Ethical Considerations for Disseminating Research Findings on Gender-Based Violence, Armed Conflict, and Mental Health: A Case Study from Rural Uganda

Jennifer J. Mootz, Lauren Taylor, Milton L. Wainberg, and Kaveh Khoshnood

Abstract

Gender-based violence (GBV) is a major public health problem that is exacerbated in armed conflict settings. While specialized guidelines exist for conducting research with GBV, guidance on disseminating findings from GBV research is scant. This paper describes ethical considerations of designing and disseminating research findings on GBV, armed conflict, and mental health (including alcohol misuse) in conflict-affected settings in Northeastern Uganda. Following completion of two research studies, we conducted a half-day dissemination meeting with local community professionals (n=21) aged 24 to 60. Attendees were divided into small groups and given a quiz-style questionnaire on research findings to prompt discussion. Two primary ethical tensions arose. One ethical consideration was how to disseminate research findings equitably at the participant level after having taken care to collect data using safe and unharmful methods. Another ethical issue concerned how to transparently share findings of widespread problems in a hopeful and contextualized way in order to facilitate community response. We recommend planning for dissemination a priori, engaging with partners at local levels, and grounding dissemination for action in evidence-based practices.

Background

Gender-based violence and armed conflict

Gender-based violence (GBV), identified as one of the most challenging and widely permeating human rights issues today, affects one-third of women worldwide.1 The United Nations has defined GBV as any act of psychological, physical, or sexual harm to women and girls.2 Some of the highest rates of GBV occur in sub-Saharan Africa. In some countries in eastern Africa, for instance, almost three-fourths of women report experiencing some type of partner-perpetrated violence in their lifetime.1 Exposure to GBV presents a tremendous public health problem through numerous physical, mental, and intergenerational deleterious effects.3

Armed conflict, between or within nations or among civilian factions, exacerbates GBV in communities.4 With critical breakdowns in legal and social infrastructures, impunity for perpetrators increases.5 Most victims are women and adolescent girls who experience GBV through sexual assault by combatants and soldiers, increase in human trafficking, and use as expendable bodies in conflict.6 There are also documented increases in household GBV in the form of intimate partner violence (IPV): intimate partner behaviors that cause physical, sexual, or psychological harm.7 While the relation between armed conflict and GBV, including IPV in households, has been observed across several contexts, little is known about mechanisms through which armed conflict affects GBV and IPV specifically.8 The first author has examined mechanisms between armed conflict and GBV in Northeastern Uganda and, in this paper, we discuss some of the ethical challenges associated with designing GBV research and disseminating findings.

Ethical principles and gender-based violence research

Traditional bioethical approaches demand an assessment of the study on the basis of four principles: beneficence, non-maleficence, autonomy, and justice. These norms are considered guideposts for moral agents working in both health care delivery and research. The American Psychological Association’s guidelines include an additional mandate for integrity, honesty, and truthfulness in scientific research and accurate representation of facts.9 Ezekiel Emanuel and colleagues provided new guidance to clinical researchers in 2001 in their highly cited article What makes clinical research ethical? Therein, they lay out seven necessary-and-sufficient requirements: value, scientific validity, fair subject selection, favorable risk-benefit ratio, independent review, informed consent, and respect for enrolled subjects.

The need for researchers to protect research participants’ confidentiality has been well-established in research ethics literature. This obligation to protect confidentiality can come into tension with the Declaration of Helsinki’s stipulation that researchers make publicly available the results of their human participant research. Myriad papers offer researchers guidance on safeguarding data collected from human subjects throughout the research process, including extensive and, at times, repeated informed consent processes; sophisticated means of anonymizing data; and design choices that favor large sample sizes so as to protect individual identities.

Ethical complexities can be amplified in cases where research is being conducted on violence against women. This area of inquiry has been identified as one deserving a special class of protections.10 For instance, it is considered best practice to recruit women for observational research related to GBV with a “vague first contact,” in which the research is described to potential participants as a general study related to health, rather than specifically about GBV.11 Recruitment of women who have been victims of assault can feel coercive.12 Dissemination of a community-based HIV prevention project aimed at women engaged in survival sex proved equally challenging, as researchers and advocates alike feared that the dissemination of study results could further stigmatize an already stigmatized group.13

In spite of these concerns, investigators with a goal of eliminating GBV have been repeatedly called to “translate and disseminate their work to people who can use it to make change.”10 The challenges of disseminating research related to GBV has generated limited scholarly attention, most often as part of larger projects describing ethical and safety challenges inherent in conducting research on domestic violence, GBV, or other types of violence against women. Rachel Jewkes et al’s treatment of the topic in 2000 argued that given the profound safety risks involved in GBV research in Southern Africa, the only rationale for conducting such research is the potential to create on-the-ground policy and culture change.14 In 2001, The World Health Organization (WHO) published seminal ethical guidelines for conducting research with violence against women in the report Putting women first: Ethical and safety recommendations for research on domestic violence against women.15 These guidelines proposed eight recommendations: ensuring safety of participants, designing methodologically-sound research to minimize underreporting, protecting confidentiality of participants, training and supporting research team members, building in ways to reduce participant distress caused by the research, training fieldworkers to refer and provide support in settings with low infrastructure, ensuring results “are properly interpreted” and used to develop policy and programming, and including violence questions in other lines of research only if these ethical requirements are met. Similar guidelines have been reiterated in the literature in the nearly two decades since.

