Managing Ethical Challenges to Mental Health Research in Post-Conflict Settings

Who Conducts the Research?

In social research gauging an appropriate distance between researcher and participant – neither too familiar nor too distant13 – is important for methodological rigour. Often the ‘appropriate’ distance is enmeshed with adherence to cultural norms, and is therefore affected by who conducts the research. This is important in mental health research where effective interventions require attention to the cultural context in which participants are embedded.14 A local research team who can advocate for culturally centred interventions,15 and research processes that respond to cultural context can aid acceptability of services and research.

Who conducts research carries implications for research capacity building, defined as the ability to conduct, manage, disseminate, and apply research in policy and practice.16 Gaps in LMIC mental health research capacity have been identified at every level: individual, organisational and national.17 The lack of sufficiently trained and experienced local researchers carries ethical implications when research is conducted by those unfamiliar with local context or without sufficient expertise to maintain ethical research standards.18 Consequently, the benefits of building local capacity should not be underestimated,19 including better integrating LMIC perspectives into research agendas and practice in global health.20 Capacity building requires long-term investment and commitment – including recognising research as a viable career.

Developing local capacity and partnerships also provide routes for initial contact with communities that can increase the acceptability of those conducting research. The way communities are approached has been highlighted as critical to ‘ethical entry’21 appropriate to local cultural norms. Partnerships with organisations embedded within local communities also provides routes for researcher accountability to the community.22 When negotiating access to communities, particularly in conflict or post-conflict situations, awareness of power relations and who is identified to represent a community are pertinent.23 The potential for researchers to be perceived as supporting one side or another, or privileging certain accounts requires careful attention24 and can be exacerbated by researchers working outside of local community systems. Researchers must remain mindful of who they are provided access to, and issues that might arise from only engaging those with the power to speak out, further disempowering those without a voice.25 Responding to these considerations is contextual, favouring a situated approach to how ethical entry is managed.

Local research teams are well placed to manage researcher safety. Craig et al.26 identify race, gender and culture as potentially impacting upon researcher safety in violent contexts. They advocate addressing safety by matching these and other important characteristics of the research team with the community. The issue of researcher safety is discussed in more detail below.

Therefore, for a range of ethical reasons it is maintained that ethical research conduct requires the incorporation of local researchers within the project team.

In the case study researcher matching and ethical entry were ensured through a local research team, critically incorporating local females including two mothers. These female researchers shared important characteristics with participants and were widely respected within the community. Gender matching researchers and study participants increased the acceptability of the research, strengthening mutual trust and rapport. Appropriate distance between participant and researcher was assured by adherence to local cultural norms such as dress codes including observing purdah (the practice of females wearing a veil and being segregated from men who are not family) and speaking the local language. Researchers were not personally known to participants, but came from the same region, deemed important for building trust and ensuring confidentiality.27 Ethical entry was achieved through partnership with a local organisation and respecting cultural norms by discussing the study with community elders and health workers prior to its commencement. Negotiating access to female research participants required attention to gender power imbalances as well as cultural norms relating to decision-making authority (discussed under ‘informed consent’ below). Ongoing activities in the research site continue to draw upon the skills and expertise of the trained research team, contributing to local capacity building and embedding local partnerships for future research.

Who Funds the Research?

The manner in which research is funded in emergencies carries at least two important ethical considerations: first, in post-conflict situations is the funding source, or country associated with the source, seen to be party to the conflict? This could put researchers and participants at risk of harm. Second, do funding structures enable research that leads to tangible improvements in participants lives in the short and long term?

In post-conflict settings who fund the research can lead to positive or negative views of the study depending upon local populations’ perceptions of outside agencies. This is particularly pertinent when bodies perceived as party to the conflict fund research, and can impact upon researcher safety.

