Part of: Enhancing evidence-based infant feeding policy implementation in Malawi �

Professor Anne Matthews

Professor Ellen Chirwa

Dr Aisling Walsh

Dr Pieternella Pieterse

Exploring how and why Care Groups work to improve infant feeding practices in Low and Middle Income Countries: a realist review

Who are we, the researchers?

Professor Anne Matthews

Dublin City University

Principal Investigator

​

Professor Ellen Chirwa

Kamuzu College of Nursing

Principal Investigator

​

Dr Aisling Walsh

Royal College of Surgeons in Ireland

Co-investigator

Dr Pieternella Pieterse

Dublin City University

Post doctorate researcher

Enhancing evidence-based infant feeding policy implementation in Malawi = 3-year research project funded by the Irish Research Council’s Collaborative Alliances for Societal Challenges fund

Exploring how and why Care Groups work to improve infant feeding practices in Low and Middle Income Countries

----

​

Realist Evaluation of Care Group implementation

in Malawi

The Baby Friendly Hospital and Baby Friendly Community Initiatives:

a scoping review

​

desk review

field research

Enhancing evidence-based infant feeding policy implementation in Malawi

​

Realist synthesis:

We will present the findings from our realist synthesis today – it is a desk review,

but used a ‘reference group’ of experts to discuss our ideas and theories with

Exploring how and why Care Groups work to improve infant feeding practices in Low and Middle Income Countries

​

Using the Realist Research approach

​

Using Realist Research allows us to:

• Conduct research that focuses on ‘how, why, when, for whom and under what circumstances’

​

• Research interventions that are complex and whereby the context can play a role in influencing the outcome.

​

Realist research focuses on three components:

1-Context: both the actual context in which an intervention takes place (so the physical

place, plus all contextual factors such as religion, patriarchy, inequality, low-literacy, etc.) and ALSO the intervention itself.

​

2-Mechanism is what happens, it can be a description of what is triggered by the context (often what happens in the coming together of the physical/social context AND the intervention).

​

3-Outcome is usually a measurable thing (behaviour change, different action, improved health status, etc.) that is the result of C and M.

​

We published a research protocol that set out the steps of the research we planned to undertake:

https://systematicreviewsjournal.biomedcentral.com/articles/10.1186/s13643-020-01497-1

We found that the published texts we found could be divided into four types:

1. Peer reviewed articles
2. Programme evaluations
3. Guidance / ‘how to’

documents

4. Case studies and articles

We found 82 texts in total and excluded 20 and ‘set aside’ 20, so we are left with 42 texts.

​

The set aside texts provide us with good insights but cannot be linked to actual Care Group implementations

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

We created 9 Initial Programme Theories and 11 C-M-O configurations:

Evidence for these findings comes primarily from extracted texts:

​

CMO-01 shows that in a context (C) where women who are elected Care Group volunteers and are provided with a group meeting structure, a manageable workload, a T-shirt or wrap that distinguishes them from others in her group, flipcharts and a bag to carry their Care Group items, these women can be motivated by this (which is all provided by the implementing NGO), which will encourage them to remain engaged in the program.

​

The examples below is one of many citations that were found in the analysed texts; both confirming and showing that an absence of the right ‘inputs/supports’ often leads to an absence of initial motivation and ‘getting off to a good start’:

​

​

[Care Group volunteers] worked solely on a voluntary basis and received supportive supervision from [project] field staff. When interviewed, [CGVs] reported that the training they received, the positive changes in community health behaviors, and the recognition they received for their contributions to their communities were the factors that motivated them to continue their work. They reported that they felt that their workload as volunteers was manageable… The volunteer dropout rate has been very low in spite of very limited material support from the project. Capps, et al 2011, p 23

​

​

The training provided by UBALE through government staff was sometimes perceived as either disjointed (many subjects with little connection) or irregular (months between related trainings) by community service-providers and leaders… which affected their ability to complete reports, understand module messaging, and engage women in true behavior change. There was minimal to no follow-up on whether or not messages were delivered from [care group] leaders to the community according to standards (p15-16)... the capacity of the project to support Lead Mothers to move beyond providing messages to engaging with beneficiaries for accompaniment on behavior change is limited… Best et al 2017, p 17

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

Key findings A:

​

Within the Care Group Volunteer theories and the Care Group Neighbourhood Group Member theories, we see two mechanisms emerge very strongly:

1. motivation/self efficacy
2. empowerment

​

We also see how these mechanisms/ feelings are generated by three different interventions or dynamics:

1. the NGO’s training, supportive supervision, etc.
2. the community’s response (and for CGVs positive feedback/relationship vis-a-vis CG-NGMs): positive and self-sustaining group dynamics
3. Self-efficacy of CGV and CG-NGMs, feeling confident and respected, trying new behaviours

​

If we analyse the texts in relation to 1, 2, and 3, we believe that at the start of every Care Groups, there is only 1. NGO’s training, supportive supervision, etc. Over time, I and II emerge if the context is right, which then leads to the emergence of 2 and 3, and a virtuous circle of more I and II leads to good outcomes.

