Development of a multi-level family planning intervention for couples in rural Uganda: key findings & adaptations made from community engaged research methods | BMC Women’s Health

The study was conducted in selected rural and peri-urban communities of Butambala District, central Uganda located approximately two hours from the capital city of Kampala. The investigative team had been engaging in collaborative research in this area for more than 10 years. Family planning services in this district are integrated into general outpatient services and are provided for free in all public health facilities. Family planning services are also provided by private not-for-profits (PNFPs) and faith-based PNFP facilities, which mainly promote natural methods (i.e., counting days). The public health facilities follow Uganda’s five level decentralized health system structure (I-IV). Health Center IIs and above offer condoms, oral pills, and injectable contraceptives. Health Centre IIIs and above offer intrauterine devices (IUDs) and implants, and Health Center IVs provides non-reversible methods (vasectomy, tubal ligation). Village Health Teams (VHTs), a cadre of community health workers, serve as liaisons between the community and health facilities, and support community family planning efforts. VHTs provide community education about family planning and distribute short-term methods (condoms, oral pills) directly in the community. Also, an international nongovernmental organization, Marie Stopes, provides regular community outreach for all contraceptive methods in selected villages within the district. The villages in this district are mostly homogenous in demographics and size with only small commercial centers (no city within the district).

Community-engaged methods for intervention refinement

A visual depiction of the community-engaged research methods used to gather feedback on and further develop the Family Health = Family Wealth intervention is presented in Fig. 2 to illustrate the overall timeline of study procedures, described in detail below. All study procedures were reviewed and approved by the Institutional Review Boards (IRBs) at the University of Texas at San Antonio (protocol # 19–253, October 2019) and Makerere University School of Public Health (protocol # 748, January 2020). The study was also approved by the Uganda National Council for Science and Technology (May 2020) and by Butambala District Health leadership, who provided formal project endorsement, entry into the health centers in the district, and introductions to key stakeholders for qualitative data collection. Subsequently, two qualitative interviewers familiar with the area of study, the Luganda local language, and experienced in qualitative research methods were hired and trained to assist in the data collection process.

Stage 1 of the intervention development process began with assembling an intervention steering committee (ISC) tasked to guide the tailoring of the intervention to the local community and health system context and to linking the study team to the local communities, clinics, and other stakeholders essential to study progress. The ISC was made up of district health officials, family planning providers, VHTs, and other community stakeholders. Ahead of the planned qualitative data collection, the investigative team first presented the proposed intervention protocol and research plan to the ISC in an in-person meeting in March 2020 to gather initial feedback and begin early refinement of the intervention. This meeting helped to raise issues that needed to be explored further in the planned formative research phase with the community participants (Stage 2 in Fig. 2, described next), and thus informed our interview and focus group tools.

Following a three-month government-mandated COVID-19 lockdown that temporarily halted all research activities (March-June 2020), the formative phase of the research began in June 2020 with the aim of drawing feedback on the intervention content and study procedures from relevant community stakeholders and community members. The research team developed and refined all qualitative data collection tools, which included questions on overall barriers and facilitators to contraceptive use in the local setting (relevant for developing intervention content), as well as questions to elicit feedback on the feasibility and acceptability of the planned intervention approach. In consultation with the ISC, we identified communities for our formative data collection as part of the process of selecting communities for the future intervention trial, aiming to identify communities that were similar across key characteristics. The communities identified were matched on population size (~ 2000), distance to health facilities offering contraceptives, and other contextual factors (e.g., demographics, distance to a trading center).

Four approximately one-hour focus group discussions (FGDs) were conducted with 26 community members (13 women, 13 men), stratified by age and gender (men < 30, men < 30, women < 30, women > 30). Focus groups were moderated by a trained facilitator experienced in qualitative research and a researcher assistant took detailed notes, used later to aid transcription.