The question of disclosing research funding raises competing ethical duties. On the one hand there is a duty to develop mental health interventions by conducting research which requires funding, and there is an established ethical obligation to disclose funding sources to participants. On the other hand following this ethical obligation could potentially put researchers at risk. Therefore two ethical principles come into conflict – the disclosure of funding sources to participants, and the duty to ensure researcher safety. To address this we recommend conducting an assessment to consider the ways funding agencies may be viewed by the local community, and how this may impact upon researcher safety. Assessment findings should be shared with funders to negotiate an approach to funding disclosure, considering adjustments to the obligation for full disclosure. This approach is not without its problems. For example, could failure to fully disclose funding sources be viewed as deception, presenting a risk should it become known that researchers were not open with participants? Resolving these tensions requires an approach that accounts for local factors, and not a ‘one size fits all’ ethical requirement.28

A further issue is the ethical obligations of funders. Schopper et al. highlight ‘reasonable availability’ of an intervention post-research, defined as a commitment to deliver services for a minimum of two years, or that it remains available through other means.29 Similarly, the Council for the International Organisation of Medical Sciences (CIOMS) identify as ‘morally praiseworthy’30 the sponsor funding services beyond the duration of research, with this commitment outlined in research protocols; guidance that is reflected elsewhere.31

The issue of research funding should be included within ethical risk / benefit analysis. In the field of mental health research benefits typically include influencing the design and delivery of services. This is the aim of WHOs ‘Building Back Better’ which advocates strengthening health systems in the immediate aftermath of emergencies for long-term benefits.32 However, policy process can be lengthy meaning participants may not see research impact. This is particularly the case where intervention delivery is tied to short-term funding rather than a commitment to embed services into routine care. Delays in research impact or short-term services can create poor perceptions of research participation.

The Hastings Centre consider the duty of ‘fair benefits’ from research participation as laid out in CIOMS.33 They argue that ‘fair benefits’ is poorly operationalised, with lack of clarity over who is to benefit (research participants, the wider community, a whole country?) and who is responsible for funding this benefit (research funders, governments, international organisations?). They support the community deciding the value of fair benefits of research participation, and what these should entail. Therefore, ethical research is not tied to continued access to services, but could include benefits such as capacity building of local service providers or researchers, contributing to health infrastructure, or financial reward. This raises a duty for researchers and funders to engage with communities to determine how benefits can be ethically distributed, delivering immediate and long term benefits of value to the community.

In the case study the issue of who funds the research was minimised as funding came from a National Higher Education body and a recognised local NGO. These funding sources increased local acceptability of the research and were fully disclosed to study participants. Future research benefits included an intervention provided by embedded health workers which continued to be delivered beyond the research, and capacity building of both healthcare providers and researchers. These benefits were discussed informally with healthcare providers prior to research, exploring how to embed research into existing services. Capacity building was viewed as of particular benefit and involved training healthcare staff at two centres in the mental healthcare needs of perinatal women.

However, in light of community mistrust it is possible that some participants considered the local funding sources as a route for government authorities to extract information regarding involvement in insurgency activity. Therefore, it is maintained that an assessment of the socio-political context be conducted, including in internal conflicts where community allegiances may lead to local funders being viewed with suspicion.

Ethical Review

Research ethics review is a procedural cornerstone of international guidelines on human subjects research.34 However, lacking or dysfunctional review boards in many LMICs contribute to inadequate ethical research standards.35 Challenges ranging from review boards’ legal status, workloads, and differences in expertise and procedures contribute to disparities in the review process.36

Developing research ethics committee members capacity is frequently highlighted as a way of ensuring reliable interpretation of international ethical guidelines for socio-economic and cultural conditions.37 This concurs with calls for in-country review to judge ‘ethical acceptability of the research in accordance with the customs and traditions of the community’, involving lay persons to review research against community cultural and moral values.38

Advocating formal ethical review is premised on the view that when conducted well feedback can be instrumental to ensuring research maintains ethical standards. A subsidiary aim is to stimulate a conversation between researchers and reviewers, seeking consensus on how to manage potential ethical issues. In the ethical review process it is important to recognise informal community-level procedures for reviewing research that operate alongside formal review, ensuring that the latter does not usurp the former: conversations with ethical review boards should not replace conversations with communities involved in research. An iterative process between ethical review boards and communities to identify, define and negotiate the way ethical challenges will be resolved is recommended. This approach problematizes the priority of formal ethical review, with the strongest process being one that balances formal review with community-led processes. An iterative approach acknowledges that most ethical issues arising in research implicate a number of principles which requires a process of judging the relative weight to be accorded each principle.39

In complex contexts it has been suggested that those conducting ethical review are often in a ‘double-bind’: they recognise the risks and potential for exploitation, but have little practical guidance to offer on the management of ethical issues.40 This suggestion concurs with recent research identifying the paucity of guidance from ethical review committees on a study conducted with IPDs.41 Involving the community to collaboratively design research, including developing responses to potential ethical issues, as well as having community review of research, offer potential mechanisms to address this ‘double-bind’.

