(P17-024-23) SBCC Intervention Reduces the Proportion of Children Becoming Ill, Though Types of Morbidities Unchanged: A Cluster-Randomized Controlled Trial in Uganda

Objectives: The ONCE (One Nutrition in Complex Environments) was a cluster randomized trial implemented in Agago, Uganda that evaluated the effectiveness of two social and behavior change approached called Nutrition Impact and Positive Practice (NIPP and NIPP+) in improving the health and nutrition of children under-2. We evaluated the interventions’ effects on decreasing the prevalence of common illnesses among children.

Methods: At baseline, 900 households in 60 clusters were randomized to Control, NIPP or NIPP+ (n=300 per arm); data were collected at baseline and endline. At endline we evaluated differences in illness composition using principal component analysis (PCA) to identify symptom patterns and number of symptoms (e.g., cough, vomit, diarrhea, etc.) reported by the caregiver in the previous 14 days. Ordinary Least Squares regression assessed differences in composition and logistic regression determined differences in ‘ever ill’ (yes/no), adjusting for clustering and intervention arm in all models. Biological factors were balanced among the arms at baseline and did not impact model specification. External factors (e.g., diet) were included in the fully adjusted models.

Results: At baseline, more children in NIPP+ (203, 68%) experienced any illness in the prior 14 days compared to Control (181, 63%) (p=0.073). By endline, more Control children (70%) reported any illness compared to NIPP (63%, p< .05) and NIPP+ (64%, p< .05).

On average, children in the control arm experienced 3 symptoms compared to 2 in NIPP and NIPP+ (p < 0.05), though this was because a greater share of children in control reported both fever and malaria. The PCA estimated null differences in illness composition between Control and either NIPP or NIPP+ at endline. In the fully adjusted model, “ever breastfed” and being in the 2 highest SES quintiles were statistically significant (p < .05) and protective against becoming ill. Neither diet nor vaccine status influenced illness in this population.

Conclusions: Both intervention arms decreased the proportion of children becoming ill, but not the types of illnesses. That ever breastfed and SES were important protective factors against illness suggests that SBCC interventions may need to be more intensively targeted to households in lower SES strata.

Funding Sources: U.S. Agency for International Development (USAID) and the Feed the Future Initiative