Multilevel and geographically weighted regression analysis of factors associated with full immunization among children aged 12-23 months in Ethiopia.

Immunization is the process of building immunity or resistance to an infectious disease, typically through administering a vaccine. It is one of the most effective strategies for lowering child morbidity and death. It protects against more than 20 potentially fatal diseases, increasing longevity and health. Despite progress, Ethiopia failed to meet its vaccination coverage target. The magnitude of full immunization is different across areas. Therefore, conducting geographically weighted regression to identify the local factors and multilevel analysis to investigate and identify factors associated with full immunization coverage among children aged 12-23 months is necessary. The study was conducted using the 2019 Ethiopian Mini Demographic Health Survey dataset. A sample of 1028 weighted children aged 12-23 months were included in the analysis. Descriptive statistics were used to describe variables. For the spatial analysis, Arc-GIS version 10.8 statistical software was used. Spatial regression (geographically weighted regression) was done to identify factors associated with the proportion of full immunization, and model comparison was based on adjusted R2 and Akaike Information Criteria (AICc). Multilevel mixed-effect binary logistic regression models were fitted to identify factors associated with full immunization. The fitted models were compared based on log-likelihood, deviance, median odds ratio, and Proportional Change in Variance. Finally, statistically significant factors were reported using an adjusted odd ratio (AOR) with a 95% Confidence Interval for fixed effect. All variables with a p-value less than 0.05 in the final model were considered statistically significant factors. In Ethiopia, the overall full immunization coverage among children aged 12-23 months was 40.58%, with spatial variation across regions in Ethiopia. The significant spatial distribution of full immunization coverage among children aged 12-23 months was detected in northern Tigray, Addis Ababa, central Oromia, and southeastern Amhara regions. The proportion of rural residents,the proportion of women aged 35-44 years,  the proportion of women who had ANC 4 and above andthe proportion of women who had PNC were local factors associated with the proportion of full immunization among children aged 12-23 months. Rural residence [AOR 0.27 (95% CI 0.10, 0.70)], family size 4 and above[AOR 0.41 (95% CI 0.17, 0.96)], never breastfeed [AOR 0.026(95% CI 0.003, 0.21)], 1-3 times ANC visit [AOR 0.45 (95% CI 0.23, 0.86)], being from Oromia region [AOR 0.23 (95% CI 0.05, 0.97)], Eastern pastoralist region [AOR 0.09 (95% CI 0.023, 0.35)], age 35-44 years [(AOR 6 (95% CI 1.57, 22.9)], and PNC [AOR 2.40 (95% CI 1.24, 4.8)] were significant factors associated with fully immunization in multilevel mixed effect analysis. Full immunization coverage in Ethiopia is below the global target with significant geographical variation. The high proportion of rural residents, the high proportion of women who had ANC 4 and above, mothers who had a high proportion of PNC, and the high proportion women age 35-44 years were local geographical factors for the proportion of full immunization among children age 12-23 months in Ethiopia. Women who had PNC, ANC visits four or more times, and increased maternal age were positively associated, whereas larger family size, no breastfeeding, rural residence, and being from Oromia and eastern pastoralist region were negatively associated with full immunization. Strengthening maternal and child health services, focusing on rural areas and low-coverage regions, is essential to increase immunization coverage in Ethiopia.