Geo-additive modelling of children under-five sleeping under Insecticide treated nets in Nigeria

Abstract

Number of children under-five at risk of malaria in Nigeria is overwhelming and despite several reported effectiveness of Insecticide Treated Nets (ITNs) in malaria prevention, many factors deterring utilization have been reported. To examine spatial variation in South-West region of Nigeria, this study examines level of utilization among under-five children in the region. Data from the 2018 Nigeria Demographic Health Survey is used to examine relationships between different covariates and spatial variation using a semi-parametric model. Gaussian, Poisson and Zero Inflated Poisson distributions were assumed for the response variable and comparisons were made. The Deviance Information Criteria (DIC) is used as selection criteria. With lowest DIC among the competing distributions, Gaussian has the best fit for the model. Spatial variation is observed in the utilization across states. Ondo state has highest net usage among children follow by Ogun state. Both Lagos and Ekiti states have the least. Higher utilizations are observed among rural residents and households without some basic amenities and mass media. Utilization is found to be higher among respondents in the poor category of wealth index. Educational status of mothers is found to affect ITNs utilization. Non-linear effect of age of the children on the number of children sleeping under net has a bathtub shape, indicating decrease in number of children sleeping under net as age increases till around two years and then starts increasing. For effective utilization of ITNs, more efforts are to be put into campaigns on the importance of utilization rather than only ownership.