Several factors may increase the vulnerability of individuals and communities participating in research, including, lower socioeconomic status, less experience with and understanding of biomedical research, and poor availability and accessibility of health services [12]. Because these vulnerabilities may be particularly prominent in resource-limited communities, additional mechanisms for ensuring adequate protection for research participants should be considered [4]. The need to apply and uphold standard ethical frameworks across significantly different communities and within vulnerable populations has led to recommendations stressing the role for community engagement as part of the ethical research process [13,14]. Involving communities in aspects of research activities like protocol development and research conduct can lead to greater protections for communities and individual participants [15,16], but little empirical data inform the best practices and models of community engagement for research in sub-Saharan Africa [17,18].

Informed consent is a fundamental ethical principle of biomedical research to protect vulnerable populations and minimize risks [19]; however, the optimal implementation of informed consent requires consideration of multiple factors. The type of scientific trial, particularly whether randomization is done and whether it is done in clusters or by individual, may shape which consent processes best protect vulnerable subjects [20]. Moreover, in certain settings, individuals may be less likely to understand the research being conducted or their rights as participants [21,22], leading to concerns about the adequacy of standard informed consent practices. Illiteracy, lack of formal education and differing views of health and research complicate informed consent in all settings and may be particularly significant in resource-limited settings [21,23]. Both decisions favoring participation and refusals to participate in research may result from a lack of information or appropriate explanations, including perceptions of access to benefits such as healthcare [24]. When a community has limited access to health services, research may be used as or considered a means to access healthcare [25,26]. While conflating research and healthcare is often referred to as the “therapeutic misconception”, this may not be a misconception in settings where the research is the only access point to services [27,28]. Finally, how the participants conceptualize individual autonomy and whether community or social autonomy supersedes the individual can have important implications for informed consent processes [29,30]. The balance between individual autonomy and social or political choices may significantly influence the implementation of research [31]. A qualitative study of researchers in developing countries emphasized the need to allow more operational flexibility in the informed consent process to account for varying community contexts [32]. Furthermore, a recent review of research on standard informed consent suggests greater community involvement may be needed to better protect participants in all settings, particularly resource-limited settings [33].

International research bodies recognize the need to increase community engagement in research activities [34,35]. How a given community understands research and how to best engage communities in the process of consent are still open questions in many international research settings [36]. A literature review related to informed consent found an absence of data from the developing world, and particularly from sub-Saharan Africa, on how communities’ perceptions of informed consent related to the experiences of participation in the research [11]. Moreover, the limited empirical research on the informed consent process in resource-limited settings suggests community perspectives require further investigation within multinational research settings [21,25,37-40]. A needs assessment of research ethics issues among a partnership in western Kenya found adapting informed consent processes to involve community consent was considered a major challenge [41].

As multinational research collaborations seek to answer these questions about how best to conduct informed consent and to increase community engagement, existing community structures for dialogue and decision-making may offer important venues for consideration. In many Kenyan cultures, traditional community assemblies called mabaraza are used for sharing information and for gathering community opinions on issues [42]. These mabaraza have the potential to serve a role in research investigation and development, as well as in the consent process. For the purposes of understanding community perspectives or beliefs, a baraza (singular form of mabaraza) can be used as a venue for qualitative investigation, and because the baraza holds cultural significance, it may produce more meaningful community involvement than researcher-organized boards or focus groups. A recent study investigated the use of mabaraza to understand community perspectives on the USAID-Academic Model Providing Access to Healthcare (AMPATH) HIV care program in western Kenya and found the forums were useful in eliciting community perspectives and understandings of health programs [42]. A qualitative case study in northern Ghana investigated the use of durbars, traditional, formal community gatherings that include cultural activities, as a form of community engagement with a biomedical research project [43]. They found that engaging the community within this existing cultural forum was effective in gaining community perspectives on various aspects of research and limited social disruption.

