a recent review indicating that effectiveness increases the greater the number of adaptations made (Harper, Heim, Chowdhary, Maercker, & Albanese, 2016). Although many reports of program implementations in different settings report some kind of cultural adaptation, often no (detailed) account of adaptation is given.

Less than half of the studies in a recent systematic review mentioned the development of locally adapted versions of outcome measures (Jordans et al., 2016). Such adaptations are necessary in order to ensure the cross-cultural validity of the measures, as expression of symptoms can vary between different cultural contexts, which can lead to inaccurate findings during intervention evaluation (O’Sullivan et al., 2016). It is imperative that further research considers the value and process of cultural adaptations of both intervention methods and assessment tools (Jordans et al., 2016; O’Sullivan et al., 2016; Tol et al., 2012).

Box 4: School-Based Psychosocial Structured Activities

The impact of a school-based Psychosocial Structured Activities (PSSA) program in Northern Uganda has been positively evaluated (Ager et al., 2011). It serves as an example of a culturally adapted intervention, based on a Classroom Based intervention that has previously been used amongst others in Palestine. The program adopted an approach that went beyond children’s individual symptomatology and incorporated different ecological levels, by asking parents and teachers to report on the child’s wellbeing, in addition to the child’s self-report. Significant improvements in wellbeing were observed for the children in the intervention group in comparison to those in the control group, indicating that girls make greater progress than boys.

Replication and Sustainability

Replicability of interventions is determined by various factors. The quality of research is crucial to assessing the validity of its outcomes. In their review of mental health promotion interventions in LMIC, Barry et al. (2013) noted that cost-effectiveness is imperative in order to establish whether the intervention is feasible in non-research settings, and whether the cost of the intervention is justified by the level of desired change achieved. However, very few studies integrate a cost-benefit element into their research design, creating challenges for replication as well as sustainability. Group-based interventions demonstrate higher cost-effectiveness, and are commonly implemented, with most indicating moderate to strong treatment effects on children’s socio-emotional wellbeing (Barry et al., 2013). Schools are an optimal delivery platform for these kind of interventions because of the structure they provide (Barry et al., 2013; Betancourt et al., 2013). Internet-based mental health interventions, which have seldom been studied in LMICs but have shown value in HICs as a relatively low-cost intervention modality, may prove to have considerable potential (Arjadi, Nauta, Chowdhary, & Bockting, 2015). Although none of the studies mentioned by Arjadi et al. (2015) refer to children, a scoping review conducted in 2014 highlighted the potential of using technology, for example, videoconferencing, mobile phone applications and internet-based applications, for mental health interventions for children in HICs (Boydell et al., 2014).

Ethical Issues

Ethical issues in intervention research with conflict-affected children and families have received comparatively limited attention in the literature. Yet there are numerous ethical points that merit consideration. Examples include; (1) the use of untested interventions, or interventions tested in other contexts, without careful monitoring for untoward effects; (2) research interventions not relevant to the actual local needs and main problems that need addressing (O’Mathúna & Siriwardhana, 2017; Tol et al., 2012); (3) a focus on getting approval for studies from Ethics Review Boards, instead of actually focusing on the goal of research ethics, which is, amongst others, the protection of participants (O’Mathúna & Siriwardhana, 2017); (4) the use of passive versus active consent procedures by caregivers, and the use or non-use of children’s own assent to participate in intervention studies; (5) weighing the benefits of longitudinal research that allow for patterns of recovery, resilience and persistent distress to be examined, versus the imperative to provide treatments when these are available; and (6) providing treatment to children without adequately assessing ongoing stressors such as child abuse that may be contributing to their distress. If these and related considerations are taken into account when planning and conducting research, investigators can improve their ethical practice in humanitarian research, thereby ensuring the actual protection of participants (O’Mathúna & Siriwardhana, 2017) – something that is crucially needed in humanitarian contexts given the especially vulnerable position of participants.

Developing a Multi-sectoral System of Care: War Child Holland’s Response

Commonly in humanitarian response programs, interventions are defined in specific sectors that focus on a particular thematic area (e.g., psychosocial, child protection, education), creating silos without a focus on a shared goal or outcome. For many children and communities this approach is not adequate nor responsive enough to meet the complex and varied mental health and psychosocial needs found within post-conflict contexts. Rather, a system of complementary support mechanisms is required, targeting mental health problems or stress directly, as well as indirectly, through addressing the major social determinants of mental health and psychosocial wellbeing (de Jong, 2002; Jordans, Komproe, et al., 2010; Saltzman, Layne, Steinberg, Arslanagic, & Pynoos, 2003; Stichick, 2001; Tol, Jordans, Reis, & de Jong, 2009; Wessells & Monteiro, 2006; Williamson & Robinson, 2006).

