Dissemination and implementation of evidence based, mental health interventions in post conflict, low resource settings

Abstract

The burden of mental health problems in (post)conflict low- and middle-income countries (LMIC) is substantial. Despite growing evidence for the effectiveness of selected mental health programs in conflict-affected LMIC and growing policy support, actual uptake and implementation have been slow. A key direction for future research, and a new frontier within science and practice, is Dissemination and Implementation (DI) which directly addresses the movement of evidence-based, effective health care approaches from experimental settings into routine use.

This paper outlines some key implementation challenges, and strategies to address these, while implementing evidence-based treatments in conflict-affected LMIC based on the authors’ collective experiences. Dissemination and implementation evaluation and research in conflict settings is an essential new research direction. Future DI work in LMIC should include: 1) defining concepts and developing measurement tools, 2) the measurement of DI outcomes for all programming, and 3) the systematic evaluation of specific implementation strategies.

Background

The burden of mental health problems in low- and middle-income countries (LMICs) is substantial. However, the gap between those who experience them and those who receive any type of treatment is large, particularly in LMICs, where up to 90% of people needing care do not receive health services (Demyttenaere et al., 2004; Saxena et al., 2007; Thornicroft, 2007).

Stemming from the recognition of the treatment gap, a growing body of evidence for treatments in humanitarian settings, and the general prioritization of mental health within the field of global health (WHO, 2008 ; 2009), there has been increased support for the use of evidence-based treatments (i.e., those interventions that have been shown to be effective in randomized controlled trials) in LMICs. For example, in the Mental Health Gap Action Programme (mhGAP; WHO, 2010a; Barbui et al., 2010; Dua et al., 2011), a set of treatment recommendations based on syntheses of research literature, the World Health Organization endorses the use of some evidence-based interventions as frontline treatment for mental health problems.

Despite growing evidence for the effectiveness of some mental health programs in conflict-affected LMICs and growing policy support (e.g., mhGAP), actual uptake and implementation by governments, non-governmental organizations (NGOs), and community-based organizations (CBOs) have been slow (Thornicroft et al., 2010). Notably, high-income countries have also documented significant challenges with dissemination and implementation of EBT (Shafran et al., 2009). A key direction for future research and the next frontier within science and practice is how effective treatments can actually be implemented in real-world health and community settings.

Dissemination and Implementation Science

Dissemination and Implementation (DI) science is an emerging field that directly addresses the movement of evidence-based, effective health care approaches from experimental settings into routine use (Peters et al., 2013; Rubenstein & Pugh, 2006; Thornicroft, 2012; Thornicroft, Lemp, & Tansella, 2011). Dissemination may be defined as “an active approach of spreading evidence-based interventions to the target audience via determined channels using planned strategies” while implementation is described as “the process of integrating evidence-based interventions within a setting” (p. 26; Brownson, Colditz & Proctor, 2012). While the translation of evidence-based interventions into practice to improve overall public health outcomes is a common theme for NGOs, governments and funders, actual knowledge on and evaluation of the process of how to disseminate information and support the use of programs is less developed.

In a brief overview of implementation science, Proctor and colleagues (2009) clarify that two important pieces are needed for effective DI: 1) an evidence-based treatment or program, and 2) a method, plan, or “technology” for implementing the evidence-based intervention in routine practice. The idea of an “evidence-based intervention” within DI can be quite broad and include practices, processes, policies, guidelines, treatments, or programs. These may have different levels of evidence depending on the program and the problem it addresses. Within health there is literature documenting the various levels of evidence, with RCTs considered to be important and one of the “highest” levels of evidence (Centre for Evidence-Based medicine, http://www.cebm.net; Burns, Rohrich, & Chung, 2011; Tansella et al., 2006). Current recommendations focus less around a particular time when an intervention is “ready” or has enough evidence to move to DI examination, and more around building hybrid studies that examine implementation challenges, processes, and outcomes during effectiveness evaluations (Curran et al., 2012).

