When partnerships work: how evidence informed UNICEF’s scale-up
By Siobhan McDonough, Lilian Lehmann, and Jeffery McManus
Since 2015, IDinsight has engaged governments and NGOs in embedded learning partnerships: we provide data-driven advisory services on clients’ most pressing questions in real-time. These long-term partnerships allow flexibility and innovation, since we can support several of a client’s highest-priority projects with whichever analytical tools best fit their needs. This post explores that process and the outcomes of a two-year learning partnership with UNICEF Kenya.
At IDinsight, we consider our work a success when clients use data to improve social sector programmes. As part of an embedded learning partnership with UNICEF Kenya, IDinsight found that a child-focused sanitation programme involving nutrition messaging had modest but statistically significant improvements on key sanitation and nutrition childcare knowledge and practices. Our evidence eventually informed UNICEF’s scale-up and revision of the programme.
The problem: child deaths relating to undernutrition and diarrhoea
Child stunting, or low height for age, is a persistent problem in many of Kenya’s rural counties where UNICEF operates. Kitui County has one of the highest stunting rates in the nation, with 46 percent of the population exhibiting signs of stunting. Factors related to unhygienic sanitation, especially faecal contamination, and poor nutrition practices, put children at risk for stunting.
A solution? Combine sanitation and nutrition programmes for caregivers
Improving sanitation and hygiene practices can be challenging, especially when it requires changing the behaviours of an entire community. One type of programme, known as community-led total sanitation or CLTS, looks to facilitate behavioural change by creating new social norms around open defecation. The goal of CLTS is to end open defecation by igniting community interest in building simple toilets, such as pit latrines. These sanitation programs have had some success: there is evidence CLTS increases latrine coverage and use, and in some cases, it reduces diarrhoea and stunting.
Nutrition programmes are another way organisations like UNICEF look to reduce child mortality and stunting. Usually these programmes promote breastfeeding for mothers and enable or encourage them to take micronutrient supplements. Prior research suggests that these interventions can reduce deaths of children under six years by 15 percent.
Despite their shared connection to addressing child undernutrition and stunting, sanitation and nutrition projects are often run separately.
UNICEF Kenya’s sanitation and nutrition teams and the Kitui County Public Health Office jointly designed an integrated programme called SanNut. The project bolstered the existing community sanitation initiative with a set of nutrition behaviour-change messages targeted at caregivers of young children.
Our evaluation and findings: modest improvements in sanitation and nutrition knowledge and practices
As part of our embedded learning partnership with UNICEF Kenya, IDinsight designed and conducted a randomised controlled trial to evaluate the impact of SanNut on caregiver knowledge and practices. A total of 604 villages were randomly assigned to either the sanitation program alone (control), or both the sanitation program and the nutrition program (treatment). IDinsight surveyed 4,322 caregivers on sanitation and nutrition practices. We also conducted a process evaluation to assess whether there were gaps between the programme’s expected and actual implementation that could be improved during programmatic scale-up.
SanNut led to modest improvements in sanitary knowledge and practices emphasized by the program, especially safe handling of child faeces — a recognised gap in CLTS, which largely focuses on adult sanitation and hygiene. Caregivers in treatment villages were more likely to mention ‘lack of handwashing’ after handling child faeces as a potential cause of diarrhoea. They were also more likely to report having safely disposed of child faeces and having a stocked handwashing station as compared to caregivers in control villages. Families in the programme were also less likely to report incidents of child diarrhoea.
There was also a modest improvement in parents’ nutrition knowledge, especially in regards to how soon after birth to initiate breastfeeding, although there was no detectable impact on self-reported nutritional practices, such as breastfeeding, Vitamin A supplementation, or deworming.
One concern we had with combining the nutrition and sanitation programs was information overload: the additional nutrition messaging could crowd out the important sanitation messages. However, additional information about nutrition did not reduce the effectiveness of the sanitation messaging on families’ sanitation practices over time: the adult sanitation practices associated with CLTS, including latrine construction and maintenance, were similar in treatment and control villages.
Informing scale-up and programme refinement
IDinsight’s evidence informed UNICEF’s approach to the programme in two major ways:
1) Since the programme improved families’ sanitation practices and nutrition knowledge without adversely affecting other sanitation components, UNICEF is scaling the integrated sanitation and nutrition programme to a second county in Kenya, West Pokot, which has an estimated 100,000+ children under 5 and the highest child stunting rates in the country.
The programme served as an add-on to the existing sanitation program, which also helped to reduce implementation costs. The cost is only a few dollars per child, which could help scale up the combined programme at a more accelerated pace.
2) Since the programme’s effects on families’ nutrition knowledge did not translate into changes in their practices, UNICEF used findings from IDinsight’s process evaluation to improve programmatic delivery in the scale-up. They are refining the nutrition component to increase caregiver attendance and make the message more salient.
This embedded learning partnership was a success because UNICEF was committed to using data to test and revise their programme. Because this was a long-term partnership, IDinsight was able to work closely with UNICEF to refine the programme even after the randomised controlled trial component was over. Use the links to read more about our embedded learning partnerships with other NGOs and five governments in Africa and Asia.
Additional References
For additional references, see our BMJ paper about this intervention.