Challenges to iteratively developing specified ethical standards for international mental health research are recognised, not least of all the time required. Another difficulty is the complexity of ethical review at multiple levels – community, in-country, international and increasingly funder review. However, these processes aid development of ethical standards for specific studies that adhere to both local and international norms, and support researcher preparedness by thinking through ethical issues prior to data collection. Multiple levels of review can also stimulate ethical review committees cross-learning: educating international committees on country context, culture and moral values; in-country committees can see how international committees work; and both local and international committees can learn from community responses to potential ethical issues. Therefore, due to potential long-term benefits for specific research studies and a broader moral conversation we argue for using the opportunity of engaging with multiple levels of review and iterative development of ethical standards to build examples of best practice for managing potential ethical issues in a range of contexts.

Whilst this proposal could be charged with being idealistic, it is countered that researchers routinely engage with procedural demands which frequently entail multiple levels of ethical review, for example in the country where the study is to be conducted and in the sponsor country. This demonstrates that with sufficient planning and researcher commitment multi-level review is possible. A further challenge is presented when seeking to follow the process of multi-level review in settings where no local review structures exist. In this case we recommend a peer-review assessment of the protocol either by local academics, practitioners or community members to ensure research is critiqued from a local perspective.

In this case study formal ethical review was obtained both in-country and at a UK University. Community perspectives were represented by the locally-based researchers who took an active role in developing the study protocol and acted as cultural brokers, identifying potential ethical challenges and suggesting routes to manage these – such as the informed consent process discussed below. Additionally, sharing proposed research with community health workers and elders provided informal community-level review. Contrary to the above discussion on the opportunities for cross-learning through ethical review processes this was not experienced in this study. Approval at all levels was provided without comment on the potential ethical issues that may arise. Therefore, management of in-practice ethical issues relied heavily upon informal local review and comment, researcher integrity, and knowledge of the study setting rather than formal ethical review processes. It is not known if this missed opportunity is a result of those conducting review feeling they were in a ‘double-bind’ or due to poor capacity and review procedures.

Voluntary Informed Consent

Research guidelines recognise informed consent in LMIC raises additional cultural considerations, including the issue of gatekeepers46 and differing conceptualisations of ethics and rights.47 Attending to power relations is also identified,48 with one study seeking to mitigate power hierarchies’ related to religious, community and political leaders, as well as the status accorded to medical professionals and researchers which can create undue inducement to participate.49

In this case study gatekeeper consent was conducted, respecting local cultural norms. This entailed obtaining prior consent from household males and elders to seek consent from the female participant. This can be viewed as taking consent from multiple levels or ‘spheres’,50 including whole communities, community leaders or elders, families, and individuals as appropriate in the setting. This process presents ethical risks as it is possible that gatekeepers may not allow an individual to participate. In this circumstance the participant is unable to exercise their right to make an informed choice to participate. This presents an ethically charged dilemma for researchers balancing adherence to cultural context with ethical and human rights norms.

Chambliss suggests informed consent ‘represents at best a polite fiction’,51 a view pertinent to emergency contexts.52 In the case study before research assistants sought informed consent the research was introduced by a community health worker who provided a short explanation of the study and asked permission for a research assistant to meet with the individual. Through training, research assistants’ self-awareness of the impact disparities in education and status could have upon making an informed choice to participate were raised, and the participants right to refuse to participate without penalty was reinforced. Therefore, each step in the informed consent process sought to protect participant rights whilst remaining compatible with local cultural norms.

Additional safeguards were also considered including taking repeated consent, an approach adopted through repeat verbal consent. This compromise aimed to minimise raising anxieties in relation to the research purpose, and formed one aspect of addressing community mistrust through consistent articulation of the research process, discussed below. A challenge to repeat consent is that it could promote higher attrition rates, something particularly relevant to randomised control trials.