Many different conceptual models have been designed and researched to better understand the multiple factors that affect the implementation and dissemination of EBTs (e.g., Tabak et al., 2012; Proctor et al., 2009; Aarons, Hurlburt, & Horwitz, 2011; Damschroder et al., 2009). Most models conceptualize the importance of the multiple ‘layers’ involved in ultimate delivery of EBTs in real-world settings and the need to consider each of these layers carefully. For example, one model describes three general layers: 1) Policy – including legislatures and governments; 2) Organizations – including decision making processes, leadership, and organizational culture and climate; and 3) Individual – including providers and consumers individual and/or group behaviors (Shortell, 2004; Proctor et al., 2009). There is also general agreement that it is crucial to consider dissemination and implementation over time: from pre-implementation (and adoption in principle), early/mid implementation, and longer-term maintenance and sustainability implementation to maintenance (Aarons et al., 2011). Furthermore, the emerging DI field has started to hone in on selection and measurement of some of the key indicators of success of implementation processes (see Table 1). Figure 1 is a reprint of the Proctor et al. (2009) conceptual model that shows one example of how the evidence-based intervention, DI strategies and DI outcomes fit together.

Table 1.

Dissemination and Implementation Outcomes	Definition
Acceptability/Satisfaction	The perception that a given EBT is agreeable, palatable, or satisfactory
Adoption/Uptake	The intention, initial decision, or action to try or employ an EBT.
Appropriateness	The perceived fit, relevance, or compatibility of the EBT: 1) for a given setting, provider, or consumer; or 2) to address a particular issue or problem
Cost	The additional expense of implementing an EBT and the cost-effectiveness of it.
Feasibility	The extent to which an EBT can be successfully used within an organisation, in a particular setting, or with a certain population.
Fidelity/ Quality of programme delivery	The degree to which an EBT was implemented as it was designed in its original protocol.
Penetration/Access to Services	The integration of an EBT within and across a service setting (e.g., across a population)
Sustainability/Standard practice of care	The extent to which the EBT is maintained or institutionalised within a setting’s ongoing operations.

Source: Proctor & Brownson, 2012

Fig. 1.

Conceptual model of implementation research

From a global perspective, Tansella and Thornicroft (2009) state that “if the development of implementation science is in its infancy, then its application to mental health practice may be considered as embryonic”. This is particularly true within LMICs and even more so in (post-)conflict LMICs where there is limited descriptive literature on DI outcomes, barriers, and/or facilitators, and even fewer rigorous studies examining specific strategies of implementation. Programmatically, dissemination and implementation is becoming more common (e.g., Ventevogel et al., 2011), but these efforts are largely void of rigorous evaluation of their success.

This paper presents DI research in conflict-settings as an essential new research direction. The authors utilize advances on some mental health interventions showing evidence through randomized controlled trials in such settings as a starting point. Most of these studies may be considered hybrid studies in that although the effectiveness of the interventions on reduction of symptoms was the primary outcome, the researchers also monitored some implementation outcomes (e.g., fidelity). Based on the authors’ collective experiences with these studies, we seek to outline key implementation challenges and strategies in implementing evidence-based interventions in conflict-affected LMICs (see Table 2 for list of authors’ studies).

Table 2.

