Harmonizing voices: A collaborative, cross-cultural approach for adolescent sexual and reproductive health research Open Access

International research studies often leverage the knowledge and lived experience of in-country investigators to inform study design and implementation. Local research staff who implement the study, such as study managers, research assistants (RAs), and data managers, usually are not engaged to participate in data collection tool development and data analysis. This exclusion is a missed opportunity to benefit from their knowledge of the local culture and create sustainable community practices while concurrently studying these practices [1, 2, 3, 4].

Maneno Yetu (“Our Words”) is a study of sexual and reproductive health behaviors among adolescents residing in informal settlements of Kisumu, Kenya. Informal settlements, that is, slums, are characterized by impoverished living conditions which contribute to poor health and high rates of morbidity and mortality [5, 6, 7, 8]. Our mixed-methods study included focus group discussions (FGDs) and a cross-sectional survey [9, 10, 11, 12, 13, 14].

Recognizing the value of integrating our Kenya-based research team in all aspects of the study, we implemented a novel collaborative, cross-cultural approach that leveraged our team members’ knowledge of the culture and vernacular of the study communities. In this article, we describe the process by which our U.S. and Kenyan research team jointly collaborated in developing the interview guide and analyzing the FGD data. We also share the team members’ reflective analysis of how their experience with learning about qualitative tool development and data analysis enriched their professional development.

The five members of the U.S.-based team included the principal investigator (PI), qualitative researchers, and study managers. The seven members of the Kenya-based team included a study manager, data manager, and RAs with a range of experience and expertise. One Kenyan member previously worked on 7 research studies; three members worked on 3–5 studies; two worked on 1–2 studies; and one was working on their first study.

Several team members were young adults who related well with participants. Others spent their adolescent years living in or raised or were currently raising adolescent children in the study communities and contributed first-hand perspectives on the experiences of adolescents residing in informal settlements [7, 8]. The community connections of the Kenyan team were essential for sensitization and mobilization efforts.

Our research team members engaged in the FGD interview guide development and data analysis. Our team initially included a Kenyan co-investigator who was co-leading the qualitative component, but then subsequently withdrew due to time constraints. Given the timeline pressure to implement the study, the Kenyan team wanted to continue developing the FGD interview guide with the support of the U.S.-based co-investigator co-leading the qualitative component.

The interview guide’s initial domains of inquiry were based on the thematic areas described in the funded grant application, which was co-written by U.S. and Kenyan investigators. We held a workshop in Kisumu to design probes and vignettes to facilitate discussions about knowledge, attitudes, behaviors, and beliefs around relationships and sex. The Kenyan team implemented the FGDs and took the lead during the data analysis phase, with mentoring by the U.S. members.

Through team meetings and impromptu conversations, our members frequently shared their perspectives with each other regarding their experiences with interview guide development and training and mentoring on qualitative data analysis. After the first two papers from the qualitative study component were published, our team members expressed an interest in developing a manuscript that described our experiences with this collaborative approach [9, 10]. To facilitate manuscript development, each team member was asked to write down perspectives that they had previously shared or new thoughts they had regarding the interview guide development and data analysis training and mentoring processes. The PI assigned each team member a unique code that included location information, for example, KE1 = Kenya-based team member 1, US2 = U.S.-based team member 2; organized the team members’ perspectives based on whether they were related to the interview guide development process or the qualitative data analysis process; and selected quotes that were representative of the breadth of perspectives.

This was the first time that most of the Kenyan team members had participated in tool development for qualitative research.

They credited their participation for gaining a better grasp of the scientific intent behind the FGDs.

The team believed it was critical that the guide be developmentally appropriate for adolescent participants.

The Kenyan team drew from their own adolescence growing up in similar environments to develop a guide that resonated with our study population.

Our Kenyan members ensured the local context was well-represented.

Discussions during the workshop and post-workshop through emails and conference calls helped craft probes that prompt a variety of perspectives to ensure all voices would be heard.

Jointly developing the FGD guide helped team members gain a better understanding of the intent behind the questions and probes.

The Kenyan team members felt that the practice sessions increased fluency in administering the interview guide and enabled sharing on how to best approach the questions to obtain rich data.

Our team viewed working together on the guide as a learning opportunity.