War Child Holland is one among few international organisations that primarily focus on the mental health and wellbeing of children and communities affected by conflict. Drawing on the aforementioned recommendations, it combines a focus on psychosocial support, child protection and education through the development of an integrated, multi-level system of care. This care system entails an integrated approach in which interventions are interconnected and mutually strengthening, with a range of intervention methods available to respond to the varying needs of children and their caregivers. The care system is multi-level, in that interventions range from low-intensity and least restrictive access interventions that aim to promote wellbeing and prevent problems from arising, to higher-intensity and more targeted interventions designed specifically for children experiencing significant and enduring distress. Finally, the care system is socio-ecological in its approach, with services targeted at different ecological levels (individual and peers, families, schools, communities, civil society and state authorities). This approach reflects the reality that children’s development is inextricably linked to the families, communities, economic situation, social values and cultural influences in which their lives are embedded and which provide for their basic needs and protection.

At the core of the War Child Holland care system (see Fig. 1) is a set of complementary interventions, outlined in Table 1. Together they target community systems (both formal and non-formal), the school as a place to enable children to reach their full potential, all children in communities affected by armed violence to promote their wellbeing, children for whom more focused support is needed (either with regard to significant psychosocial distress or severe protection issues), and families who may have been adversely affected by conflict. We believe that this care system will therefore address both mental health consequences and related social determinants, by responding to the needs of individual children whilst also strengthening child protection and education services, building the mechanisms and confidence necessary within communities to facilitate the care and protection of children under their care. Two interventions directly target children’s common mental health problems, either through an intervention that combines specific therapeutic components or through a family systems treatment. Other interventions prevent these problems from developing through either life skills interventions, which aim to strengthen children’s resources to cope with the adversity they experience, or interventions focused on improving parental wellbeing and strengthening parenting under conditions of adversity. Reinstating structural education supports normalization, stability and continuity and also promotes continuation of the social role of a student, all of which are particularly important in conflict or post-conflict settings. Furthermore, a protective education environment is important to reduce the risk of children becoming involved in the worst forms of child labour, including being associated with armed groups and armed forces. Similarly, child protection services aim to deal with root causes of children’s psychosocial and mental health problems, for example, by establishing community structures to prevent, or respond to, child abuse. While each intervention has specific goals and expected outcomes, overall the care system aims to contribute to (1) building increased resilience at different socio-ecological levels and; (2) improving psychosocial wellbeing in children. In short, this is done by targeting groups of children (with universal and targeted interventions), parents, families and key structures and processes within communities as a whole.

Fig. 1
figure 1

Care system

Full size image
Table 1 Overview of interventions
Full size table

We recognise the need for the care system to be implemented within the real-world contexts of humanitarian work, which are often complex, variable and unpredictable. A number of enabling components will support the development and implementation of the interventions and promote access to care.

First, we are developing a structured tool to proactively identify children and families who may benefit from more targeted interventions, and to increase awareness, access and the uptake of these services. Using this tool will enable us to more effectively allocate interventions to higher risk children and families, based on identified needs. Particularly, because the (multi-level) care system includes interventions aimed at small groups of children with more severe problems, such a detection tool will become an integral component of this approach. Previous studies have found that a similar detection methodology with adults, using pictorial vignettes, led to acceptable accuracy in detection (Jordans, Kohrt, Luitel, Komproe, & Lund, 2015) and subsequent help seeking (Jordans, Kohrt, Luitel, Lund, & Komproe, 2017).

Second, we are aiming to ensure equitable inclusion of children. It is crucial to view this integrated programming framework from an inclusive perspective as, due to a variety of barriers and factors, some children are excluded from participating in their societies. These barriers need to be recognised and overcome to ensure that the care system is inclusive and responsive to all. This may include interventions that directly address stigmatization and discrimination to ensure that services can be accessed by all children and families who might benefit from them.

Lastly, we are developing a thorough and sensitive process of ‘community entry’, which will ensure that the services being offered are in line with the needs and expectations of relevant community stakeholders, and build on existing resources and structures.

