Journal Articles

UNICEF operates in 190 countries and territories, where it advocates for the protection of children’s rights and helps meet children’s basic needs to reach their full potential. Embedded implementation research (IR) is an approach to health systems strengthening in which (a) generation and use of research is led by decision-makers and implementers; (b) local context, priorities, and system complexity are taken into account; and (c) research is an integrated and systematic part of decision-making and implementation. By addressing research questions of direct relevance to programs, embedded IR increases the likelihood of evidence-informed policies and programs, with the ultimate goal of improving child health and nutrition.

This paper presents UNICEF’s embedded IR approach, describes its application to challenges and lessons learned, and considers implications for future work.

From 2015, UNICEF has collaborated with global development partners (e.g. WHO, USAID), governments and research institutions to conduct embedded IR studies in over 25 high burden countries. These studies focused on a variety of programs, including immunization, prevention of mother-to-child transmission of HIV, birth registration, nutrition, and newborn and child health services in emergency settings. The studies also used a variety of methods, including quantitative, qualitative and mixed-methods.

UNICEF has found that this systematically embedding research in programs to identify implementation barriers can address concerns of implementers in country programs and support action to improve implementation. In addition, it can be used to test innovations, in particular applicability of approaches for introduction and scaling of programs across different contexts (e.g., geographic, political, physical environment, social, economic, etc.). UNICEF aims to generate evidence as to what implementation strategies will lead to more effective programs and better outcomes for children, accounting for local context and complexity, and as prioritized by local service providers. The adaptation of implementation research theory and practice within a large, multi-sectoral program has shown positive results in UNICEF-supported programs for children and taking them to scale.

We examine the effect of the Zambia Child Grant Programme – an unconditional cash transfer (CT) targeted to rural households with children under age five – on height-for-age up to four years after programme initiation. The CT scheme had large positive effects on nutritional inputs like food expenditure and meal frequency, but no impact on child height-for-age. Production function estimates indicate that food carries little weight in the production of child height in the study sample. In settings with poor health infrastructure and harsh disease environments, a stand-alone CT is unlikely to address long-term chronic malnutrition unless accompanied by complementary interventions.

The ability to correct deficiencies in early childhood malnutrition, what is known as catch-up growth, has widespread consequences for economic and social development. While clinical evidence of catch-up has been observed, less clear is the ability to correct for chronic malnutrition found in impoverished environments in the absence of extensive and focused interventions. This paper investigates whether nutritional status at early age affects nutritional status a few years later among children, using panel data from China, South Africa and Nicaragua. The key research question is the extent to which state dependence in linear growth exists among young children, and what family and community level factors mediate state dependency. The answer to this question is crucial for public policy due to the long-term economic consequences of poor childhood nutrition. Results show strong but not perfect persistence in nutritional status across all countries, indicating that catch-up growth is possible though unobserved household behaviours tend to worsen the possibility of catch-up growth. Public policy that can influence these behaviours, especially when children are under 24 months old, can significantly alter nutrition outcomes in South Africa and Nicaragua.