Ethics of research dissemination

The impetus for researchers to disseminate results has more recently gained traction in health sciences generally and global health research in particular. Various ethical underpinnings for this obligation have been proposed, though none appears to be standard. Standard bioethics analyses ground the obligation to disseminate research in a respect for persons or justice.16 Zlotnik et al. argued that these bases are insufficient to capture the unique connection between what research subjects offer and the specific benefit they are owed in return: information. In place of respect for persons or justice, they ground the obligation to disseminate in an overarching goal of researchers to be accountable, defined as “answerability for a responsibility conferred.”17 Dauda and Dierickx, for instance, grounded the obligation to disseminate results in a broader responsibility for researchers to share benefits, defined as “giving a portion of advantages or profits derived from the research to resource providers in order to achieve justice in exchange.”18 They, and other authors writing on benefit sharing, cited the Universal Declaration of Bioethics and Human Rights as a source text for the duty to benefit-share.19 In a qualitative study of the perceptions of research participants regarding dissemination of research findings, respondents confirmed previous theorists’ rationales for dissemination, reporting that they felt a sense of ownership over the findings.20 They also overwhelmingly favored receiving updates on progress throughout, rather than at the conclusion of the research only.

The Council for International Organizations of Medical Science (CIOMS), in collaboration with WHO, released 25 guidelines for international ethical guidelines for health-related research in 2016.21 Guideline 7, “Community Engagement,” specifies that researchers should engage potential “participants and communities in a meaningful participatory process that involves them early in a sustained manner” throughout the research process, including dissemination. Guideline 23, “Requirements for Establishing Research Ethics Committee and Their Review Protocols,” returns to the theme of dissemination. It states that researchers are accountable for making their results publicly available and accessible to the lay public and communities where research was conducted. Guideline 7 of the WHO Putting women first document concerns dissemination of findings and suggests that researchers have a “moral obligation” to provide feedback to communities and inform policy. Specific dissemination recommendations are to partner with local groups from the outset to inform dissemination procedures as members of an advisory committee and infuse findings into existing policy and programming. Moreover, Guideline 7 suggests that researchers ensure that findings do not stigmatize or exacerbate negative stereotypes of groups based on ethnic and social identification. In 2016, WHO developed a companion guide to the 2001 Putting women first, adding a new set of recommendations for intervention research. The dissemination guideline, underpinned by biomedical research ethics, remained mostly unchanged with the exception of recommendations that highlight the ethical mandate to consider scale-up and sustainability of effective interventions.22

Enthusiasm for dissemination of research results has appeared particularly strong within the community-based participatory research (CBPR) literature, where engagement of stakeholders and communities at all phases of research is widely endorsed. Much of CBPR dissemination literature, however, comes out of research and community partnerships developed in the US and Canada. A systematic review of dissemination practices in CBPR research (n=101) in the US found that half of the studies reported dissemination beyond publication. Most researchers (77%) disseminated findings via meetings and media coverage (~50%). Lesser-used formats were posters and flyers, phone calls, and word of mouth. This review included follow-up surveys with researchers to understand how they framed dissemination within CBPR practice. Content analysis identified five themes:

  • dissemination is valuable as a component of CBPR and in the maintenance of researcher-community relationships;
  • dissemination is affected by myriad goals of diverse stakeholders;
  • cultural differences are important variables to consider for effective dissemination, as are time and resources; and
  • dissemination is imperative for sustainability.23

Researchers in the United States have outlined the need to embrace CBPR for numerous reasons, including a potential strengthening of future research because of feedback about research processes from stakeholders.24 A CBPR study with indigenous communities in rural Alaska involved a carefully designed dissemination process wherein community-wide presentations were held. These presentations were followed by community planning group meetings, wherein community members discussed implications of the findings.25 Theoretical frameworks providing a full account of the rationales and benefits of community and stakeholder engagement are emerging, as is the evidence base for what type of research behaviors constitute effective stakeholder engagement.26

The two ethical imperatives—the need to safeguard human subjects’ confidentiality and the need to disseminate research results—can conflict. This conflict has been considered most concertedly in the genomics and bio-banking literature, where new technologies are enabling specimen re-use and re-identification in ways that were previously impossible.27 In the global health literature, the emphasis on CBPR and stakeholder engagement is prompting different concerns about how researchers can maintain their commitments to both confidentiality and dissemination. While much of the treatment of these ethical dilemmas in the genomics and biobanking literature suggests providing upfront clarity in the protocol about results notification, the protocol is often not fully defined ex ante in instances of CBPR fieldwork.

