Estimated effect of media use on mothers' vaccination of their children in Sub-Saharan Africa: a quasi-experimental propensity score matching analysis using DHS data.

BACKROUND

Media channels are a primary source of health information for mothers/caregivers to enhance their children's vaccination uptake, particularly in Sub-Saharan Africa, where there is a dearth of health workers. WHO set mass media exposure as a potential strategy to increase vaccination uptake. However, estimating the effect of media exposure on child vaccination using observational studies suffers from problems of selection bias. Thus, this study used the propensity score matching (PSM) analysis to estimate the effect of media exposure on childhood vaccinations.

METHODS

We used nationally representative demography and health survey data collected from 2019 to 2024 among 23,194 children aged 12-23 months in SSA countries. A total of 23,194 weighted sample were included in the analysis. We employed PSM analysis with the logit model using the psmatch2 command package in STATA to estimate the causal effect of media exposure on women's uptake of child immunization. The Average Treatment Effect (ATE), average treatment effect on the treated (ATT), and the average treatment on the untreated (ATU) represent the overall effect of media exposure on the entire population, among women who had media exposure, and the hypothetical effect if unexposed women had experienced media exposure, respectively. Radius matching with a caliper width of 0.01 was employed to match the groups. The quality of matching was examined statistically and graphically. Sensitivity analysis was done to test the robustness of the PSM estimate using the Mantel-Haenszel test statistics.

RESULTS

The overall prevalence of full child vaccination was 61.27% (95% CI (60.64%, 61.89%)) in the treatment group. In the PSM analysis, the ATT values in the treatment and control groups were 0.68 and 0.59, respectively, indicating that the childhood immunization uptake increased by 8.76% because of media exposure. The ATU values in the treatment and control groups were 0.51 and 0.62, respectively. This showed that for the women or caregivers who did not have media exposure, the probability of vaccinating their children would have increased by 10.8% if they had media exposure. The final ATE estimate was 0.0828 among the samples. Quality of matching was good, and the estimates were insensitive to hidden bias.

CONCLUSION AND RECOMMENDATIONS

This study finding highlighted that policymakers and planners in SSA countries should give great emphasis to further enhancing media exposure coverage for target groups since it has a substantial impact on improving childhood vaccination uptake. Further research is recommended to examine the cause-effect relationship of media exposure on childhood vaccination by including more proximal variables that were not observed in the present study.