Impact of Behavior Change Communications and Market-based Approach to Delivering Micronutrient Powders on Stunting, Infant Feeding Practices and Anemia in Bangladesh

Overview

This complex evaluation of a large-scale program uses a cluster-randomized design where 20 geographic clusters (subdistricts) were randomized to study two sets of interventions. For the evaluation of behavior change interventions only, the 20 clusters were randomized to 10 intensive and 10 non-intensive interventions areas. For a sub-study to evaluate a market-based model for delivering micronutrient powders (MNP) along with behavior change interventions, there was subsequent randomization to comparison area, MNP-only area, Behavior Change Communication (BCC)-only area and MNP+BCC areas.

The design uses a repeated cross-sectional design, with multiple age-group samples for multiple outcomes \[since each set of outcomes can only be assessed in the relevant age group for the outcome; e.g., exclusive breastfeeding is only assessed in children 0-6 months of age\]. These include: 1. For the evaluation of behavior change intervention only: * 0-6 months for breastfeeding outcomes * 6-23.9 months for complementary feeding outcomes (in the behavior change intervention evaluation) * 24-48 months for child anthropometric outcomes (related to the behavior change interventions). 2. for the substudy on behavior change and micronutrient powder interventions combined - 6-23.9 months for anemia and anthropometric outcomes A community-based random sample of children 0-6 months, 6-23.9 months, 24-48 months old was surveyed at baseline (April-June 2010). The original anemia sub-study design had proposed surveying children 6.23.9 months of age for the MNP intervention in April-June 2013 and for the behavior change intervention in April-June 2014. Based on program implementation timelines, the endline survey dates were extended by one year such that the above age groups were then sampled and surveyed in community-based surveys in April-June 2014. Infant feeding practices will be assessed using before-after intervention-comparison area group differences between 2010, and 2013 and 2014. Specifically, breast feeding will be assessed in children 0-6 months of age (sampled separately) and complementary feeding will be assessed in children 6-23.9 months of age (sampled separately). Stunting will be assessed in the sample of children 24-48 months of age using before-after intervention-comparison group differences between 2010 and 2014. In addition, the investigators will evaluate the processes through which the programs roll out using a mix of qualitative and quantitative research methods. Qualitative research methods will help to understand drivers of promotion and sales of MNPs by frontline health workers, as well as household level determinants and dynamics related to enabling purchase and use of the MNP and adhering to behavioral recommendations. Quantitative surveys of frontline health workers will help document their awareness about IYCF, MNPs, training, and sales, while quantitative surveys of the households will help document household level awareness, purchase and use of MNPs and awareness and adherence to IYCF practices for children in the target age range.