This tension to follow conflicting ethical principles and other concerns has only begun to be described in the global health literature. Two recent papers offer candid description of challenges in disseminating sensitive findings. Turcotte-Tremblay and McSween-Cadieux described challenges disseminating findings confidentially following fieldwork on improving health systems in West Africa. In particular, their concerns related to professional repercussions for interviewees whose supervisors discovered their identities. Stepping back from their own experience, the authors suggested four environmental factors that interact and can influence potential for identification of confidential research participation: hierarchical structure, small milieu, immersion in a few sites, and vested interests of decision makers.28 Baylor et al. similarly described a conscientious dissemination process wherein they gathered more than 400 Ugandan participants in a study of people living with HIV for a ceremony modeled on a local wedding for a dissemination conference.29 Members of the research team discussed confidentiality concerns with participants over the phone when inviting participants to the event. Most participants were willing to attend despite the inherent loss of confidentiality to other participants.

Our contribution to ethics of disseminating GBV research 

Despite recognition of the need for specialized ethical guidelines when conducting research with GBV, recommendations for disseminating research findings have been less developed. To our knowledge, no recommendations exist for disseminating research on GBV in conflict-affected populations. In this paper, we share our experiences of designing and disseminating research findings on these difficult topics, including the risks and ethical considerations. Rather than being prescriptive, we describe our ethical considerations with the aim of stimulating discussion towards the development of recommendations for the identification and response to ethical issues in GBV research in underserved and marginalized contexts.

Study setting

We have conducted two research projects in the Teso subregion of Northeastern Uganda. Teso borders Karamoja, where cattle raiding (violent looting of cattle and other livestock) among subtribes of the pastoralist nomadic Karamojong was highly ritualized and commonplace. When Tanzania overthrew Idi Amin’s army in 1979, many of Amin’s soldiers fled and left behind their AK-47s. One of the Karamojong subtribes acquired these weapons, which quickly escalated the violence associated with cattle raids and had other repercussions, such as inflated bride price (a man’s payment of cattle to a woman’s family). With little international notice, the Karamojong subtribes’ cattle raiding spread to Teso.30 The raiding induced large-scale displacement into camps protected by the Ugandan government army. While this was transpiring, another conflict between the Lord’s Resistance Army and the Ugandan government impacted the region when the Lord’s Resistance Army invaded Teso in 2003. As was the case in the conflict in Northern Uganda, the combatants perpetrated widespread atrocities against civilians, including abduction of children, murder, looting and destruction of property, and sexual assault.30 Many communities, especially those along the border between Teso and Karamoja, remain militarized, with government soldiers positioned there to protect against cattle rustling, despite the Ugandan government’s disarmament program that began in in Karamoja in 2006.

Study 1: Community-partnered qualitative study

The first research project was a qualitative project based of collaborations between the first author, a local nonprofit organization, and a community volunteer group. Aiming to understand how the community conceptualizes GBV and how armed conflict impacts GBV, the first author and collaborators conducted focus groups and in-depth interviews (n=77) with community members ages 9 to 80, obtaining both assent for minors and consent from caregivers and adults. This study resulted in the development of a socio-ecological conceptual model that describes four pathways between armed conflict and GBV.31 Community participants described one direct pathway, which occurred when Karamojong warriors sexually assaulted women and girls during the cattle raids. The three indirect pathways were complex, involved numerous variables, and initiated by 1) looting of property, 2) increased militarization of communities, and 3) death of civilians. The younger participants’ input was integral to constructing these pathways. For example, according to adolescent girls, increased militarization of communities put them at risk for 1) early marriage with payment of bride price, or 2) sexual assault, because soldiers have income and the community members mostly subsist off agriculture. Every focus group and individual interview highlighted problematic male alcohol misuse in relation to household IPV, noting that male alcohol misuse interacted with husbands’ perceptions that their wives were denying them something (such as food or sex) or challenging their authority in some way.

