A Pair-matched Community Intervention Trial to Assess the Impact of an Integrated Health and Development Intervention on Child Survival and the Millennium Development Goals in 10 Sub- Saharan African Countries

Design and population The design is a pair-matched community intervention trial. Village clusters with high levels of malnutrition were selected from rural areas in ten sub-Saharan African countries to reflect a diverse range of agro-ecological zones, farming systems, disease profiles, and infrastructure challenges. MVP sites represent 80 villages in 14 clusters across 10 countries, covering nearly 500,000 people. For each intervention cluster, a matched comparison cluster has been selected at random to participate in the evaluation.

Outcomes The primary outcome is the under-5 mortality rate. Secondary outcomes are levels of coverage with essential maternal-child health interventions and related MDG indicators for poverty, nutrition, education, and environmental health.

Sample size calculation The assessment follows 6000 households across intervention and matched comparison villages at baseline, and after 3 and 5 years of intervention exposure. With 10 paired clusters, the study is powered to detect a 40% difference in the U5MR between the two groups.

Analysis plan The analysis will use a two-staged pair-matched cluster level analysis, and will be complemented with multilevel modeling. Reporting will adhere to Transparent Reporting of Evaluations with Non-randomized Designs (TREND) guidelines.

Implementation science A portfolio of qualitative implementation science (process evaluation) will complement the quantitative assessment, and involves interviews with implementers, partners, and project beneficiaries. This analysis will address questions about: the feasibility of the interventions; the timing and sequence of their introduction; key contextual barriers and facilitators to implementation; and potential synergies achieved from the integrated multisector approach.