Spatiotemporal variability in socioeconomic inequalities in vaccination against COVID-19 in Catalonia, Spain

Objectives: Socioeconomic inequalities have played a significant role in the unequal coverage of the COVID-19 vaccine. Our objectives were to assess the existence of socioeconomic inequalities in vaccination coverage against COVID-19 in Catalonia, Spain; to analyse the spatial variation over time of these inequalities; and to assess variations in time and space in the effect of vaccination on inequalities in COVID-19 outcomes.

Study design: We used a mixed longitudinal ecological design, in which the analysis units were the 373 Basic Health Areas into which Catalonia is divided, between the last week of December 2020 until the first week of March of 2022.

Methods: The study population consisted of health areas in Catalonia, observed between the last week of 2020 until the first week of March, 2022. We specified a joint spatio-temporal model, with dependent variables: vaccination and COVID19 outcomes, which we estimated with a Bayesian approach using R INLA. We controlled for observed confounders, unobserved heterogeneity, and spatial and temporal dependencies. We allow the effect of the explanatory variables on the dependent variables to vary in space and in time.

Results: The health areas with a lower socioeconomic level were those with the lowest vaccination rates and the highest risk of COVID-19 outcomes. In general, those areas located in the upper two quartiles of average net income per person and in the lower two quartiles of unemployment rate (i.e., the least economically disadvantaged) had a higher propensity to be vaccinated than the most economically disadvantaged areas. In the same sense, the greater the percentage of the population 65 years or over the area had, the greater was the propensity to be vaccinated, while areas located in the two upper quartiles of population density and the higher the percentage of poor housing in an area was, the lower the propensity. Higher vaccination rates reduced the risk of COVID-19 outcomes, while COVID-19 outcomes did not influence the propensity to get vaccinated. The effects of the explanatory variables were not the same in all areas or between the different waves of the pandemic, and clusters of excess risk of low vaccination in the most disadvantaged areas were detected.

Conclusions: What caused inequalities in the most disadvantaged areas could be structural barriers, such as the lack of access to information about the vaccination process, and/or logistic challenges, such as lack of transportation, limited Internet access or difficulty in scheduling appointments.