Study 2: Population-based mixed methods study

The second research project quantitatively tested the most commonly described indirect pathway in Study 1’s socioecological conceptual model: men’s poor mental health experiences as a result of looting, loss of livelihood, and other interacting factors. Because the community participants in Study 1 had identified IPV as the most ubiquitous form of GBV in their settings, the first author specified the dependent variable to focus on IPV. Study 2 involved a population-based survey across three districts in Teso to understand the scope of the problems of exposure to armed conflict, male alcohol use, and IPV. Using structural equation modeling, this study tested a pathway between exposure to armed conflict, men’s alcohol misuse and socioeconomic status, and IPV, which demonstrated an excellent fit with the population data. The association between men’s alcohol misuse and IPV was moderated by decision-making practices of couples (that is, alcohol misuse associated with violence for couples who had inequitable decision-making for women’s health care).32

Dissemination method

In August 2017, we returned to Uganda to share the findings of these two projects in a commonly used dissemination meeting format.23 As is noted in the WHO guidelines on researching GBV, partnerships are key. We convened a dissemination committee that consisted of three local, bilingual (English/Ateso) research assistants who had collected and helped analyze data in the population-based study, and a community liaison, a program manager with a local NGO who had experience with community advocacy organizing, development work, and GBV programming. Our local partners identified key community members from the three surveyed districts to participate in a half-day dissemination meeting held in a central location that included compensation for travel, attendance, lunch, and refreshments. Invited attendees represented various community sectors, such as locally elected political representatives; specialized police personnel who respond to family violence; personnel from nonprofit organizations with programming in mental health, development, GBV, and peace-building; clergymen; and community volunteers. A dissemination committee member traveled by motorbike to deliver in-person invitations to identified key community officials.

Meeting procedure

We initiated the dissemination meeting by asking attendees to complete a brief survey with questions about basic demographics, professional title and experience, hopes for learning and sharing, and permission to publish anonymous contents of our meeting. Following introductions and an opening prayer, the first author described how the researcher and local partner collaboration developed, as well as the data collection procedures of the two research studies. A representative from the dissemination committee joined each group to facilitate discussion. Following small group discussion, we reconvened as a larger group. Each group had an opportunity to share main discussion points from their small group, after which the first author shared the research findings and then facilitated discussion with the larger group on any topics where groups expressed surprise, difficulty understanding, or perceptions of problems that were incongruent with findings. Per feedback from facilitators and observation, all groups were lively, and members participated actively.

Descriptive characteristics of meeting attendees

The dissemination meeting included 21 attendees: 16 men and 5 women. The imbalanced representation of men parallels the larger sociocultural gendered trend for professional employment. The ages of respondents ranged from 24 to 60 with an average of 39.95 years. The majority of respondents (n=15) identified their ethnicity as Etesot. Other ethnicities were Gishm, Kumum, Largo, Samia, Etero, and Kimmam (n=1 each).

Ethical considerations

While attention to all ethical principles during various stages of research is crucial, we found that ethical principles became more or less prominent based on the stage of research. Ethical considerations for data collection conflicted with those for dissemination, producing tensions described below.

Beneficence/nonmaleficence and distributive justice

The driving ethical agenda for data collection was beneficence (do good) and nonmaleficence (do no harm). Maintaining beneficence meant conducting research that could not only build knowledge but also inform the development of interventions that benefit local communities, in particular women and girls who had experienced GBV in these settings.33

Practicing nonmaleficence signified conducting the research in a way that would not induce or exacerbate violence or mental distress: Ensuring the safety of participants and the research team was a primary emphasis. The research studies assessed three areas of notable ethical concern in terms of physical and psychological safety: exposure to armed conflict, exposure to IPV, and mental health symptoms and experiences (including suicidal ideation). These areas required the development of targeted safety protocols, a challenge in rural communities situated in one of the poorest countries in the world, where decades of armed conflict has impoverished communities and weakened infrastructure. For example, no mental health services exist in these rural areas and participants would have to finance and travel long distances to receive care for imminent suicidal ideation or mental health problems. Moreover, in these communities, due to patriarchal laws and sociocultural practices, women have few rights to children and land. Separation of families is highly stigmatized, and no shelters or temporary subsidized housing exist for women who experience IPV. The development of safety protocols accounting for local realities occurred in consultation with global experts, local partners, and scientific literature. Resulting safety procedures included intensive training of the research team to evaluate risk level through a series of assessment questions, and establishment of a mobile team that could respond to women at high risk of self-harm or partner violence.

To ensure safety, it was critical to maintain confidentiality of participants and the research team. The methods of Study 1 were focus groups and individual interviews. Thus, to reduce potential problems associated with confidentiality, Study 1 did not inquire about personal experiences with these problems and instead framed questions about what participants observed in their communities. Yet, some participants shared personal experiences and, in anticipation of that eventuality, the first author advised focus group members to keep the contents of the discussion confidential and reminded them of the limits to confidentiality given the format of focus groups.

