Exploring factors influencing childhood immunization status in East Africa using multilevel ordinal logistic...

Childhood vaccinations are crucial in safeguarding children from infectious diseases and are recognized as one of the most cost-effective public health interventions. However, children in East African countries face more than a fifteen-fold increased risk of death from vaccine-preventable diseases compared to those in high-income nations. This study aimed to identify the factors influencing childhood immunization status in East Africa.

A sample of 22,734 children aged 12 to 23 months was included to assess immunization status, utilizing recent data from the Demographic and Health Survey conducted between 2015 and 2022 across ten East African countries. A Level-3 multilevel generalized odds model with a logit link function was employed for the analysis.

Among the 22,734 children in the sample, only 67.4% were fully immunized, 27.7% were partially immunized, and the remaining were not immunized at all. The null hypothesis of proportionality was rejected based on the Brant test. Consequently, various partial and non-proportional odds models were fitted, with the generalized odds model demonstrating the best fit compared to other ordinal regression models. The findings indicated that 43.14% of the variation in children's immunization status was attributable to differences between countries, while 18.18% was due to variations between regions. Specific factors associated with immunization status revealed that mothers who attended antenatal care were 1.23 times more likely to fully immunize their children compared to those who did not, and those who received postnatal care were 1.13 times more likely to do so. Additionally, mothers who had antenatal and postnatal services were 1.07 and 1.08 times more likely, respectively, to fully or partially immunize their children compared to those who did not.

The fitted generalized odds model indicated that several factors significantly associated with childhood immunization status included maternal age, number of antenatal and postnatal care visits, tetanus injections received by mothers, vitamin A intake, presence of health documentation, place of delivery, birth order, mother's occupation, sex of the household head, distance to health facilities, maternal education, community maternal education, community wealth index, and community media exposure. Therefore, it is recommended that interventions focus on enhancing household wealth, educating mothers, and improving health systems.