Authors, year	Setting	Design	Study conditions and evidence base	Evidence base for treatment	Participants	Providers	Outcomes assessed	Conflict affected description
Bass et al, 2013	DRC, South Kivu;	Randomized controlled trial	Group CPT compared with access to individual support	CPT has multiple RCT from the U.S.	Adults	Psychosocial assistants employed by NGO.	Trauma, Depression Functioning Anxiety	Sexual violence survivors, exposure to ongoing rebel violence in their villages.
	Community based with services provided in rural settings					Previous training in case mgmt and supportive counseling
Bolton et al., (submitted)	Thailand/Burm a border; community based services provided.	Randomized Controlled trial	CETA compared to wait list control.	Other common elements have been tested in the U.S. but the CETA version for LMIC was first tested in this trial and the Weiss et al.	Adults	Lay community members; no formal training in MH and limited education	Trauma, Depression Functioning Anxiety	Burmese refugees subject to arrests and deportation
Weiss et al., (submitted)	Southern Iraq; community based integrated into MoH clinics	Parallel Randomized controlled trial	CETA compared to WLC CPT compared to WLC	CPT has multiple RCT from the U.S.	Adults,	Community health workers in the MoH system; no previous MH training; educated	Trauma, Depression Functioning Anxiety	Ongoing bombings, road closures and political events throughout the study.
				See above note on CETA
Bolton et al., (submitted)	Northern Iraq; community based interated into MoH clinics	Randomized controlled trial	IPT, BA, and a psychosocial intervention designed by the NGO	IPT and BA have multiple RCT from the U.S.	Adults	Community health workers in the MoH system, no previous MH training, educated	Trauma, Depression Functioning Anxiety
				The psychosocial intervention has not been tested.
Jordans et al, 2013	Sri Lanka, Burundi, Indonesia, Sudan, Nepal	Multiple designs (including CRTs; n=1 series)	Classroom Based Intervention (CBI)	CBI has had multiple trials across multiple countries	Children	Community volunteers; Teachers	Multiple	Ongoing community violence throughout or in the period prior to the program
Bolton et al., 2007	Northern Uganda	Randomized controlled trial	IPT compared to Creative Play and WLC	IPT has multiple RCT in U.S.	Adolescents	Community members; no formal training in MH and limited education	Trauma, Internalizing problems, Externalizing	Internally displaced persons camp; Ongoing violence in/around the area.
				Creative play has not been tested
Bonilla-Escobar et al. 2014 (report to USAID)	Buenaventura and Quibdo, Colombia	Randomized controlled trial	CETA compared to WLC Community intervention compared to WLC	See above on CETA	Adults	Community members; no formal training in MH and limited education	Trauma, Depression Functioning Anxiety, Anger	Most affected areas by displacement in Colombia, armed strikes during the program, and unofficial armed conflict for territory.
				Community intervention has not previously been tested

Caption:

Interpersonal Psychotherapy (IPT)

Behavioral Activation (BA)

Common Elements Treatment Approach (CETA)

Wait List Control (WLC)

Classroom Based Intervention (CBI)

Implementation Challenges and Strategies

Delivering an evidence-based intervention in a conflict-affected LMIC is rife with challenges, some which are relevant across settings and others that are context specific. Many of these challenges and potential strategies we discuss below cut across the time frames of implementation (pre-, during, and maintenance), as well as the various levels of implementation (policy, organization, and individual).

Lack of Properly Trained Personnel

A difficult challenge is the shortage of individuals with professional mental health training (Eaton et al., 2011; Kakuma et al., 2011; Kieling et al., 2011; Saraceno et al., 2007). The comparisons between high-income countries (HICs) and LMICs are stark: for example, one general psychiatrist is available in lower resource settings, on average, for a population of 1.7 million (Patel & Thornicroft, 2009; Jacob et al., 2007). Most LMICs do not have training programs dedicated to mental health services (WHO, 2011).

This challenge has led to the growing movement advocating use of non-specialists in so-called task-sharing models to provide evidence-based psychotherapeutic treatments (Verdeli et al., 2003; Patel et al., 2007; Saxena et al., 2007; Jordans & Tol 2013; Chatterjee et al., in press). In all the studies highlighted within this paper conducted within conflict-affected areas (Table 2), lay workers provided the evidence-based interventions. A recent Cochrane review (van Ginneken et al., 2013) identified preliminary evidence for the effectiveness of mental health interventions provided by such non-specialists in LMICs, for certain mental health problems. Although this is a currently recommended approach, there is agreement that there is still much to learn about how to do this well depending on the context (Pérez -Sales et al., 2011). For example, a recent review of a task-sharing approach concluded that there is a significant lack of data on the acceptability and feasibility of this model, with many studies not measuring and/or reporting on these variables (Padmanathan & De Silva, 2013). The task-sharing model must also adequately address the fact that individuals with limited training need different support than providers with more extensive mental health services training (who themselves need ongoing supervision). A multi-level system apprenticeship model of training and supervision has been used to provide layers of ongoing support and supervision, and ensure fidelity (Murray et al., 2011; Jordans et al., 2012). The apprenticeship model employs layers of support (e.g., direct providers, local supervisors, experts) and involves a process over time including initial training, practice among peers, limited clients with close supervision, and eventual (based on skill) graduation to more independent use of the EBT. For example, in the trial in the Democratic Republic of the Congo (Bass et al., 2013), a system was set up with community-based service providers who were supervised by Congolese psychosocial supervisors associated with an international NGO. For the trial, the supervisors were provided in-country supervision by a U.S.-trained social worker (limited Cognitive Processing Therapy specific expertise), who in turn was supported long-distance by the U.S.-based Cognitive Processing Therapy expert trainers. Three more recent studies have been completed on a Common Elements Treatment Approach (CETA), which is a transdiagnostic approach that teaches counselors certain components and then how to put the components together in particular ways to treat varying symptom presentation (see Murray et al., 2013b for details). In the CETA trials done in Iraq, Colombia, and the Thailand-Burma border, local lay counselors were supervised by local supervisors who were all trained simultaneously. Supervisors had weekly phone calls with U.S.-based CETA trainers for the duration of the study reviewing every case for fidelity and continually training supervisors – both on the treatment components and the ordering/sequence.