Study 2, on the other hand, involved collecting data at the individual level. In an effort to do no harm and knowing that surveying both men and women could be problematic, the first ethical decision point was who to survey. Given that men might underreport IPV or become agitated when asked about it, the first author elected to survey women. When introducing the study to community officials and participants, in accordance with the WHO guidelines, the research team masked the purpose of research as a survey on women’s health. Another precaution, also informed by the WHO guidelines, was to train research assistants to pause the interview or change the subject to benign conversation if privacy was compromised during the interview process. Occasionally, practicing nonmaleficence required anticipating indirect pathways of potential harm. For instance, because respondents in Study 1 had identified male alcohol misuse as a major problem in the area and as strongly related to IPV, there was concern that male partners might take women’s compensation funds and spend it on alcohol, thereby indirectly exacerbating the problem. Thus, the local research team determined that soap, a basic necessity, would be useful compensation for women and was less likely to be taken. Finally, safety procedures to protect the research team included traveling in pairs and in-depth debriefing sessions following various forms of data collection.

When the research transitioned into the dissemination phase following data collection and analysis, the ethical principle of distributive justice gained prominence. The goal of distributive justice is to ensure that no discrete group of people disproportionately bears the risks of research, and that all have equal access to the benefits of research. Balancing this principle with beneficence and nonmaleficence proved challenging. Working to equitably distribute information learned from this research, an ethical consideration was determining the level at which to disseminate research findings (for example, population, community, and/or media). While the first author had intentionally employed mixed methods to highlight marginalized women’s voices, it was an ethical dilemma to work out how to deliver findings on GBV at a grassroots, population level, so the findings would benefit and not harm participants. One possibility, for example, was to facilitate open community events and professionals and nonprofessionals to attend. However, in consultation with the dissemination committee and local experts, this approach was deemed too risky. Given that the purpose of the survey had been masked for safety and the research team followed up the surveys with two home visits to women who had endorsed experiencing physical or sexual IPV, there was concern that participants would be identifiable and subsequently at risk for harm if the communities were apprised of the methods and results. These risks informed our decision to include community-based professionals only.

Integrity and beneficence: Dissemination for action

Community-partnered approaches such as ours aim to be collaborative, involve local capacity building, embolden partners, and balance action with research.34 These cooperative approaches are meant to benefit local communities through action, information dissemination, and sharing power.34 In the data collection phase, the impetus for beneficent research was to design a rigorous study that contributes to the literature base. In the dissemination phase, an ethical issue became how to disseminate findings transparently and with integrity about difficult issues. This dissemination needed to follow the ethical principle of beneficence by facilitating change, in a way that would not shut down community partners.

Community-partnered approaches stress the importance of operating from a strengths-based perspective. While Studies 1 and 2 asked about community strengths and coping, the main research questions evaluated problems of armed conflict, IPV, and mental health outcomes. Findings revealed high rates of IPV, exposure to armed conflict, male alcohol misuse, and mental health problems. These findings showed that IPV was mostly perpetrated by male partners in the household. In a highly patriarchal setting, it was necessary to consider carefully the ramifications of identifying men as perpetrators. Study 1, for instance, revealed a common perception among men that women were to blame for driving violence in the home. Furthermore, research participants identified unjust legal and protection practices that favored men (most dissemination meeting attendees were male) and permeated formal institutions (some of which were represented in the dissemination meeting). Key informants described community resistance as one of their main challenges in raising awareness about IPV. The ethical dilemma became how to share these problems that were deeply entrenched in larger sociopolitical conditions.

Understanding that these issues had the potential to evoke resistance, it was of central importance to select a framework within which to build a message about findings. Disseminating findings beneficently meant avoiding pathologizing communities and assigning meritocratic responsibility for problems identified in the research. Thus, to conceptualize the dissemination process, we leveraged the framework of historical trauma to contextualize the findings in historical and larger sociopolitical conditions. The concept of historical trauma, which grounds current collective distress in past suffering, emerged after immersed clinical work with American Indian populations in the US.35 We used the comprehensive socioecological model to conceptualize problems evaluated in Studies 1 and 2. Thus, we could follow the framework of historical trauma by demonstrating connections between community exposure to armed conflict inflicted by outsiders to the community and ongoing problems faced in communities and homes. We concluded our dissemination meeting by eliciting discussion about community strengths that could prevent and respond to these problems.

In addition to operating from a contextualized framework, we also designed the methods of dissemination with beneficence through action in mind. Effective collaborative work involves bidirectional dialogue about research findings.36 Despite the importance of dialogue among researchers and community partners that is equitable, scarce literature exists describing best practices for engagement of community partners, especially those communities who have few resources and have experienced injustices through colonization and other forms of armed conflict.

Guided by the ethical imperative to promote bidirectional dialogue for change, instead of making a lecture-style passive format, we used a variety of active learning strategies to prompt discussion in small group format. We selected core quantitative and qualitative findings for use in quiz-style questions. Small groups worked through items intentioned to elicit dialogue and engage them in diverse mental exercises. For example, to facilitate discussion on the scope of these problems, the dissemination committee asked multiple choice questions about prevalence rates. One question asked: “How many men in our study drank alcohol daily? a. 60%; b. 30%; c. 5%.” The small groups deliberated and came to consensus about their answers. Another type of question was designed to stimulate discussion about the deeper meanings behind sensitive results. For instance, “A young boy said this: ‘When dad goes to drink, it tends to poison him.’ What do you think this boy was talking about?”

