Use of an Adapted Participatory Learning and Action Cycle to Increase Knowledge and Uptake of Child Vaccination in Internally Displaced Persons Camps (IVACS): A Cluster-Randomised Controlled Trial

"[R]esults show that an adapted hPLA method, when implemented in partnership with into pre-existing social groups, was effective at increasing maternal knowledge and improving child vaccination coverage within a 3-month period. The speed of the intervention effect makes it feasible to consider hPLA as an intervention suitable for use in humanitarian emergencies."

In Somalia, which has been chronically affected by conflict since 1991, only about half of the children are immunised against the six major childhood diseases (measles, poliomyelitis, diphtheria, pertussis, tetanus, and tuberculosis). This paper describes a randomised cluster trial, Improving Vaccination Awareness & Coverage in Somalia (IVACS), implemented in camps for internally displaced persons (IDPs) near Mogadishu from June to October 2021. An adapted participatory learning and action (PLA) approach was used in partnership with indigenous Abaay-Abaay women's social groups.

The women's group PLA approach is based on the ideas of Paulo Freire and the importance of social empowerment to address health issues. Teachers and learners engage in a dialogue, exchanging ideas and experiences and implementing a cycle of learning, action, and reflection. PLA interventions have been effective in improving maternal and newborn health in a variety of countries. However, the involvement of stakeholders and the political and security context require particular consideration when implementing social interventions in a conflict setting.

The PLA approach can be implemented either with new groups, specifically formed for that purpose, or with preexisting endogenous social groups. Abaay-Abaay groups (AAGs) are traditional female social groups that emerged in the 1880s and that are found in many parts of Somalia. Typically comprising 10-20 members, these groups usually meet every week or month to discuss local issues and sing traditional religious songs (Digri). Anecdotal reports and formative research indicate that the groups are commonly found in stressed and displaced populations, including most IDP camps in Mogadishu. The majority of members are from the same clan, village, or district and live within the same IDP camp or group of camps. Groups are led by the Khalifada (lead woman), who is usually elected by the group members based on attributes, including her perceived leadership ability and ability to provide exemplary childcare.

As part of the randomised cluster trial, an adapted PLA approach (hPLA) was used in partnership with indigenous AAGs. Adaptation of the PLA manual developed by Women and Children First (UK) for use in this humanitarian emergency context involved, for example, reducing the number of group meetings and developing localised picture cards for common childhood health problems in IDP camps. The hPLA intervention entailed weekly facilitated AAG meetings, a total of 40 meetings from July 27 2021 to September 9 2021; an average of 38 members attended each meeting. All the groups were open to all community members; those who were pregnant or had children were particularly encouraged to attend. There were 5 AAGs, one in each of the 5 intervention camps. Trained facilitators (2 locally recruited women who had completed university education) guided AAG participants through a 4-phase PLA cycle:

1. AAG members identified 9-16 child health problems and then used a voting system to prioritise 3 of them.
2. AAG members identified disease prevention and management practices and barriers to implementing such practices. Popular strategies included awareness raising, mobilising mothers during vaccination days, helping working mothers to take their children to vaccination centres, and environmental cleaning. After phase 2, there was a stakeholders' meeting where the research team facilitated a meeting between AAG member and the humanitarian agencies in the area who were providing vaccination services. Two such meetings were held; one for a group of 2 camps and one for a group of 3 camps that were close to each other. In these meetings, AAG members presented the progress of their PLA meetings to the stakeholders and shared the challenges that they were facing in health seeking, such as long distances to health facilities. One of the agencies (Action Against Hunger) that was running mobile teams in the Khada area offered to extend their mobile teams to the intervention camps as a solution to their challenges.
3. AAG members planned and implemented the solutions and strategies that had been developed in phase 2. Facilitators also visited selected households to check if women were receiving any messages and if they received the right messages. If facilitators felt that some women didn't receive any information and/or didn't retain the information given, they requested volunteer mothers to go back for more sessions. In addition, Action Against Hunger started visiting the camps and vaccinating children up to the age of 23 months with all the routine vaccines.
4. AAGs reflected on their progress in addressing common childhood health problems. All 5 groups in the intervention camps evaluated strategies through self-reflection. These meetings were facilitated by one of the study facilitators using a self-evaluation tool. Feedback was also provided to the group about how things have changed as a result of their interest and solution implementation. Lastly, non-facilitated meetings were attended to assess if groups sustained interest in child vaccination and/or related topics after the end of the facilitated PLA cycles.

The 5 camps allocated to the control group received the usual standard of health care. Community health workers also routinely carried out outreach programmes where they encouraged families to take their children for vaccinations. In the control camps, women continued with their usual AAG meetings. There was no facilitation by study staff, and, to avoid any possible observer effect, the meetings were not observed. However, after completion of the endline data collection, the research officer visited the control camps and provided feedback about the study findings.

Overall, 64.6% of mothers were AAG members at baseline. Maternal preference for getting young children vaccinated was >95% at baseline and did not change. The hPLA ntervention significantly improved measles vaccination (MCV1) vaccination coverage and Penta series completion, which increased by 19 and 20 percentage points, respectively, compared to the baseline values, with an adjusted odds ratio (aOR) of 2.4 for both vaccines. However, adherence to timely vaccination did not (aOR 1.12 95% confidence interval (CI) 0.39, 3.26; p = 0.828). Analysis by vaccine and child age revealed that delivery of oral polio vaccine at birth (OPV0) was very low, which accounted for the lack of improvement in the timely vaccination indicator, despite the large improvements in age appropriate coverage of other vaccines.

The hPLA intervention also improved the adjusted maternal/caregiver knowledge score by 7.9 points (maximum possible score 21) compared to the control (95% CI 6.93, 8.85; p < 0.0001). Possession of a home-based, child health record card increased in the intervention arm from 18 to 35% (aOR 2.86 95% CI 1.35, 6.06; p = 0.006).

This study had high initial levels of AAG membership, which increased by 31 percentage points in the intervention arm. "The high rates of participation are likely to have contributed to the success of the intervention, and argue in favour of utilising indigenous social groups where these are known to be active and popular. Another important factor was the flexibility and responsive nature of the health service suppliers....The presence and willingness of health actors to respond to the group requests was perceived as important in achieving a rapid and successful intervention. So, while the hPLA intervention was designed to work via increasing demand it also indirectly led to important changes in the supply of services. Such a change in service supply may not always be possible in other contexts."

In conclusion, the "results show that an adapted PLA approach, run in partnership with local, indigenous social groups, can help achieve important changes in public health knowledge and practice in a protracted humanitarian context. The approach shows promise for adaptation to additional public health priorities and population target groups. Further work to scale up the approach and to address other vaccines and population groups, is warranted."