Background:
The COVID-19 pandemic highlighted the links between socio-economic disparities and health inequalities, notably with respect to vaccination. In this context, the present contribution investigated the association between COVID-19 vaccination and demographic and socio-economic factors in Belgium at area and individual levels.
Methods:
The Belgian vaccine register for the COVID-19 vaccination campaign (VACCINNET+) was linked at individual level with demographic and socio-economic variables from the DEMOBEL database. For all adult individuals tested for SARS-CoV-2 (LINK-VACC sample), demographic and socio-economic indicators were derived and their impact on vaccination coverages at an aggregated geographical level (municipality) was quantified via the use of a composite factor. The same indicators were calculated for the full Belgian population for comparison purposes.
In a second step, a multilevel approach was considered by fitting hierarchical logistic regression models to the individual level LINK-VACC data to disentangle the individual and municipality effects allowing to evaluate the added value of the availability of individual level data in this context.
Results:
The composite factor built from income deciles, migration background, and household composition shows consistent results with earlier findings based on individual level data for similar indicators. Indeed, it was seen at the individual level that persons with lower household incomes, a migration background, and/or belonging to a household with only one adult were less likely to be vaccinated, and these inequalities translate into disparities observed at the municipality level.
The hierarchical models show that taking into account municipality effects when analyzing individual level data does not dramatically change the estimates related to demographic and socio-economic indicators in this case. However, they provide a more accurate description by modelling explicitly part of the variability related to the neighborhood.
Conclusion:
The most important effects observed at the individual level are reflected in the aggregated data at the municipality level. Multilevel analyses show that most of the demographic and socio-economic impacts on vaccination are captured at the individual level. Nevertheless, accounting for area level in individual level analyses improves the overall description.