These participatory methods allowed for several insights about the research findings, all of which were informative for the development of an applied intervention. For example, regarding prevalence of male alcohol misuse, all groups thought there should have been a response item that was above 60% (the highest rate listed) and closer to 80%. This discrepancy between attendees’ perceptions and our findings prompted discussion about whether we had accurately assessed male alcohol use and the understanding that it was likely even more problematic than found in the research. We also learned valuable information about social norms of alcohol use. Attendees described rituals where children as young as two years old were introduced to alcohol. They also highlighted the important role that alcohol played in maintaining male social networks. Therefore, reducing alcohol consumption among men could isolate them from their peers, while research is clear that social support and relationships are important for optimal mental health.37 Moreover, women brew and sell most alcohol. Many male partners have been killed in conflict and few other options exist for women to sustain their families. Thus, a community-based intervention that effectively decreased alcohol consumption could deplete the already tenuous incomes of women who sell alcohol.

Globally, mental health problems are stigmatized. Deciding again how to disseminate these findings without pathologizing women or explicitly implicating men was another ethical decision-making point. The community participants (Study 1) and women (Study 2) had described numerous deleterious mental health outcomes as a result of IPV. To illustrate relations among these mental health problems and IPV, analysts in the US designed a conceptual figure of boxes and directional arrows. We asked dissemination meeting attendees to engage in a similar mental exercise. This resulted in ways of modeling relations among these problems from local perspectives. One small group, for instance, conceptualized women’s mental health experiences relating to IPV from a more organic perspective than US analysts had depicted. They articulated that female survivors of IPV and their mental health experiences might be conceptualized as a tree. The violence women experience signifies the root system that sustains the tree, and the leaves and foliage represent symptoms that cyclically interact by feeding back into the soil. This locally grounded perspective has informed how the framing of future interventions can illustrate IPV’s deleterious mental health outcomes in a contextualized way.

Recommendations

Several recommendations follow from wrestling with the ethical considerations involved in disseminating findings on armed conflict, GBV, and mental health in a low-income country. First, planning to disseminate findings beyond publication for academic audiences should occur a priori to the initiation of research. Initial plans should include funding for dissemination in the research budget and building dissemination into grant proposal timelines. While the role of funders in providing incentives or requirements to disseminate findings beyond publication has yet to be explored, procuring funding for dissemination meetings and events is especially important when working with low-resourced, marginalized communities that have endured decades of armed conflict.38 In addition to researcher travel, costs to consider are payment for dissemination committee members’ time and work, meeting venue, refreshments and meals, and travel costs for attendees, among others.

Second, engaging in meaningful research-community partnerships at the local level is key. National populations are heterogenous with differences between urban and rural settings. In Uganda, for instance, numerous ethnic groups and languages exist.39 To account for these variabilities, consultation and collaboration should include liaisons at the most local level possible. Initially overlooked for inclusion in the dissemination meeting, we advocated for inviting community volunteers who were unaffiliated with formal institutions and had less training and education. Moreover, it would be helpful to first determine with local partners the level at which data will be disseminated. If the first author had anticipated wanting to disseminate findings at the participant level a priori, for instance, she could have included open-ended questions for participants about their ideas for dissemination, thereby broadening the participatory scope of the study.

As a component of the research plan, evidence and best practices should inform dissemination methods with an ethical goal of doing good through action. Evidence-based practices in pedagogy, such as methods of teaching with active learning strategies, should guide design of dissemination activities. Models of behavior change, such as the transtheoretical model and associated techniques, have been used successfully in other training and implementation activities.40 Methods of designing participatory logic models and theory of change should be considered. The development of such models can start with identifying desired outcomes and move backward to determine what is needed to achieve those outcomes. Finally, the use of contextualized frameworks, such as historical trauma, should guide researcher-community partners in communicating findings so as to facilitate change and reduce resistance to the message.

Disseminating GBV research and human rights

While we did not resolve ethical dilemmas inherent in these research studies by adhering to a human rights framework, we recognize retrospectively that we had, in fact, made ethical decisions that favored a human rights perspective. Research-oriented ethical guidelines, such as the Declaration of Helsinki or the Council for International Organizations of Medical Sciences (CIOMS), have generally been remiss in discussing practical considerations for dissemination of GBV research findings. However, we have discovered that the aims of human rights documents, such as the Universal Declaration of Bioethics and Human Rights, provide guidance that was missing in bioethics literature. Article 4 of the Universal Declaration of Bioethics and Human Rights (optimize benefit and minimize harm) parallels beneficence and non-maleficence mandated by ethical guidelines. Other articles, such as Article 3, which focus on human dignity and human rights, provide more applicable guidance by highlighting the importance of prioritizing individual rights and freedoms over scientific benefits to society at large. We recognize the principle of dignity because one of our challenges was protecting individual rights and freedoms of participants while facilitating equitable access to scientific findings through dissemination. Article 8 on respect for human vulnerability and personal integrity is likewise relevant, as it stresses the importance of respecting and protecting vulnerable individuals and groups and their personal integrity. In accordance with WHO’s Putting women first guidelines that emphasize women’s safety, our resolution of ethical tensions between research design and dissemination prioritized this latter mandate above all. 