Relating to procedural ethics, in the case study, due to high rates of illiteracy all research information was explained verbally in the local dialect with a thumb print accepted in lieu of a signature, following standard practice in the context. It is important to note that in different settings a thumb print can itself carry negative connotations. Other alternatives for recording consent with illiterate populations are to record verbal consent or have researchers witness and verify consent on behalf of participants. However, both strategies present ethical challenges. In some contexts, including the case study, recording is not acceptable to the local community or presents risks to confidentiality. Equally, to have researchers verify consent on participants’ behalf can be considered insufficient protection against coercion. Consequently, decisions relating to how to record consent must be carefully examined with local researchers who can act as ‘cultural brokers’ to ensure acceptability of consent processes.

In the case study despite providing information in the local language and attempting to overcome illiteracy through verbal explanations of the research, rumours of threats to the research team called into question how far confidentiality, anonymity, and protection of participant rights was understood. Difficulties translating concepts such as anonymity and confidentiality into the local dialect raise questions as to the meaning participants ascribed to them. Perceptions that interviews are collecting information to pass onto intelligence authorities are more likely to occur where the population feel threatened, such as conflict or post-conflict settings. This indicates that whilst consent processes can be culturally adapted, they may be unreliable when undertaken with an illiterate population who feel threatened.

Accordingly, the case study consent process in some cases failed to meet Horman's element 2: comprehension of research information.53 This raises a critical ethical dilemma: how to ensure information is fully comprehended at the time of obtaining consent? Moreover, what are the implications for consent should it transpire that information was not fully comprehended? These strike at the heart of the principle of informed consent, and have been discussed elsewhere.54 Strategies for managing this in LMIC have been proposed including: placing emphasis upon the process of information exchange over formal recording of consent;55 providing information in lay language appropriate to local literacy,56 where required including images or video to aid communication;57 and conducting an oral examination to gauge sufficient knowledge to make an informed decision about participation.58 The first two strategies offer routes to enhance the quality of information exchange and are deemed appropriate in emergencies. An examination to verify comprehension is deemed inappropriate given the potential for disempowering potential participants. However, the principle of asking participants to reflect back in their own words their understanding of research has been recommended59 and is considered a less formal approach to verifying study comprehension to make an informed decision about participation. Furthermore, relating to information, concerns about the way information is constructed and presented have been raised.60 This is important in emergencies where clear and unbiased presentation of information is critical to avoiding exploitation.

As this brief discussion emphasises, the practice of ensuring voluntary informed consent is frequently complex, requiring researchers to judge the quality of consent. It may only be once research is underway that it emerges to what extent the information provided during consent was understood by participants. Therefore, ethical standards may require acknowledgement of situations where it may not be possible to obtain fully informed consent due to contextual realities such as a climate of fear. In such circumstances a more nuanced view of consent may be appropriate with alternative guarantees of ethical research practice. In this regard the biomedical field could learn from the work of social scientists and anthropologists.61 The possibility of negotiated consent involving collaboration with the community and flexibility as to what consent ‘looks like’ by research ethics review boards (i.e. not dependent upon signing a form), is believed to offer potential for more culturally appropriate and robust consent processes. This more nuanced view is felt to be suitable for emergencies.

Community Mistrust

Managing paranoia or mistrust over the way information collected during research is to be used requires careful handling in communities exposed to conflict. Craig et al. identify that it can be necessary to equip researchers with tools to respond to strong feelings or angry reactions to research by participants and the local community.62 This is particularly important in mental health research where stigma and discrimination are common, reinforcing the importance of a trained research team.63

Promoting trust requires active communication and mutual understanding between researchers and the community. At a minimum communities should be consulted during the research planning stage, on an ad-hoc basis whilst the research is conducted, and provided with research findings in an appropriate format and timely manner.64 In addition, care should be taken to distinguish between routine care and participating in research in order to avoid therapeutic misconception.65 These measures are essential to reducing community misperceptions about research.

In the case study community mistrust was an important ethical concern. When going to homes in the community researchers were confronted with families fearful for their safety. On rare occasions researcher safety was threatened when rumours of threats against the research team spread as a result of questions relating to exposure to violence. With families active in the insurgency these questions were interpreted as collecting information to pass onto intelligence authorities, carrying implications for participant recruitment and increasing risk to the research team perceived as acting at the behest of the government.

