Effect of urban vs rural context on effectiveness of a community intervention to prevent diarrhoea and stunting in young children in Mali

The Problem
Globally 1.7 billion diarrhoea episodes result in approximately 2 million deaths in 2010.1 Regionally, Africa has the greatest burden. Alongside poor infant feeding, diarrhoea diseases also contribute to infant malnutrition.2-4 Despite new vaccines, treatments and public health measures,3 diarrhoea and malnutrition remain considerable public health problems in low- and middle-income countries (LMIC). The period when a child starts eating solids (usually 6-24 months called complementary food (CF)), is associated with the highest rates of diarrhoea: over 50% of all diarrhoea deaths occur at 6-11 months.1 Most low-income households in LMIC cook ingredients at home or obtain food from informal street food-sellers who prepare the goods in their homes. Therefore home food handling, preparation and storage determine the scale of food contamination.

Although many studies explore the effects of improved water supply, hygiene and nutrition of infant diets on infant diarrhoea and growth development, much less attention paid to studies of food safety remain scarce. Research in this area has been too general to reduce diarrhoea through food contamination. The World Health Organisation (WHO) advocates targeted interventions to support CF safety and hygiene. Ideally CF safety needs to be accompanied with achieving optimal dietary intake for young children, which also remains a challenge in LMIC. Research shows that infant health and safety advice has limited impact on behaviour change unless accompanied by means to motivate and empower mothers in the community. Yet previous interventions targeting diet or diarrhoea have seldom drawn on cultural dramatic arts and community assets to motivate behaviour change. African communities have a particularly strong cultural heritage to underpin such potential impact.

Our Aim
We propose a low-cost, scalable, and adaptable community intervention to reduce diarrhoea and improve the growth of young children in urban-poor and rural Mali. We will assess the effects in both settings, to inform replication and scaling of the intervention, because the dynamics of community life vary in each.

Our Previous Work
We combine two complementary interventions shown to be effective elsewhere in LMIC. A trial in the Gambia (developed through former work in Bangladesh, Nepal and Pakistan) evaluated efforts to improve hygiene and safety of CF, while a trial in Kenya evaluated a community programme to improve breastfeeding and weaning food content.

Our Plan
After adaptation with communities, our intervention will empower local families to implement behaviour change. It will include campaign-like activities such as culturally relevant dramatic arts (drama, songs, stories), public meetings, certifications, and home visits, delivered by a small team: 5 days of community campaign visits dispersed during 35 days and including home visits by trained female volunteers, plus a reminder campaign day at 9 months. We will allocate 120 urban and rural sites in Mali by chance to receive the intervention, or not, and assess 27 households in each site after 15 months.

The study is designed to quantify the influence of urban vs rural context, and to examine other societal influences (e.g. household poverty, women's work, and education, etc). Using observations, interviews, discussion groups, surveys and laboratory tests we will compare the implementation of the intervention in urban-poor and rural settings. Importantly, the intervention is designed to be sustainable through peer-education among mothers and older female volunteers (demonstrated after 32 months in Gambia), thus requiring only small levels of coordination resources from central government.

We will undertake a mixed-method evaluation of a low-intensity, culturally engendered, and scalable community intervention to promote food safety and hygiene and child nutrition (breastfeeding and CF content) practices. We aim to evaluate (qualitatively and quantitatively) the effect of intervention on diarrhoea and growth parameters, understand mediating factors contributing to the outcome along the causal pathway to health outcomes, and assess the role of context in the outcomes, with a particular focus on the influence of urban-poor and rural areas.
Design: Mixed-methods - qualitative/ethnography and a parallel cluster randomised controlled trial
Study duration: 36m
Trial site: Mali: cities of Bamako, Kati and Ségou and rural villages in the same provinces
Trial Management Centre: the University of Sciences, Techniques and Technologies (USTT), Bamako
Intervention: 5 days of campaign community visits dispersed across 35 days (days 1,2,17,15,35) plus volunteer home visits and a 9-month reminder campaign day
Control: 1 day campaign community activities on use of water in households and gardens
Follow-up: 15-months
Minimisation Randomisation after clusters have ratified participation and after baseline
Outcomes: Primary: Observed and reported 24-hour diarrhoea, (primary outcome is observed photographed diarrhoea during 7-9 hours home observation from early morning till afternoon)
Other outcomes include: 4 health, 3 microbiological, 4 behavioural, 1 environmental, and 3 social outcomes; all helping to explain any changes in the causal pathway to transmission of infection and growth parameters.
Sample size: 120 clusters stratified by urban and rural, ethnicity and community size, with 27 mother-child pairs in each: 90% power to estimate a 25% reduction in diarrhoea and 25% relative difference between rural/urban with interaction and 3 measurement points.
Dissemination and Capacity Building Plan will be established and monitored.

A successful intervention to improve complementary food safety and hygiene and improve infant feeding has the potential to impact the health, economic and social dimensions of individual children, families and communities. This benefit can transfer to societal level benefits, and by leading to healthier and more productive populations, result in more prosperous societies. In our trial, this can be achieved for individuals participating in the intervention arms during the life of the trial. If replicated and scaled in the rest of Mali or other African and LMIC countries, this impact could be felt more widely due to reductions in the rate of preventable diseases in children (SDG3). The Community Led Total Sanitation program was one such community-wide intervention tested in Mali (reported 2014), which as the result of its beneficial effect is now being scaled at national level in Mali and adopted by other countries in the region. Robust evidence from our trial on the effect of intervention on improving health outcomes, the mediating variables along the causal pathway, and with reference to the effect of urban/rural context on these, will be important for policy makers and implementers to fund and scale up our intervention. This trial provides such evidence and will also provide ample data on other contextual factors and societal impacts that can guide policy makers on future replication and scalability. The draft intervention manual which we will update after this trial will provide important guidance for non-academic implementers to adapt and deliver the intervention in new LMICs.

These data and the new knowledge produced will be impactful for the academic community as it will answer important research questions while driving new lines of enquiry and research. Immediate academic impact is summarised in the Academic Beneficiaries section. Benefits will extend beyond the immediate researchers involved in this project because, we plan to engage other academics in Mali and more widely within our host institutions. In Mali, much needed PhD degrees will be facilitated (Capacity Building Annex), and the trials unit and laboratory facilities will be developed through training with new methodologies.
If our intervention is proven effective, it could have significant impact across the health system in Mali and other LMICs. Diarrhoea and nutritional related conditions in children are extremely common in LICs such as Mali and utilise a great deal of health system resources as well as leading to days lost from work for family members which affects family income and prosperity. Health services are often over stretched and respond to high rates of communicable diseases and child health related conditions which overwhelm them. Should our intervention be effective, the resulting reduced rates of diarrhoea and illnesses secondary to malnutrition and stunting are likely to release resources within health services in primary and secondary care, which could then be diverted to the care of other conditions.

The PtI section outlines various impactful products produced by the research, and outlines groups of individuals or systems that can benefit in the short and long term. It describes how we intend to realize these impacts. We have a dedicated work package (WP4) with Co-Is in Mali and UK jointly responsible for its delivery. At the heart of WP4 is the Communication and Dissemination Plan (CDP) which will be monitored by the WP Co-Is (OT, OPG). Individual Co-Is will have responsibility for delivering aspects of the CDP and impact related to them and the Project Management Group will monitor the progress of WP4 and CDP on a monthly basis with SMH ultimately responsible. We are confident that since we have involved national and international policy makers, academic and implementing experts from the start in the Advisory Groups we can realise impact from various aspects of our investigations at numerous levels and into the future.