The process also created cohesiveness for the newly assembled U.S. and Kenyan team.

For the qualitative data analysis process, U.S. members provided training and mentoring for the Kenyan team through 2 week-long intensive in-person workshops in Kisumu over a 6-month period and five 1-hour online workshops over a 2-month period, followed by ongoing mentoring through emails and online sessions. This stepwise approach afforded time to “contextualize” qualitative research and “digest the material.” We reviewed transcripts as a team using an open coding process to develop initial codes that were close to the data. We wrote potential codes on sticky notes and categorized the notes on a board along the lines of parent and child codes. This process facilitated the transition to using qualitative data analysis software (Dedoose). Team members strengthened their skills by working in pairs to conduct double-blind coding, presented back to the group to review areas of convergence and divergence, finalized the codebook, created analytic tables from code reports, and drafted memos.

The team viewed the trainings as a “great initiative” that afforded them the opportunity to learn from their U.S. and Kenyan colleagues and from each other.

Trainings helped them “understand core concepts and processes involved in qualitative data analysis” and working in groups “helped with the exchange of ideas.”

Analyzing the data was a novel experience and a chance to expand upon their skill set.

The team would have liked to have additional in-person workshops to allot more time to cover topics and foster a “deeper understanding and synthesizing of the concepts learnt.” Workshops enabled more interpersonal interactions than online sessions that were conducted every other week and not always sufficient to address all questions and concerns in a timely manner.

Other challenges with the online sessions included poor network connections at times, which made communication less effective, and time zone differences.

Applying their newly acquired data analysis skills, the Kenyan team took the lead with coding, creating analytical tables and developing memos, with ongoing review and mentorship by the U.S. members. The full team collaborated on finalizing the analyses.

Transcripts were double-blind coded and then discussed among team members.

U.S. members provided constructive feedback, which the local team found helpful.

Trainers were impressed by how rapidly the team developed their analytic skills.

The entire team considered the collaborative interview guide development and data analysis processes a success.

The collaborative process created a strong bonding experience, transforming co-workers into a close family. This opinion was expressed by both Kenya- and U.S.-based team members.

Our collaborative, cross-cultural approach created a team structure and research environment that promoted egalitarianism, representativeness, and collectiveness within the rigor of a National Institutes of Health study. Our team members considered each other as peers, irrespective of education degrees or position titles. The teamwork fostered a supportive working environment, created family-like dynamics, empowered equal partnerships, and amplified investment in the success of the study.

Our approach aligns with promoting respectful and collaborative interactions principles of decolonization of global health research practices [15]. Engaging both U.S.- and Kenya-based members in all study phases and integrating capacity building helped strengthen our team’s qualitative research skills. The team approach strengthened the validity of findings and provided professional development opportunities.

The Kenyan team has built upon the knowledge and confidence gained from the qualitative data analysis experience to expand their skill set further. Two members were selected to participate in a week-long writing workshop led by faculty and scientists from the Kenya Medical Research Institute, Maseno University, and UCSF. Several members are lead authors on manuscripts in development that stem from the FGD data, with mentoring from the PI to hone their writing skills. The experience also inspired several members of our team to further their educational pursuits during the course of the study. One member obtained a bachelor’s degree, one received a master’s degree, and another pursued a doctoral degree.

It is important to note that our collaborative approach developed organically, stemming from ongoing conversations with the team members who expressed a keen interest in taking a more active role in the design of the interview guide and analysis of the qualitative data. Had we planned to formally assess this approach, we would have used an evaluation framework such as the Kirkpatrick’s model to guide data collection and analysis [16]. Despite this limitation, we believe it is important to relay our real-world experience.

We learned several key lessons in implementing our collaborative approach. For future studies, we will ask our team members from the onset whether they are interested in being actively involved with developing the study tools. We will conduct more in-person training sessions and have an in-country investigator to facilitate trainings and provide ongoing mentoring. We will also hold writing workshops to assist our team members develop manuscripts from the study data.

Success stories such as ours demonstrate how future studies can benefit greatly from leveraging the valuable contributions of all team members. Our approach resulted in professional skill development and educational capacity building. Given our team’s overwhelmingly positive experience, we encourage other investigators to integrate this collaborative approach into their research programs.