Conclusions

Our experience researching and disseminating research findings related to GBV and mental health in conflict-affected communities revealed that the prominence of various ethical principles shifted dynamically. Depending on the stage of research, some principles became more central than others, requiring accordingly responsive procedures and protocols. The fluctuating nature and significance of ethical principles produced identifiable tensions when it came time to disseminate findings. One particular tension arose between collecting data using safe and unharmful methods and distributing research findings equitably. Another ethical issue concerned transparently sharing findings of problems in a hopeful and contextualized way that would facilitate community action to address problems. Recommended dissemination strategies are to plan for dissemination a priori, engage with partners at local levels, and ground dissemination for action in evidence-based practices.

Jennifer J. Mootz is a research scientist at Columbia University Department of Psychiatry in New York, NY, USA.

Lauren A. Taylor is a doctoral candidate at Harvard Business School in Boston, MA, USA.

Milton L. Wainberg is professor of clinical psychiatry at Columbia University Department of Psychiatry in New York, NY, USA.

Kaveh Khoshnood is associate professor of public health at Yale University in New Haven, CT, USA.

Please address correspondence to the authors c/o Jennifer Mootz. Email: Jennifer.Mootz@nyspi.columbia.edu.

Competing interests: None declared.

Copyright © 2019 Mootz, Taylor, Wainberg, and Khoshnood. This is an open access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original author and source are credited.

