A qualitative exploration of community ownership of a maternity waiting home model in rural zambia

Context: Ownership is an important construct of sustainability for community-based health programming, though it is often not clearly defined or measured. We implemented and evaluated a community-driven maternity waiting home (MWH) model in rural Zambia. We engaged stakeholders at all levels and provided intensive mentorship to an MWH governance committee comprised of community-selected members. We then examined how different stakeholders perceive community ownership of the MWH. Methods: We conducted 42 focus group discussions with community stakeholders (pregnant women, fathers, elders, and community health volunteers) and 161 in-depth interviews with MWH stakeholders (health facility staff, district health officials, and MWH governance committee and management unit members) at multiple time-points over 24 months. We conducted a content analysis and triangulated findings to understand community ownership of the MWH and observe changes in perceptions of ownership over time. Results: Community members’ perceptions of ownership were related to their ability to use the MWH and a responsibility toward its success. Community and MWH stakeholders described increasingly more specific responsibilities over time. Governance committee and management unit members perceived their ability to represent the community as a crucial component of their role. Multiple respondent types saw collaboration between the governance committee and the health facility staff as key to allowing the MWH to meet its goal of serving the community. Conclusion: The perceptions of community ownership evolved as the intervention became more established. Use of the MWH, and clear understanding of roles and responsibilities in management of the MWH, seemed to foster feelings of community ownership. To improve the sustainability of community-based maternal and child health programs, interventions should be accessible to target communities and clear roles should be established among stakeholders. The Maternity Homes Access in Zambia project constructed 10 MWHs adjacent to rural health centers able to provide obstetric care for uncomplicated deliveries and within 2 hours of time to a referral hospital equipped to care for women experiencing obstetric complications. The intervention was implemented in 4 districts of rural Zambia: Choma, Pemba, and Kalomo (in Southern Province) and Nyimba (in Eastern Province). All study districts are primarily rural with some peri-urban pockets. Choma has 247,860 people, 76% of whom live in rural areas. At the time of the 2010 census, Pemba was part of Choma. Kalomo has 258,570 people, most of whom live in a rural area (93%).24 Nyimba has 77,359 people, 91% of whom live in rural areas.24 We gathered community input from community members and relevant stakeholders in the health system and traditional leadership structures to design an intervention that would meet community standards of acceptability.4,8,9 The resulting 3-pillar conceptual model (core MWH model) focused on: (1) the establishment of quality MWH structures with functional infrastructure and amenities; (2) the need for a community-based system to oversee the daily management, finances, and future maintenance requirements of the MWHs without overburdening the existing health system; and (3) the need to be linked with the health system for clinical care of waiting women and education. The core MWH model met cultural-appropriateness and was aligned with Ministry of Health policy.8,9,23 In accordance with the management pillar, we engaged stakeholders at all levels of the MWH ecosystem before and during the intervention implementation (Figure). We engaged community members, including traditional leadership (i.e., chiefs and the village headmen who represent the chiefs), to sensitize them on the benefits of an MWH, actively participate in the governance and management of the MWH through selected community members, and contribute to the financial and operational sustainability of the MWH. We also engaged health system staff, which included staff at the health facility and district health office levels, to ensure our goals were aligned. For example, we engaged district health staff to participate in steering committees to advise the creation of the MWH governance committees and MWH management units. We engaged health facility staff and community health outreach workers to actively participate in the governance and management of the MWHs and to ensure linkage of the MWH to the facility. The Maternity Waiting Home Ecosystem in 4 Districts in Zambia We provided training and ongoing mentorship to community-elected MWH governance committees and management units (GCMU). The governance committees are comprised of community members and health facility staff. The management units are comprised of community members or health facility staff selected by the governance committees. The governance committee is responsible for managing the MWH, mobilizing resources, and overseeing the management unit to ensure sustainability of the MWH. The management unit is responsible for the daily operations of the MWH and management of MWH assets. Additionally, we covered the start-up costs for community-led income-generating activities that could help support costs associated with the MWH and contribute to its financial sustainability. The project phased out supporting implementation in April 2018 but continued to monitor intervention activities through October 2018. The evaluation of the implementation of the intervention has been described elsewhere.25 To evaluate the sustainability of our MWH intervention, we relied on findings from the formative evaluation and Scheirer and Dearing’s framework for the sustainability of public health programs.26 The framework determines sustainability by asking: (1) whether program activities were continued after external support ends, (2) whether community-level partnerships or coalitions developed during the funded program were maintained, and (3) whether new organizational practices, procedures, and policies that were started during program implementation were maintained. We hypothesized that community ownership may be an important mediator of these constructs and therefore an essential component of sustainability. This hypothesis was in line with findings from our formative work, which qualitatively underscored the importance of ownership of the MWH intervention by the community.4,8,9 We deliberately did not define community ownership, but rather allowed our stakeholders to explain ownership in their own words. We conducted 42 focus group discussions (FGDs) and 161 in-depth interviews (IDIs). The FGDs were conducted with 412 community members (14 groups of pregnant or recently delivered women, 10 groups of men with a child under age 1, 9 groups of community elders, and 9 groups of community health volunteers). Safe Motherhood Action Groups made up the majority of community health volunteers, but traditional birth attendants were also part of the FGDs. The IDIs were conducted with MWH governance committee and management unit members, and health systems staff (health facility staff, district health officials). FGDs were conducted at 3 timepoints: immediately following intervention launch (Octo-ber 2016 to January 2017); during the intervention (August 2017 to September 2017); and after implementation phaseout (April 2018 to May 2018). IDIs were conducted at 4 timepoints: immediately following intervention launch (October 2016 to January 2017); during the intervention (April 2017 to June 2017 and November 2017 to January 2018); and after implementation phaseout (July 2018 to October 2018). We used convenience sampling to select the most senior person available on the day of visit for the district staff, health facility staff, governance committee, and management unit IDIs. Community health volunteers recruited FGDs participants from varying distances from the health facility. Both qualitative instruments captured basic demographics and had questions that elicited perceptions of the MWH operations and stakeholder roles as well as perspectives on health facility engagement, community ownership, and long-term sustainability. Local data collectors fluent in English and the local languages, who were trained in qualitative interviewing techniques, the interview guides, and research ethics, administered the IDIs and FGDs. Data collectors were not members of the intervention implementation team, which provided direct mentorship and support to the GCMU, as described above. Data collectors participated in a refresher training before each round of qualitative interviews. Predefined probes were adapted and refined based on results from each previous round. IDIs and FGDs were audio recorded, translated into English, and transcribed verbatim into Microsoft Word. Transcripts were systematically coded in NVivo version 11 (QSR International). The main coding nodes were identified a priori based on the questions and probes in interview guides. Transcripts were double coded against the theoretical framework and to a topic or theme. Additional nodes were added as themes emerged during coding. We conducted a content analysis to assess respondent definitions of community ownership and applicability to the MWH intervention among respondent types and over time.27 Demographic data were captured in Survey-CTO Collect version 2.212 (Dobility, Inc.) and analyzed in SAS version 9.4 (SAS Institute Inc.). Proportions were calculated for respondent sex, occupation, and school attendance. Means and standard deviations (SD) were calculated for respondent age and highest grade completed. We had missing data (n=24) for years of education for elders and community health volunteers at project phaseout. We obtained ethical approval through the Boston University Medical Campus Institutional Review Board and the ERES Converge Institutional Re-view Board in Lusaka, Zambia, and approval by the Zambian National Health Research Authority. Written informed consent was obtained from respondents in the language they were most comfortable using: English, Chinyanja, or Chitonga.