RESUMEN
INTRODUCTION: Vitamin A has been one of the most important micronutrients which are necessary for the health of the children. In developing countries, the supplementation of vitamins under a regular schedule had different constraints. Awareness, access, and resource limitations were usually the problem. In the current study, we analyzed the data from the demographic health survey (EDHS) 2016 to uncover the spatial distribution, predictors, and to provide additional information for policymaking and interventions. METHODS: In this analysis, we applied intra-community correlation to measure the random effect; global Moran's I to test the nature of variance in the null model; proportional change in variance to check the variance of null and neighborhood in subsequent models. We used STATA 15 for prediction; ArcGIS 10.7 for the spatial distribution of vitamin A supplementation; SaTscan 9.6.1 to specify location of clustering were the applied soft wares. After confirming that the traditional logistic regression cannot explore the variances, we applied multilevel logistic regression to examine predictors where p-value <0.25 was used to include variables into the model and p-value<0.05 was used to declare associations. We presented the result using means, standard deviations, numbers, and proportions or percent, and AOR with 95% CI. RESULT: The vitamin A coverage was 4,029.22 (44.90%) in Ethiopia in 2016. The distribution followed some spatial geo-locations where Afar, Somali were severely affected (RR = 1.46, P-value < 0.001), some pockets of Addis Ababa (RR = 1.47, p-value <0.001), and the poor distribution also affected all other regions partially. Place of delivery 1.2(1-1.34), primary and secondary education 1.3 (1-1.6), media exposure 1.2(1.1-1.4), having work 1.4(1.2-1.5), and all visits of ANC were positively influenced the distribution. CONCLUSION: The distribution of vitamin A coverage was not random as per the EDHS 2016 data. Regions like Afar, Somali, and some pocket areas in Addis inquires immediate interventions. Pastoralist, agrarian, and city administrations were all involved from severe to the lesser coverage in order. Since factors like Place of delivery, education, ANC, media exposure, and having work were showed positive associations, interventions considering awareness, access, and availability of service need more attention than ever.