Attrition of Personnel

A related challenge is attrition of lay workers taught to provide evidence-based interventions. Most countries or projects have yet to devise a formal job position or title for the individuals now trained to deliver evidence-based interventions in non-specialist settings. Without a formal “place” in the health infrastructure, problems arise such as limited future work opportunities and/or individuals not using their skills past a particular project life (Padmanathan & De Silva, 2013). Attrition due to situations like health issues, maternity leave, and/or displacement is also common and requires the ability to constantly re-train a non-specialist workforce in evidence-based interventions. The relocating of health workers to other areas was also a factor in Iraq and Timor-Leste based on the author’s experiences.

One idea yet to be realized is to engage policy makers enough to create a new “place” within the health care system for those trained in evidence-based interventions for mental health. Research is needed to evaluate the added value of dedicated mental health workers (e.g. counsellors) providing EBTs as a new cadre within the health system. This research is currently ongoing in Nepal, as part of the PRIME program (www.prime.uct.ac.za). To deal with loss of counselors over time, one strategy often suggested is to train locally based individuals in how to train on the evidence-based interventions. A Training of Trainers is a specific form of training that seeks to provide would-be trainers with the necessary knowledge and skills to become trainers themselves (International Labour Organisation, 2002; see also USAID, 2008). There is limited research on whether existing Training Of Trainers efforts or programs have achieved a desired level of change in capacity of individual trainees (Davis et al., 1999 ; Steinert et al., 2006).

Another strategy to reduce attrition of personnel is to build the capacity of an educational institution so that there is a formal degree program related to the ongoing implementation of EBTs in mental health (Thornicroft et al., 2012). For example, institutionalizing a new curriculum on effective treatments within the local mainstream educational system. In Nepal, for example, this was tried by developing a one-year post-graduate diploma course in psychosocial counseling (Jordans et al., 2003). In conflict-affected settings, the lack of stability or organization within a government often makes this particularly challenging. In Iraq, the authors (LM, PB, GZ, AA), as well as members of the Ministry of Health, have discussed this strategy with local universities and attempted to engage them. University representatives stated challenges such as finances, lack of human resources (professors, trainers), and low priority for mental health education compared to other academic interests.

Danger and Instability

Perhaps more specific to conflict affected regions are dangerous events that occur prohibiting ideal implementation. For example, during the trial in Southern Iraq (Weiss et al., submitted), many of the study sites experienced bombings, political events that prohibited movement, and/or increased security check-points. During the trial in the Democratic Republic of the Congo (Bass et al., 2013), rebel attacks on the study villages were not uncommon, resulting in study participants and the treatment providers having to sometimes spend the night in the bush. In Colombia, there were multiple political events that caused an area to be ‘shut down’, where individuals were not allowed to travel for a period of time, and occasionally abductions took place. Events such as these can drastically affect the ability of clients to access services, for counselors to see clients safely, and for supervisors to communicate with counselors.