Employing the baraza -- a culturally accepted public assembly for the purpose of community dialogue [42] -- we gathered novel data on community members’ understanding of research, informed consent, and community consent. We built on previous qualitative work in Kenya that used focus group discussions with community leaders and members to better understand their perceptions of research and informed consent [17,26], and we tried to fill the continued gaps in understanding how a community’s perspective on informed consent relates to participation in the research [11]. While the community’s understanding of research generally encompassed the principles of inquiry and investigation, the strong expectation to receive direct benefits for research participation was an important belief to consider and understand. The community members often focused on the goods, services, capacity building or development that might be offered as part of the research activities. This is consistent with other findings from poor communities, where research activities were equated with development projects whose aim is to improve conditions in the community, rather than to collect information or test interventions [25]. While some research studies offer access to health care or other resources of value to the participants, others might not benefit one’s health, education, or finances. In our work, the participants expressed that the potential benefits of the research would be scrutinized – rather than the protocol. These expectations for direct benefits should inform how researchers present research proposals to the community and how they protect vulnerable individuals within that community. This community view could shift the risk-benefit ratio such that research benefiting the community would be endorsed at the community-level despite individual risks to the participants. These findings also point to the need to be explicit about possible risks. In this context, studies with few direct benefits may be unlikely to recruit participants from similarly underserved or impoverished settings, while studies with obvious benefits might draw in participants despite having significant risks. Considering this interpretation of the risk/benefit ratio for this particular setting and striving to make clear the risks that the participants might not otherwise consider would be important targets for researchers within this cultural context. A multinational research team might be seen as even more likely to offer direct benefits for research participants, reinforcing the need for researchers to understand the unique community perspectives in order to adequately address community concerns and provide protections for vulnerable populations. Understanding how the history, poverty, and unique resources of a given community shape the community’s attitude towards research is vital for researchers entering that community.

Another important cultural perception revealed in the mabaraza was the conflation of HIV and research. Given that the most sophisticated healthcare system in this community in western Kenya was designed for HIV care and that much of the research which has taken place has been related to HIV, it is conceivable why participants would assume biomedical research involved HIV testing or HIV-related outcomes. AMPATH has conducted research in this setting since its inception in 2001, and AMPATH’s services within this county have included home-based HIV counseling and testing, and multiple locations provide voluntary counseling and testing for HIV. Conflating HIV and research may thus be a unique finding to a setting with significant HIV services. However, since HIV infection continues to be significantly stigmatized in this society, the equation of research with HIV means that HIV-related concerns for privacy and confidentiality are prominent in the minds of community members. In fact, “risks” associated with research were largely interpreted to mean risks related to diagnosis and disclosure of HIV status, rather than any other harm. Studies with no HIV-related components might want to take particular care to make this clear to potential participants. Issues of confidentiality, stigma, and HIV disclosure need to be addressed carefully within any research project in this setting.

Community consultation commonly refers to engaging communities and eliciting feedback from community members that researchers can incorporate into the design of their research work to enhance protection and benefits for the community, increase legitimacy and share responsibility of activities and outcomes [48]. Community advisory boards have been utilized for community consultation in HIV research [49,50], but there is little empirical data on the effectiveness of these boards in the consultation process [51]. Communities in western Kenya have worldviews that emphasize communitarian rather than individualistic living, a philosophy that may render community consultation all the more necessary [52]. The use of a traditional community assembly such as the baraza offers a potential venue for community consultation, and for additional research activities such as community consent. Involving the community in review of research proposals in assemblies such as mabaraza was widely perceived as culturally acceptable and an important protective measure. Community consent is a process of seeking permission from the community to solicit individuals to participate in research [48]. Our findings from Kenya suggest that this is a setting in which community consent may be appropriate and necessary, even before individual consent, in order to extend protections from the individual to the community-level [41,53]. Using community consent as a supplement to individual consent would be in accordance with the 2002 Council for International Organizations of Medical Sciences International Ethical Guidelines for Biomedical Research Involving Human Subjects, which endorses “obtaining permission from a community leader, a council of elders or another designated authority.” [35] Our findings begin to answer key questions needed to implement this guideline: 1) defining the community, 2) describing the content and purpose of community consent, and 3) identifying the legitimately empowered local leaders who could make decisions about research on the community’s behalf [54,55]. As there have been very few studies investigating community consent processes in sub-Saharan Africa [18,38,40] and no guidelines as to its implementation, these formative data offer an important description of how one community would answer these questions. The limited body of research investigating effective strategies for community engagement in consultation and consent processes warrants further investigation [36].

Community consent might offer a particular benefit to especially vulnerable populations, such as orphaned children, where the community’s consideration of the research could offer an additional level of protection against potential harms or the opportunity to advocate for including these children when benefits might be present. The potential advantages of community consent need to be examined carefully against potential harms that could result if the community were to value community benefits over the protection of an individual; however, the usual measures for protecting against abuse, from emphasizing researcher integrity to requiring local IRB review, could help minimize those risks. Not only must researchers ensure that the procedures, risks, and benefits to their proposed work are well understood, but they must take precautions to minimize risks for vulnerable individuals.