Usually the right context seems to be generated by NGOs devoting sufficient time and staff to the set up and ongoing supportive supervision of the Care Groups, and by conducting research into the target area and addressing, if needed, contextual hindrances.

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

NGO provides CGV w training, incentives, supportive supervision

CGV equipped and confident to convene groups and pass on messages

Neighbourhood groups in action, CGVs and CG-NGMs attend regular meetings

C

M

O

These findings visualised:

over time…

Confident and respected CGV continues to enjoy regular training, supervision

Group dynamics, confidence, trust empowerment

virtuous circle

Neighbourhood groups in action, CGVs and CG-NGMs attend regular meetings

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

CMO-01

CMO-02

CMO-03

Confident and respected CGVs continue to host neighbourhood groups, enjoy good feedback

Positive group support, trust, confidence in themselves, self-empowerment

Neighbourhood groups active, CGVs and CG-NGMs drive continuation.

Self empowerment, feel more respected

Key findings B:

​

Within the CMOs that focus on NGO interaction with family members/community leadership and with district-level MoH, we see another interesting finding emerging:

​

The outcomes of these CMOs cannot be conceived as an outcome in their own right, they are merely an alteration to the context in which the community-level Care Group activities take place.

​

These types of ‘alterations to the context’ take two different shapes:

1. The NGO staff (often promoters) influence the community leadership, men within the community, others who ‘hold power over women’s reproductive health decision making’ (e.g. TBAs) to create a conducive environment in which it is easier for CG beneficiaries to adopt new behaviours.
2. NGO staff has also been recorded as removing practical barriers to the adoption of healthy behaviours, most often in the form of promoting, expanding or establishing local drug sellers (iCCM, socoristas, etc.). In one other example the NGO worked with market traders to improve the sale of iodised salt.

​

We believe that establishing a step with the Care Group implementation guidelines that should involve checking the need for any ‘alteration of the context’ should be added, as this is too often overlooked by NGOs that are new to Care Group implementation.

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

CG promoter talk to community leadership about ANC importance

Regular visits create trust and acceptance among community leadership

Community leadership endorses regular ANC visits, promotes this in meetings

It becomes acceptable for women to attend ANC meetings, even if this reduces time wife can devote to chores

C

M

O

One practical example of what the alteration of context can look like (alteration type 1):

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

CMO-07

CG project management works with MoH to allow expansion of iCCM with additional medicines

Healthcare supply and demand are in balance

Mothers have learnt to recognise possible malaria and can access test + treatment

C

M

O

One practical example of what the alteration of context can look like (alteration type 2):

NGO and MoH recognise they have shared objectives (population health improvement)

CG project management includes responsibility of training of iCCM within project activities

CGV promote timely accessing of healthcare for children under 2 + teach danger signs for key illnesses

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

CMO-09

Evidence for these findings comes primarily from extracted texts that show:

• Care Group interventions that successfully achieve their outcomes, tend to increase demand for healthcare due to promotion of tests and treatment of lifesaving childhood ailments, for immunisation, ANC, facility delivery, etc.
• Only be simultaneously improving demand can NGO-driven Care Group interventions ensure a lack of healthcare supply does not stifle future demand.

​

Project‘s capacity to build partnerships with the communities, promote behavior change at the community level, and build community-based programs for treatment and referral. Therefore, it recognized that the Project was filling an important void. As one MOH official noted ‘The Project is doing what the MOH could not do on its own’.(p26)… Making Socorristas available in many of the Project villages provided ready access to antibiotic treatment for childhood malaria and pneumonia and assistance in determining whether referral was warranted or not… The Village Health Committees and the Socorristas were fulfilling a felt need in the community, and Many of those interviewed indicated that they thought the village-level work would continue, now that villagers have learned new skills and practices and how to teach these to others. they had earned the full support of the MOH. No additional funding or assistance is required for them to continue functioning. Perry et al 2006, p25

​

Although ‘Knowledge Practice Coverage’ [survey] results for the indicator of the “percentage of mothers of children age 0 to 23 months who know at least two signs of childhood illness that indicate the need for referral” showed great improvement (40.4% at Baseline, 94.7% at Mid-Term Evaluation), improvement was not found for specific care seeking indicators for malaria nor use of ORS. It is likely that the lack of essential medicines seen during Health Facility Assessment in Years 1 and Years 2 … is affecting care seeking decisions. Jennings, 2010, p19

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

Exploring how and why Care Groups work to improve infant feeding practices in LMICs

Conclusion:

​

Our research brought to light some interesting findings about ‘how Care Groups work’

• The multi-directional nature of group dynamics and motivation demonstrated how important it is to ensure that a Care Group intervention gets the intervention right, in order to generate the right group dynamics and motivational forces
• We also unpacked the inner workings of Care Group implementers ‘creating a conducive environment’ – this is probably not discussed often enough in the Care Group guidance

​

​

• Overall, I believe that we have demonstrated that realist research can be a very useful approach to evidence synthesis and evaluation.

​

​

​

We welcome your questions now

​

and are happy to continue the discussion via email:

pieternella.pieterse@dcu.ie

​

​