Approaches to manage this risk have varied widely depending on the context and extent of danger. In the studies listed in Table 2, the authors worked closely with locally based organizations that had intimate knowledge of the context and already had a wide range of safety procedures in place, and these were adopted allowing for mental health program implementation despite the danger. Additional strategies have been used with some success. For example, in the Democratic Republic of the Congo, if supervisors could not get to the villages to provide in person supervision, phone-based supervision was instituted. In the trial completed in Southern Iraq, the same solution was used when road blocks were put up prohibiting travel due to bombings and/or violence. During the trial on the Thailand/Burma border with torture-affected Burmese refugees, counselors saw many of their clients in or around their homes due to the challenges of a refugee moving around in the open. In conflict-affected areas, implementation plans that take into account possible disruptions like these is critical.

Lack of facilities

Although integration of mental health within primary health care settings is a recommendation (e.g., Yasamy et al., 2011; Lund et al., 2012), these efforts are often not accompanied by the resources needed to expand primary care facilities to accommodate more and/or different patients (i.e., those coming for mental health-related problems). A severe lack of a private space, particularly locations that are confidential enough for discussing mental health issues, is often a major challenge. This is closely linked to stigma (reviewed below). In addition, health care settings are often not close enough to allow for regular attendance for counseling/therapy sessions. This is particularly challenging in rural settings, such as our study sites in the Democratic Republic of the Congo and a mountainous country like Nepal, where primary health centers are often a day or two walk away from villages. Finally, there is often an unmistakable scarcity of facilities that address high-risk situations – for example child abuse and neglect, gender-based violence or suicidal behaviors.

One approach to the lack of space is for counselors to meet the clients in other locations – spaces that are comfortable and private and more proximal to where the clients live. Some of these may include community centers or religious spaces. In the Democratic Republic of the Congo, group therapy was provided in a small building located in the study participant’s village so they could easily access it for treatment and also see the counselor outside of session if they needed additional services. Engagement with clinic directors and the larger Ministries of Health in some cases has helped in developing schedules for the limited space. For children, especially, it is deemed important to aim to provide services in a non-stigmatizing, separate environment – for example, in schools (Jordans et al., 2013).

An approach to the lack of facilities for high-risk situations has been to develop creative, setting-specific safety plans that find and/or create local services or organizations that can help with issues like suicide, homicide, and abuse (Murray et al., 2014). On the Thailand/Burma border, a safety plan began with calling supervisors to further assess the risk, and then engage a local doctor who was willing to be ‘on call’ for high risk cases. The protocol included engagement of multiple staff and resources on the ground to find a solution specific to each individual case (e.g., a personal visit by the doctor, a 24-hour watch set-up, entry to a local shelter for a short period of time). In the event of a safety issue in northern Uganda (within an internally displaced persons camp), research personnel also called a local supervisor who may personally assess the adolescent. Given the long distance to a hospital, if the supervisor was concerned about the youth they would work with the family and others within the camp (e.g., neighbors, camp leaders) to develop an overnight watch until the supervisor returned the next day to re-assess.

Transportation

Travel to receive services is an issue for mental health programs across almost all areas of the world and arguably may be even more problematic in conflict affected areas. In all the studies completed by the Applied Mental Health Group (http://www.jhsph.edu/research/centers-and-institutes/center-for-refugee-and-disaster-response/response service/amhr/index.html), a significant proportion of the budget had to be dedicated to transport of clients to be able to access services, counselors to be able to serve clients, and/or supervisors to consult with counselors. In terms of DI, this ‘logistical’ category significantly affects sustainability post-project as it is unlikely governments in LMIC would be able to provide this. Literature from the Middle East also cites insufficient local transportation as a major problem (Eapen and Ghubash, 2004).

The option to move services further into the community and away from primary care settings where services are sparse has been and should continue to be investigated. In (post-)conflict settings, such an approach may have the added benefit of addressing the damage to the social fabric caused by armed conflict (Scholte et al., 2011). For example, health houses or posts are increasingly cited in policy documents as a potentially closer layer of primary health care services to the population (MoH/Afghanistan, 2005, MoH, 2009). In our work, this strategy is helpful in some ways, but still requires transport for supervision in most cases. In Democratic Republic of the Congo, services were provided closer to consumers in the villages, but still required significant transport resources for supervisors. The Basic Health Services Package for Iraq also utilizes this concept where a rurally-located center has 2-3 community health workers trained in the provision of basic health services (Ministry of Health Iraq, 2009). The strategy to move services into the community is also being implemented in Nepal. In most cases, resources for ongoing supervision would still be needed, unless over time trained supervisors were also positioned in these rural settings, or in some instances if technology was dependable enough that they could be reached if needed.

