Local knowledge to reduce under-five mortality: Initiating participatory action research in rural South Africa

We asked Dr Lucia D’Ambruoso, Deputy Director of the Centre for Global Development, University of Aberdeen and co-author of this recent publication, to tell us about using participatory approaches to strengthen mortality surveillance and reporting systems. The process addresses deaths of children under five years of age, which includes newborn deaths. The approaches can be applied to surveillance of perinatal mortality.

U5 mortality article_MDSR Sept 2017 newsletter_image 3Despite a progressive health policy context, deep social inequalities characterise the disease burden in South Africa. The public health system is also severely under-resourced, fragmented and suffers a human resource crisis. In this context, robust evidence on health, illness and service utilisation among socially disadvantaged people is crucial to inform effective responses.

Participatory action research (PAR) is an approach concerned with redressing power asymmetries between communities and the political and administrative forces that shape health policies. We initiated a PAR process in the Agincourt health and socio-demographic surveillance site (HDSS) in rural north-east South Africa. We convened three village-based groups and held a series of discussions on causes and potential actions to reduce under-five mortality.

U5 mortality article_MDSR Sept 2017 newsletter_image 2

We used ranking, diagramming and participatory photography (Photovoice) to elicit people’s perspectives and build a shared understanding of under-five mortality. The process also examined verbal autopsy (VA) data, inclusive of new indicators on circumstances of mortality, to further verify and amplify these data with local knowledge. The process concluded with a priority setting exercise to consider actions to respond to the issues identified. We also consulted with a Provincial Directorate for Maternal, Child, Women, Youth Health and Nutrition (MCWY&N) to ensure that the process was generating evidence that responded to information needs in the health system.

Poverty, unemployment, inadequate housing, unsafe environments and shortages of clean water were identified as fundamental root causes of under-five mortality by the village-based groups. A series of contributing factors in clinics were also identified: lack of confidentiality, disrespect and abuse, misuse of medications, poorly staffed and equipped clinics, long waiting times and overcrowded facilities.

Actions were prioritised as: expanding clinics, improving accountability and responsiveness of health workers, improving employment, providing clean water, and expanding community engagement. These were fed back to the Provincial MCWYH&N Directorate, and a feedback forum was held with Department of Health (DoH). In these fora, a useful feature of the VA and PAR evidence noted by DoH stakeholders was the complementarity between the statistical VA data, and contextual, qualitative and visual PAR data.

U5 mortality article_MDSR Sept 2017 newsletter_image 1

With the time and resources available, it was not possible to develop the process further, by taking and learning from action, so PAR was not fully achieved. However, commitments were secured among participants, health authorities and funders to continue the process, and move towards acting on the evidence generated.

We aim to extend the PAR process through the following considerations:            

  • Participation for whom? Expanding the process to include the perspectives of the most marginalized people is critically important. We have maintained links with the groups convened to date, and through these we are expanding PAR to include new discussion groups with people who are severely socially disadvantaged.
  • By which means? Increasing participant control over the process, choice of topics, how outputs are discussed, communicated, acted on and learned from is intended to build ownership further. Photovoice generated powerful evidence. Using it to investigate care in facilities may be more contentious, however, and is an area for consideration with providers and participants.
  • Change in health systems. Action, and learning from action, are key elements of the PAR process. We plan to engage in the health system at different levels to enable a process of action and learning from action. Engagement with sectors such as labour, housing, and sanitation will help foster an integrated approach to using the information to achieve change.
  • The role of research. Engaging communities developed rich accounts of avoidable mortality, and responded to information needs in health systems. With HDSS, using PAR can connect communities, researchers and health authorities to develop robust evidence for service delivery, policy and planning. This is highly relevant in South Africa as the government consolidates and expands HDSS infrastructure to inform public policy.

This piece was written by Dr Lucia D’Ambruoso, and has been developed from:

Wariri, O., D’Ambruoso, L., Twine, R., Ngobeni, S., Van Der Merwe, M., Spies, B., Kahn, K., Tollman, S., Wagner, R. & Byass, P. (2017). Initiating a participatory action research process in the Agincourt health and socio–demographic surveillance site. Journal of Global Health, 7(1), 010413. Full text is available at: http://jogh.org/documents/issue201701/jogh-07-010413.pdf

The content has been adapted and shared under the Journal of Global Health’s CC BY 4.0 licence https://creativecommons.org/licenses/by/4.0/

All image credits: © D’Ambruoso 2015

Image 1: Systematising subjective perspectives – ranking; image 2: Participatory photography; image 3: Validating by consensus – diagramming