JMIR Publications

ABSTRACT

Background:

Over the past decade in Indonesia, the prevalence of stunting has remained close to 37% of under-five children. The Baduta Program is a multi-component package of interventions developed by the Global Alliance for Improved Nutrition (GAIN), which aims to improve maternal and infant nutrition in Indonesia.

Objective:

To assess the impact of the Baduta Program, a package of health systems strengthening and behavior change interventions, compared with the standard village health services on maternal and child nutrition.

Methods:

The impact evaluation uses a cluster randomized controlled trial design and two outcome assessments. The first will use repeated cross-sectional surveys of mothers of children 0-23 months and pregnant women before and after the interventions. The second will be a cohort study of pregnant women followed until their child is 18 months from a subset of study clusters. We will also conduct a process evaluation guided by the program impact pathway to assess program coverage, fidelity, and acceptance. We will carry out the study in Malang and Sidoarjo districts of East Java, Indonesia. The unit of randomization will be sub-districts. As it is only feasible to randomly allocate the interventions to six sub-districts, we will use constrained randomization to ensure the balance of baseline covariates between treatment groups. The first Baduta Program intervention is health system strengthening, including the Baby-Friendly Hospital Initiative, and training on counseling for appropriate IYCF. The second is nutrition behavior change that includes “Emo-Demos,” a national television advertising campaign, local screening TV spots, a free, text-message service, and promotion of low-cost water filters and hygiene practices. The primary trial outcome is child stunting (length-for-age <-2 Z scores), and secondary outcomes include length-for-age Z scores, wasting (low weight-for-length), anemia, child morbidity, IYCF indicators, and maternal and child nutrient intakes. In the cross-sectional evaluation, the sample size for each survey is 1,400 mothers and their children aged 0-23 months, and 200 pregnant women, in each of the treatment groups. The cohort evaluation will require a sample size of 420 mother/infant pairs in each treatment group. We will seek “Gatekeeper” consent and written informed consent from women participants. The intention to treat principle will guide our data analysis, and we will apply CONSORT guidelines for clustered randomized trials in the analysis.

Results:

For the cross-sectional evaluation, we conducted the baseline survey in February 2015 and recruited 2435 mothers with under-two children and 409 pregnant women. In the February 2017 endline survey, we recruited 2740 mothers with under-two children and 642 pregnant women. The cohort evaluation began in February 2015, and we recruited 729 pregnant women. We completed the follow-up in December 2016.

Conclusions:

Our results will help guide policies to support effective packages of behavior change interventions to prevent child stunting in Indonesia. Clinical Trial: Registry for International Development Impact Evaluations - RIDIE-STUDY-ID-55ad3b0f60c57