Stigma

Lack of knowledge and/or stigma about mental health services has long been identified and documented as a key implementation barrier (e.g., Thornicroft, 2006). Patterns of stigma have been identified across the globe’s regions, but specific patterns and their impacts on mental health vary across socio-cultural settings. One example may be when a primary health clinic is “known” in the community as treating diseases like HIV/AIDS, and thus seem more stigmatizing to those attending the clinic for mental health treatment.

We have employed two overlapping strategies. First, to adopt language that is least stigmatizing within the local setting. This means moving beyond simple translations of mental health terms, instead assessing through ethno-psychological studies what are the best suitable terms to be used (e.g., Kohrt & Hruschka, 2010) and/or conducting qualitative studies to understand local terminology (e.g. Bass et al., 2008). Second, in several of our programs, psychotherapeutic services are systematically combined with efforts to reduce the levels of stigma among people within the community, as well as among service providers themselves (e.g., trainings, dramas done within community settings). Another strategy is to engage the support from the community at all the stages of implementation. For example, in Colombia, the implementing organization worked with individuals identified as ‘community leaders’. These leaders helped the implementation process by facilitating access to remote areas and addressing stigma in the community.

Going beyond this, despite clear evidence that stigma and discrimination are both common and severe worldwide (Lasalvia et al., 2012; Thornicroft et al., 2009), there is at present a clear information gap concerning effective interventions to reduce stigma and discrimination in LMICs at all levels. Developing and testing candidate interventions for stigma is a priority in the coming years and could greatly affect the DI of evidence-based interventions.

Lack of Leadership

Leadership has the capacity to help or hinder implementation of an innovation and change within a system (Aarons, Sommerfeld, & Wilging, 2011; Damanpour & Schneider, 2006). Identifying leadership in health systems in post-conflict settings may be particularly challenging when there is frequent turn-over of personnel and disorganization of previously established health system hierarchies. This may affect the range of health system actors from the level of the Ministry of Health to remote community health systems. For example, experience by authors from Iraq (AA, GZ) describes how effective leadership on mental health initiatives was achieved within the Ministry of Health, but quickly challenged due to personnel changes. Similarly, in humanitarian settings, funding priorities may change drastically with frequent changes in personnel of donor organizations – thereby affecting the longer-term commitments that recipients of funding can make.

In the process of a program that aims to integrate mental health into primary health care in post-conflict Nepal, some authors have found changes at the policy level are highly challenging with a fragile government system. With no mental health focal point within the Ministry of Health, different people function as de facto focal points. After considerable investment and headway in explaining the importance of the Ministry of Health’s role in supporting the process of development of mental health services, the process begins from scratch again when the person is shifted to an entirely different position. With little structure or policy to fall back on, this can happen several times over a couple of years.

Despite the fact that personnel may change, there is still a general agreement that DI efforts need to include engagement of leaders, stakeholders, and policy level personnel. Attempts to expand the reach of these engagement efforts so that buy-in is ‘deeper’ than one person has been met with mixed success. In cases where government leadership is still in flux, as is often the case in post-conflict settings, we are attempting to engage leadership from other organizations that are closely linked and/or respected by the government. In Burma, for instance, there are a number of organizations that are able to influence policy at times and have been helpful advocates for the implementation of certain evidence-based interventions.

Lack of Trust in the system

Trust in the quality of health services, treatment fidelity, and human treatment of clients and patients are crucial requirements for the successful implementation of services. Many factors undermine the trust of populations and individuals in health services provided at public health facilities (Gilson, 2003; Whetten et al., 2006), including the destruction of the physical infrastructure, limited investment, the politicization and militarization of health services, the inequitable distribution of health care resources, unregulated privatization, and the brain drain of professionals – many factors which are exacerbated in times of conflict (Dewachi et al., 2014). In addition, increasingly vocal media and civil society organizations are exposing medical errors and incidents on some areas further contributing to the skepticism towards traditionally patriarchal health service provisions (El-Jardali et al., 2010; World Bank, 2013).