Research and Development

The War Child Holland care system is complemented by a comprehensive research agenda to support the development process and evaluate its core interventions and their inter-related functioning. The research program works towards combining evidence-based care services with quality implementation standards. This entails going beyond demonstrating evidence for interventions, and towards creating an improved understanding of how interventions are implemented adequately and with sufficient quality in real-world low-resource settings. This combination will allow us to make real progress towards replication and scale-up efforts and closing the evidence-practice gap in humanitarian settings (Proctor et al., 2009; Shidhaye, 2015; Tol, Patel, et al., 2011).

A systematic approach of development and evaluation designed especially for complex interventions will be followed (Craig et al., 2008). This iterative process focuses on: (a) formative research towards the development and modelling of interventions, which can be achieved through systematic reviews of existing intervention evidence and related constructs, preliminary qualitative work and development of a Theory of Change (Anderson, 2004); (b) pilot studies to test procedures and assess relevance of interventions within the target settings and populations, via small-scale qualitative and quantitative studies, typically resulting in adaptions to interventions and evaluation protocols; (c) evaluation to assess effectiveness of the interventions, ideally with the use of randomized controlled trials, and; (d) an implementation phase focusing on surveillance of implementation outside a controlled study setting, which centrally includes assessment of quality of care.

To achieve these ends, we will develop and pilot-test standards for quality of implementation for each of these interventions. Following Miller’s (1990) clinical skills hierarchy, we operationalize these standards as the extent to which a service provider has the knowledge and skill required to deliver a treatment to the standard needed for it to achieve its expected effects (competence) and the extent to which a psychological treatment was delivered well enough for it to achieve its expected effects (adherence) (Fairburn & Cooper, 2011). Similar work has already been conducted with adults (Kohrt et al., 2015), which will need to be replicated both for children and for interventions covering multiple sectors (child protection, education and psychosocial support). Furthermore, additional implementation research will need to address remaining gaps in understanding; for example, the optimal model and dose of training and supervision, the most adequate delivery agent, strategies for integration into existing care systems, recruitment and retention strategies for competent lay workers, and the extent of equity in proposed service delivery models.

In addition to intervention-level research, there is a vast need for system-level thinking in the design and evaluation of care. We have argued that a broader systems approach is more capable of addressing the variety of needs of children affected by armed conflict. A systems-of-care approach raises obvious questions with regard to feasibility and sustainability (Jordans & Tol, 2013, 2015). Especially in settings where existing support systems are weakened by conflict, the development and continuation of a system of care will be challenging. Consequently, we will review whether and how a care system is feasible and applicable in terms of service uptake (coverage) and cost – again developing standards to assess these in practice. At the same time, it will allow for an assessment of the added value of an integrated or multi-sectoral approach towards improving wellbeing over single intervention or ‘silo-ed’ sector-specific approaches.

In summary, we are advocating, and making progress towards, a care system that can be implemented in humanitarian settings, and is evidence-based, scalable and equitable. This requires a trajectory that moves from a research space through an implementation space to achieve actual impact (see Fig. 2). The research space involves intervention-level work that transfers current practice to meaningful evidence-based practice by establishing efficacy/effectiveness on the one hand and relevance on the other. Subsequently, the implementation space involves system-level work that transfers meaningful evidence-based interventions to large-scale impact through assessing maintenance of quality of care via a set of competence and adherence standards, as well as assessing feasibility of the multi-sectoral service delivery framework through cost and coverage standards. The thinking behind this model is that once evidence for interventions has been established, implementation of services at scale can be assessed through a limited and defined set of standards for service providers and planners, with the assumption that if these standards are met impact is achieved.

Fig. 2
figure 2

Overview research and development trajectory

Full size image

Conclusions

To address the myriad psychosocial and mental health issues of children in areas of armed conflict, there is a need for comprehensive care systems that bring together prevention and treatment approaches. This can only truly be achieved if the social determinants of children’s mental health are adequately addressed. We therefore propose a care system that integrates targeted psychosocial support and mental health care with child protection services that address and prevent issues of child abuse and neglect as well as interventions that aim to promote the school environment – jointly geared towards improving children’s wellbeing and promoting resilience. The care system combines a set of interventions at different levels of the child’s ecology and of differing intensity. It addresses common barriers to care, notably stigmatization and under-detection of children in need of care. An enormous ‘service gap’ exists whereby the support needs in low resource settings vastly outweigh the capacity of available services, and thus for any care package to have impact it must be capable of being provided at scale. Consequently, this requires the research agenda to gradually shift from demonstrating effectiveness of such interventions in LMIC to demonstrating how they can work optimally outside the boundaries of a study context. A set of standards to monitor quality and feasibility of service provision when implemented at scale will be vital to contribute towards impacting large groups of children affected by armed conflict that are currently not receiving the care they might need.