Dissenting views also merit further consideration when considering community consultation or consent. Balancing individual autonomy and the choices of the broader social context must be carefully weighed in implementing ethical research [31]. Child dissent was not discussed, nor was the possibility that the voices of some members of the community might not be heard in the context of community deliberations or assemblies. Within this setting in Kenya, women, younger community members, and families of lower socio-economic status in the village may not feel able to speak freely in group deliberations, and thus their opinions may go unheard in community discussions. To try to counter this tendency in our assemblies, the facilitator of the mabaraza specifically asked for responses from women within the groups and periodically asked participants who had been quiet to offer their opinions. This increased participation among groups that might face gender or social inequalities, and did result in hearing some diverse opinions. However, their concerns may not have been fully expressed, and addressing these inequalities within the context of a community assembly requires direct attention or may require complementing the community mabaraza with smaller, more homogeneous group discussions (men only, women only, youth only, etc.). A family or individual’s decision to withdraw from a study for which community consent has been given is another important form of dissent to address from the beginning. The participants in these discussions did not talk about that issue. The weight given to community consent compared to individual consent or dissent may also be altered by the design of the study, taking into account factors such as cluster or individual randomization and whether individual consent is possible. Additional investigation of how to address, protect, and respond to minority views within this cultural context would be helpful.

This study does have limitations that merit consideration. The model relies on the contextual data and lived experiences of subjects in a very particular part of the world – two locations in western Kenya. Thus, the results may not generalize to other geographic locations. The mabaraza took place in peri-urban areas that would be classified as slums, which may make the findings less generalizable to more rural or urban settings. Nonetheless, this is a population often under-represented in research, and the population caring for many of the orphans and vulnerable children we sought to evaluate. Furthermore, the methodology has the inherent limitations of using relatively small, convenience samples which can limit generalizability. However, the lived experiences of subjects in this particular resource-limited setting may have more generalizability for subjects in other resource-limited settings than do similar studies conducted in resource-rich settings. Despite these limits to the sample, validity was supported by thematic saturation. In addition, there have been no published studies specifically using mabaraza, or similar traditional assemblies, to gain community perspectives on informed consent processes in biomedical research settings, and so the data from this population offer an important window into ethical cross-cultural research practice. A baraza gathering might not lend itself to hearing from every participant or to collecting in-depth responses from any one participant in the manner of an individual interview or a smaller focus group. Nonetheless, it does provide a community overview and broader perspective on the community’s beliefs. Moreover, in this context, it made use of a structure being proposed to serve a similar purpose in future research endeavors and thus testing its own possibilities.

Eunice Kamaara is a professor in the Department of Philosophy and Religious Studies at Moi University in Eldoret, Kenya. She is also a student in the Masters in International Health Research Ethics program. Her research interests focus on informed consent in post-colonial contexts.

Allan Kamanda is the study coordinator for the Orphaned and Separated Children’s Assessments Related to their Health and Well-being (OSCAR) research study in western Kenya, which is following the outcomes of a large cohort of several thousand orphaned and separated children. His research interests focus on the consenting process among vulnerable populations such as orphaned children and street children.

David Ayuku is an associate professor in the Department of Behavioural Sciences, School of Medicine, Moi University College of Health Sciences in Eldoret, Kenya and Co-Principal Investigator for the Orphaned and Separated Children’s Assessments Related to their Health and Well-being (OSCAR) research study in western Kenya. His research interests focus on the issues of street children and orphans.

Winstone Nyandiko is an Associate Professor and the Head of the Department of Child Health and Paediatrics, School of Medicine, Moi University College of Health Sciences in Eldoret, Kenya and Associate Program Manager for the AMPATH partnership, serving as the Co-Director for the AMPATH Research Network. Dr. Nyandiko is also the Pediatrician-In-Charge for the AMPATH Pediatric HIV Care program. His expertise lies in evaluating the clinical management of HIV-exposed and HIV-infected children and implementing research within the Kenyan health care system.

Lukoye Atwoli is a Lecturer in the Department of Mental Health in the School of Medicine, Moi University College of Health Sciences in Eldoret, Kenya and secretary of the Kenya Psychiatric Association. His research interests are in psychotrauma in children and adults, general hospital psychiatry and the interface between HIV/AIDS and mental health.

Samuel Ayaya is a Professor in the Department of Child Health and Paediatrics, School of Medicine, Moi University College of Health Sciences in Eldoret, Kenya and Co-Chair of the Kenya Pediatric Research Working Group. His research interests include children’s nutrition, growth, and development, the progression of HIV in children, and child abuse and neglect.

Peter Gisore is a Lecturer in the Department of Child Health and Paediatrics, School of Medicine, Moi University College of Health Sciences in Eldoret, Kenya. His interests center around clinic and community-based research on maternal and newborn health.

Michael Scanlon is a Global Pediatric Clinical Research Scholar with Indiana University School of Medicine and the AMPATH partnership. He is focused on research to support long-term care of children with HIV in resource-limited settings, including disclosure of HIV status and adherence to HIV therapy, as well as community-based participatory research.