As a concrete example, the Maoist insurgency in Nepal was partially aimed at reducing the unequal distribution of governmental (health) services in urban and rural areas. However, the armed conflict increased this divide (Tol et al., 2010a). As another example, in Iraq, many individuals who are able are leaving middle-eastern countries to receive health care in other countries because they do not trust the existing medical services (e.g., Dewachi et al., 2014). If the community does not trust the overall health system and mental health services are integrated into this broader system, this will become yet another barrier.

Certain strategies have been discussed and/or promoted, but none have been scientifically evaluated in LMICs. For example, there have been calls for engaging general community members and mental health service users in the design, formulation, implementation, and evaluation of health policy in general and mental health services’ provision in particular. This type of participatory ‘people-centered approach’ to the provision of mental health services may include community input into legislation or the engagement of civil society organizations in building trust and enhancing knowledge about mental health disorders (Ruggeri & Tansella, 2012). In some LMIC this may include reaching out to traditional healers and treating them as partners, with a goal of agreeing on a delineation of tasks and responsibilities.

Policy

Policy plays a critical role in the dissemination and implementation of services. A fundamental problem underlying the treatment gap is the low priority accorded to mental health issues by most governments in LMICs (Saxena et al., 2007, Bird et al., 2011). In humanitarian settings, a collection of ten case examples in low- and middle-income countries highlights the important role that policy reform played in ensuring that the initial emergency-related mental health services were eventually translated into sustainably improved mental health care systems (WHO, 2013). Humanitarian settings may in fact form a good time to reform or implement existing policies, as there may be a heightened willingness on behalf of policy makers to consider the importance of mental health services.

Some researchers suggest using an argument of advocacy to policy makers – essentially positing that access to mental health services as a fundamental human right (Funk et al., 2006). Others suggest using education to influence policy, stating that there is a significant lack of broad public health perspective among leaders in the mental health community (Saraceno et al., 2007; Jenkins et al., 2010). A recent article outlines an attempt to put such solutions into action (Abdulmalik et al., 2014). The Mental Health Leadership and Advocacy Program was described as a process for individuals in a leadership role in government and civil society with the goal of educating leaders on mental health and promoting its advancement among policy priorities (Abdulmalik et al., 2014). The authors document success based on outcomes such as planning meetings, advocacy activities (e.g., radio shows), and, for some countries, a mental health plan within the government.

Funding

Perhaps most daunting and cross-cutting among challenges in conflict-affected LMICs is the shortage of funding to implement, support, and sustain services. According to the WHO, US$60 per capita healthcare spending is necessary to cover basic health needs in LMICs (Gostin, 2012; WHO, 2010b). Mental health care allocations within health care budgets are virtually non-existent, with LMICs devoting on average 1% or less of their health budget on mental health, with this funding often going to psychiatric hospitals in large cities (Saxena et al., 2007). As a result, individuals and families depend on out-of-pocket expenditure with potential impoverishing effects (Kankeu et al., 2013; WHO, 2010b).

Furthermore, mental health problems themselves represent a significant factor in earning loss for affected individuals and families (Levinson et al., 2010). In (post)-conflict settings these resource limitations often lead to an over-reliance on external emergency funds and/or NGOs to provide and finance mental health services. Financing of NGOs can vary widely and tends to fluctuate with changes in funder priorities and political influences. For example, an analysis of funding for mental health and psychosocial support programs in humanitarian settings found that less than 15% of funding was disbursed through existing medical, social welfare, or primary education systems – thus raising concerns about the sustainability of funding through shorter-term emergency mechanisms (Tol et al., 2011).

Although well-known, this challenge has not been well addressed. One strategy the authors have used is to engage the Ministry of Health (or other governmental agency) early in post-conflict work funded by outside organizations so they can see the importance, challenges, and successes. These efforts may include working to have mental health listed as a priority within a National Strategy Plan through education around the effects of mental health problems – for example, on productivity. This has been met with limited success. For researchers and program evaluation staff, another idea tried is to include cost-effectiveness analyses that aid governments and funders in understanding the costs of mental health programming and the associated outputs. There remain many financial challenges in transitioning funding from outside donors to governmental health authorities within the country, even when there is a commitment by the government (WHO, 2013). This is clearly an area where attention is needed to more rigorously evaluate promising models.