References

  1. C. Garcia-Moreno, H. Jansen, M. Ellsberg, et al., “Prevalence of intimate partner violence: findings from the WHO multi-country study on women’s health and domestic violence,” The Lancet 368 (2006), pp. 1260–1269.
  2. United Nations. Declaration on the Elimination of Violence against Women (1993).
  3. M. Ellsberg, H. A. F. M. Jansen, L. Heise, et al., “Intimate partner violence and women’s physical and mental health in the WHO multi-country study on women’s health and domestic violence: an observational study,” The Lancet 371/9619 (2008), pp. 1165–1172.
  4. Human Security Research Group, Human Security Report 2013: The Decline in Global Violence: Evidence, Explanation, and Contestation (2014).
  5. United Nations Population Fund, Gender-based violence in humanitarian settings, 2014. Available at https://www.unfpa.org/resources/gender-based-violence-humanitarian-settings.
  6. United Nations, Conflict-related sexual violence (2017).
  7. World Health Organization, Violence against women (2016).
  8. L. Stark and A. Ager, “A systematic review of prevalence studies of gender-based violence in complex emergencies,” Trauma Violence Abuse 12/3 (2011), pp. 127–134.
  9. American Psychological Association, Ethical Principles of Psychologists and Code of Conduct, 2018. Available at https://www.apa.org/ethics/code/.
  10. L. A. Fontes, “Ethics in violence against women research: the sensitive, the dangerous, and the overlooked,” Ethics & Behavior 14/2 (2004), pp. 141–174.
  11. A. K. Blanchard, C. A. T. M. Sangha, S. G. Nair, et al., “Pursuing authenticity from process to outcome in a community-based participatory research study of intimate partner violence and HIV Vulnerability in North Karnataka, India,” Qualitative Health Research 27/2 (2017), pp. 204–214.
  12. A. Aldous, M. Magnus, A. Roberts, et al., “Challenges in conducting research on sexual violence and HIV and approaches to overcome them,” American Journal of Reproductive Immunology (New York, N.Y.: 1989) 78/1 (2017).
  13. K. Shannon, V. Bright, S. Allinott, et al., “Community-based HIV prevention research among substance-using women in survival sex work: the Maka Project Partnership,” Harm Reduction Journal 4 (2007), p. 20.
  14. R. Jewkes, C. Watts, N. Abrahams, et al., “Ethical and methodological issues in conducting research on gender-based violence in Southern Africa,” Reproductive Health Matters 8 (2000), pp. 93–103.
  15. World Health Organization, Putting women first: Ethical and safety recommendations for research on domestic violence against women (2001) .
  16. A. Norris, A. Jackson, and K. Khoshnood, “Exploring the ethics of observational research: The case of an HIV study in Tanzania,” AJOB Primary Research 3/4 (2012), pp. 30–39.
  17. R. Zlotnik Shaul, L. Reid, B. Essue, et al., “Dissemination to research subjects: operationalizing investigator accountability,” Accountability in Research 12/1 (2005), pp. 1–16.
  18. B. Dauda and K. Dierickx, “Benefit sharing: an exploration on the contextual discourse of a changing concept,” BMC Medical Ethics 14 (2013), p. 36.
  19. J. H. Solbakk, “In the ruins of Babel: pitfalls on the way toward a universal language for research ethics and benefit sharing,” Cambridge Quarterly of Healthcare Ethics: CQ: The International Journal of Healthcare Ethics Committees 20/3 (2011), pp. 341–355.
  20. R. S. Purvis, T. H. Abraham, C. R. Long, et al., “Qualitative study of participants’ perceptions and preferences regarding research dissemination,” AJOB Empirical Bioethics 8/2 (2017), pp. 69–74.
  21. Council for International Organizations of Medical Sciences (CIOMS, International Ethical Guidelines for Biomedical Research Involving Human Subjects.
  22. World Health Organization, “Ethical and safety recommendations for intervention research on violence against women,” Putting Women First: Ethical and Safety Recommendations for Research on Domestic Violence against Women (2016), pp. 1–40.
  23. P. G. Chen, N. Diaz, G. Lucas, et al., “Dissemination of results in community-based participatory research,” Am J Prev Med 39/4 (2010), pp. 372–378.
  24. C. R. Horowitz, M. Robinson, and S. Seifer, “Community-based participatory research from the margin to the mainstream: are researchers prepared?,” Circulation 119/19 (2009), pp. 2633–2642.
  25. I. Rivkin, J. Trimble, E. D. S. Lopez, et al., “Disseminating research in rural Yup’ik communities: challenges and ethical considerations in moving from discovery to intervention development,” International Journal of Circumpolar Health 72 (2013).
  26. M. S. Goodman, S. Thompson, and V. L, “The science of stakeholder engagement in research: classification, implementation, and evaluation,” Translational Behavioral Medicine 7/3 (2017), pp. 486–491.
  27. J. Kaye, “The tension between data sharing and the protection of privacy in genomics research,” Annual Review of Genomics and Human Genetics 13 (2012), pp. 415–431.
  28. A.-M. Turcotte-Tremblay, and E. Mc Sween-Cadieux, “A reflection on the challenge of protecting confidentiality of participants while disseminating research results locally,” BMC Medical Ethics 19/Suppl 1 (2018), p. 45.
  29. A. Baylor, C. Muzoora, M. Bwana, et al., “Dissemination of research findings to research participants living with HIV in rural Uganda: challenges and rewards,” PLoS Medicine 10/3 (2013), p. e1001397.
  30. Internal Displacement Monitoring Center & Norwegian Refugee Council, Uganda: Difficulties continue for returnees and remaining IDPs as development phase begins, 2010. Available at http://www.internal-displacement.org/countries/uganda.
  31. J. J. Mootz, S. D. Stabb, and D. Mollen, “Gender-based violence and armed conflict: A community-informed socioecological conceptual model from Northeastern Uganda,” Psychology of Women Quarterly 41/3 (2017).
  32. J. J. Mootz, F. K. Muhanguzi, P. Panko, et al., “Armed conflict , alcohol misuse , decision-making , and intimate partner violence among women in Northeastern Uganda : a population level study,” Conflict and Health 12 (2018), pp. 1–11.
  33. WHO – World Health Organization, “Putting Women First: Ethical and Safety Recommendations for Research on Domestic Violence Against Women,” World Health Organization (2001), p. 33.
  34. B. A. Israel, A. Schulz, E. Parker, et al., “Critical issues in developing and following CBPR principles,” in M. Minkler and N. Wallerstein (eds), Community-based particip. res. heal. from process to outcomes (Jossey-Bass, 2008), pp. 47–66.
  35. M. Y. H. Brave Heart, “The historical trauma response among natives and its relationship with substance abuse: A Lakota illustration,” Journal of Psychoactive Drugs 35/1 (2003), pp. 7–13.
  36. B. McDavitt, L. M. Bogart, M. G. Mutchler, et al., “Dissemination as dialogue: building trust and sharing research findings through community engagement,” Preventing Chronic Disease 13 (2016), p. 150473.
  37. M. Weissman, J. Markowitz, and G. Klerman, The guide to interpersonal psychotherapy: Updated and expanded version (Oxford University Press, 2018).
  38. S. D. Ritchie, M. J. Wabano, J. Beardy, et al., “Community-based participatory research with Indigenous communities: The proximity paradox,” (2013).
  39. Central Intelligence Agency, The World Factbook: Africa, 2018.
  40. J. Strang, J. McCambridge, S. Platts, et al., “Engaging the reluctant GP in care of the opiate misuser Pilot study of change-orientated reflective listening (CORL),” Family Practice 21/2 (2004).