These issues were managed by (a) integrating local community members into the research team, and (b) suspending research activities whilst mistrust was addressed by the study lead, a precaution designed to ensure researcher safety. The locally based study lead and researchers ensured an ‘ear to the ground’, seeking to anticipate community mistrust and respond accordingly. In the post-conflict study setting the integration of local researchers increased the credibility of claims that research was not collecting information for intelligence authorities.

During the suspension of research the study lead conducted repeated research information events with those making threats. This involved detailing the broad study topic, what participation entailed, how information would be used including protection of participant anonymity and that personal information would remain confidential to the research team. Opportunities for questions and discussion were provided. In addition, as identified above, researchers conducted repeat verbal consent, reinforcing key principles of voluntary participation and protection of participant rights. The presentation and re-presentation of information sought to counter community rumours and mistrust, re-engaging participants and the wider community in research. This strategy proved effective in this context. Therefore, the response to this ethical issue was locally specific, developed by embedded researchers in accordance with local cultural norms.

Risk to the Research Team

Research investigating sensitive topics needs to assess the potential risks research poses to both participants and researchers.66 This is equally important when researchers are working in sensitive contexts such as emergencies. Examples of safety risks include threats to physical safety; risk of psychological distress; potential for accusations of improper behaviour; and increased exposure to everyday risks such as infectious illnesses or accidents.67

Risk assessments are an integral element of developing a research protocol, including budgeting and planning to manage potential risks to participants and researchers.68 Managing risks to the research team is essential to ethical research which is dependent upon researcher competency to practice,69 including addressing stress and fear. Guidelines to support development of protocols to manage researcher safety have been suggested. These include steps to assess the situation, identifying and responding to threats, and developing preventative strategies and follow-up procedures,70 including addressing potential psychological harm.71 Psychological support referral pathways for research teams are essential when conducting research with populations exposed to violence or trauma, or where upsetting or difficult disclosure may arise. In addition, in emergencies processes for consulting local security experts72 as well as those coordinating the emergency response are also essential.

Recognising the potential emotional and psychological impacts of discussing sensitive topics is important to preventing researcher burnout.73 Mental health and trauma-related research may lead to researchers experiencing vicarious trauma: traumatisation through the act of bearing witness to the suffering of interviewees.74 Equally, failure to ask these questions results in an incomplete picture of mental health status and may miss cases of abuse or degradation, hence can be viewed as unethical not to ask.

Adequate attention to personal and psychological safety of researchers requires specialist training of research teams that emphasises strategies for researcher self-care, supervision and support.75 Such training addresses the concerns of Dunn that ‘the novice researcher is usually taught that the research process is orderly and straightforward’.76 They also offer an opportunity to bridge the gap between procedural ethics and ethics in practice, engaging with potential real-world difficulties that may arise in research conduct. Training therefore promotes a more nuanced approach to the way in which specific ethical challenges will be managed.

In the case study support was provided though daily meetings between the study lead and the research team. In these meetings the research team detailed the day's field activities and reported any events that had happened. These collective meetings provided opportunities for peer support and raising concerns. They also ensured the study lead was aware of field challenges and could monitor researchers’ wellbeing. In addition to these daily meetings the study lead was available via mobile phone for immediate contact in the cases of crisis.

One crisis arose during the case study where military raids of homes occurred whilst interviews were being conducted. This resulted in research team members becoming fearful for their safety, compounded by high profile insurgent activity targeting women and health workers. These threats to safety were managed through pausing the research for one week so as not to coincide with activities being targeted by insurgents. Given the similarity of the study to activities being targeted – a health campaign conducting house-to-house calls – this was felt to be appropriate, with research resuming only after the other activities had ceased.

This highlights the ethical duty to be flexible in the research schedule, suspending research to not coincide with activities of a similar nature being targeted in order to protect both researchers and participants. This carries ethical implications when projects are externally funded as hostile activity may prevent a study being concluded, including withdrawing when a mental health intervention is ongoing which may leave participants at risk, and the ethical implications of perceived wasted financial resources. Despite the potential risks it is important that mental health research with groups or in settings perceived high risk is conducted77 which requires robust risk management. This necessitates an approach to ethical research practice that responds to risks inherent to different contexts, employing culturally appropriate strategies to address and mitigate risk to ensure that research adheres to the principle ‘do no harm’. Sharing strategies for responding to risk in different contexts will promote a moral conversation to identify best-practice approaches to minimising risks.