Looking Ahead: A DI Research Agenda for (post)conflict LMICs

This paper introduced the importance of DI research and laid out some of the primary implementation barriers experienced in the testing of evidence-based interventions in post-(conflict) LMICs. Although the studies highlighted within conflict-affected areas included creative strategies for addressing these challenges of implementation, there is much work to be done. Rigorous research in the dissemination and implementation of evidence-based mental health practices globally remains sparse and is methodologically limited, particularly in post-(conflict), lower resourced settings.. We advocate that DI research needs to become a priority within LMICs to realize actual implementation of the growing body of interventions that have evidence of effectiveness, with the eventual aim of reducing the treatment gap.

In our opinion, a DI research/evaluation agenda should aim to address at least three key issues. First, more knowledge is required on defining concepts and developing measurement tools for low resource settings. Even in high-income settings, these are identified challenges ‘at an early stage, leaving the field without clear directions for conceptualizing and evaluating [DI] success’ (Proctor & Brownson, 2012; Proctor et al., 2011). The development of DI measurement tools in LMICs will need to factor in key differences from high-income settings such as limited specialist work force, diverse organizational structures, no accessibility to ongoing training, and weak funding infrastructures. For example, existing DI instruments from higher income settings have items asking about insurance plans or conference attendance that would not be relevant to most low-resource contexts (Murray et al., 2013a). Assessment tools will be critical in order to advance any dissemination and/or implementation goals.

Second, there should be deliberate examination of implementation outcomes measured before, during, and after implementation of evidence-based interventions for all programming. This was considered a major research priority in a consensus-based research agenda for mental health and psychosocial support in humanitarian settings (Tol et al., 2011b). There is substantial documented experience to build on, including years of field-based mental health programming leading to anecdotal reports and/or case studies (e.g., WHO, 2013; see also an Intervention special issue from 2011 on integration of mental health into primary health care in emergency contexts – Ventevogal et al., 2011). However, without more rigorous knowledge on DI outcomes across various settings, as well as the synthesis of this data, the effort in identifying effective treatments in LMICs’ post-conflict settings will likely not translate to actual improvements in accessibility to services for affected populations.

Finally, a critical area of the dissemination and implementation science field that is often neglected is the systematic evaluation of the effectiveness of certain implementation strategies through randomized controlled trials (Proctor et al., 2009). Specific to LMICs, and even more so in conflict-affected areas, part of the challenge is to design and evaluate implementation strategies that both work and are sustainable in such low resource contexts. For example, although an apprenticeship model has been used in many trials, it requires investments of time, resources, and expertise that are difficult to sustain. One direction would be to conduct studies to unpack the critical components and minimal time frames of training and supervision to still produce well-trained staff in order to make the best possible use of scarce resources. Certainly implementation strategies need to be examined to determine how to effectively engage policy and governments in mental health integration.

In delivering these three aspects of future DI research, collaborations between universities, governments, and implementing organizations will be crucial. Collaborative DI projects between academic, implementing, and policy organizations provide ideal contexts for learning to flow both ways. For instance, to academics about which effective programs have better chances for sustainable implementation, and to implementing and policy organizations about systematic evaluation of implementation. Such collaboration could take the form of adding DI indicators into routine program monitoring and evaluation and/or becoming a secondary aim in an effectiveness trial. Feedback loops will also be critical in that if a treatment or program is difficult to scale up, it may be too complex and need to be re-evaluated in a simplified format to aid in dissemination and implementation efforts.

Conclusion

The field of global dissemination and implementation needs to advance significantly to realize any longer term sustainability of implementation of EBTs leading to a reduction in the treatment gap. Three major directions laid out include: 1) definition of DI concepts and development of DI measurements appropriate for low resource settings, 2) evaluation of implementation outcomes within the numerous programs running in LMICs; and 3) study of of implementation strategies (rather than anecdotal, or descriptive case studies).