Inclusion criteria included being from the selected communities, being of reproductive age (women: 18–40, or an emancipated minor, defined as individuals below 18 years who are married, have a child, or are self-sufficient; men: 18–50 or an emancipated minor), considering oneself married, speaks Luganda, not currently pregnant, and having an unmet need for modern contraceptive methods. An unmet need for family planning was defined as wanting to delay pregnancy for at least a year but not currently using a high-efficacy method of modern contraception; methods (tailored to availability in the local setting) included oral pills, intrauterine device, implants, injectables, and condom use 100% of the time. Since contraceptive uptake among those with an unmet is the goal of the intervention, those already using non-reversible methods (tubal ligation, vasectomy) were not included, as their need is met. While natural methods of contraception (e.g., counting days) can be high-efficacy when used correctly and can be an appropriate person-centered outcome, we only considered high-efficacy methods given that many women use natural methods because of low knowledge and access barriers to modern methods in the local area. FGD participants were compensated 22,000 Ugandan Shillings (~ 6 USD) for their time. See Table 1 for an overview of focus group participant characteristics. The final sample’s demographics (largely Muganda tribe, Muslim as well as Catholic and Protestant religion, and low lifetime experience with modern contraceptives) is representative of the communities selected for the intervention.

Fifteen key informant interviews (KIIs) with community stakeholders who were identified and recruited with help from the ISC were also conducted including: district health officials, family planning providers, VHTs, and cultural, religious, and political leaders from the selected communities. KII participants were compensated 25,000 Ugandan Shillings (~ 7 USD) for their time. All FGD and KII participants provided written informed consent. See Table 2 for an overview of KII participants.

Data from these interviews were transcribed, translated, and summarized. Data were analyzed thematically [27]. Through an iterative review of the transcripts by the investigative team (CM, KMS, SMK) we developed a coding guide informed by the social ecological model to classify barriers and facilitators to contraceptive use in order to inform the development of intervention content. Our specific research questions on the development of intervention content and procedures were used to organize data specific to intervention refinement. Two trained research assistants used an iterative process to apply codes manually within the transcripts and coded narratives were extracted and organized in a Microsoft Excel spreadsheet. Coders met weekly with KMS to discuss new codes and potential themes, and to resolve discrepancies through discussion and consensus. The coders independently coded the transcripts deductively following the coding scheme. New codes drawn inductively from the data were created at this stage. KMS reviewed all excerpts after data were fully coded for consensus or re-coding. Codes that represented thematic elements were collated within the spreadsheet and a final round of review/revision of coded excerpts was conducted to confirm thematic validity. To answer the specific questions relevant for intervention refinement, KMS, CM, and SMK independently created initial impression summaries outlining major themes based on review of the coded excerpts. Through several rounds of discussion and revisions between KMS, CM, and SMK, they merged their separate summaries into one finalized summary of thematic results with representative quotations.”

After completing the analysis of the formative research, we convened a hybrid in-person/virtual meeting (due to COVID-19) in October 2020 with the ISC (see Stage 3 in Fig. 2). The investigative team presented a summary of the primary findings of our qualitative data to the ISC members. In this meeting, we gathered the ISC’s input on the interpretation of our qualitative findings and elicited further feedback on outstanding questions specific to intervention content and procedures.

Using the qualitative research findings and ISC feedback, the research team subsequently refined the intervention protocol outline including the proposed activities per intervention session. This outline was further revised based on an additional round of review and feedback from the ISC, as well as review from the broader investigative team. The intervention protocol was finalized, and the associated training manuals developed and later shared with ISC members for final review.

Finally, the manuals were piloted by CM and two trained facilitators with a single group of community members (7 couples) (Stage 4 in Fig. 2). Couples were identified and recruited with help of the community health worker from Wakiso, a neighboring district with characteristics relatively similar to the study district chosen for the larger intervention pilot in December 2020.

Overall, the pilot group sessions had 14 participants (7 couples); the majority were aged 25-34 (n = 8, 57%), Muganda by tribe (n = 9, 43%), Christian (n = 10, 71%), had attained secondary level of education (n = 7, 50%), and had been married between 1 and 5 years (n = 10, 71%), as described in Table 3.

Each intervention session during the small pilot was audio-recorded and transcribed for investigators to review and give final feedback to the facilitators on the delivery of the materials (e.g., fidelity to the protocol) and make final adjustments. Through this pilot, we sought to assess the facilitators’ accurate delivery of the session content, their experience with the manuals/study materials (e.g., ease of use, flow of sessions), the response from participants (e.g., active engagement, comprehension, issues within couples), the perceived acceptability of the content to participants, and to identify any other issues with implementation (e.g., the total time of sessions). These issues were assessed through CM’s direct observation of sessions, the investigative team’s review of the session transcripts, and feedback from the facilitators, all of which were considered and discussed by the investigative team. The intervention was finalized and delivered thereafter by two trained intervention facilitators in the planned larger trial (evaluation reported in another paper) [26].