Prioritization of COVID-19 vaccination is improved by including health and socioeconomic factors

By James DuckerReviewed by Emily Henderson, B.Sc.Jul 13 2021

Allocating COVID-19 vaccines has been determined primarily by age group, but a new study shows that other factors including health risks and socioeconomic status are also important to consider when implementing large-scale vaccination programs.

Image Credit: Kandula S and Shaman J, 2021, PLOS Medicine

Improving vaccine prioritization by using population-level health and socioeconomic indicators

The rapid and large-scale implementation of COVID-19 vaccination programs across the USA is well underway, with an estimated 48.4% of the overall population vaccinated as of July 2021.

To determine the focal areas of vaccination programs, individual characteristics such as age and occupation were used to identify populations at risk. However, a new study by Sasikiran Kandula and Jeffrey Shaman from the Department of Environmental Health Sciences, Columbia University, New York, USA, suggests that other factors are also important to consider.

The study, published in the open-access journal PLOS Medicine, uses a series of computer models to demonstrate that including population-level health-related as well as socioeconomic indicators as prioritization criteria for vaccination reduces COVID-19 mortality levels.

Initial efforts to reduce COVID-19 mortality rates in the US focused on prioritizing vaccination for those at a higher risk of severe outcomes, focusing on age group and health to determine where and when populations will receive vaccines.

However, the inclusion of population-level indicators in COVID-19 prioritization of vaccines is understudied, with limited empirical evidence based on a narrow range of factors. To test whether other population-level health and socioeconomic indicators contribute accurately to the risk of COVID-19 mortality, researchers extracted county-level estimates of 14 indicators associated with COVID-19 mortality from public data sources.

Using the data sources on COVID-19 mortality and a range of indicators, researchers used spatial simultaneous autoregressive (SAR) models to assess the proportion of county-level COVID-19 mortality that can be explained by specific health/socioeconomic indicators as well as the effect estimates of each predictor considered.

Models accounted for spatial autocorrelation in response as well as predictors and corrected for biases including case rate at the county level.

Differential risks of severe outcomes from COVID-19 across populations
The SAR models showed high levels of association between COVID-19 mortality and 9 health and socioeconomic indicators. The spatial relationship was prevalent in specific cases including in populations with a higher proportion of chronic kidney diseases and resident populations in nursing homes, which both had the largest individual effect on COVID-19 mortality

These factors further increased effect sizes when adding socioeconomic status, as models explained up to 43% of variance by including both indicator types.

Despite the research demonstrating the existing correlation between health and socioeconomic indicators and COVID-19 mortality, findings are limited due to the temporal lags in reporting COVID-19 infection and death. Due to this lag, estimates of mortality may be underestimated.

Our findings here show that differential risks of severe outcomes from COVID-19 across populations can be in part estimated from the structures and contexts in which the outbreak occurs, for example, a population's quality of health, its access to healthcare, and the disparities therein.

While vaccine supply continues to be limited for most, and especially low- and middle-income, countries, these population-level indicators may help inform optimal allocation."

Further research could build upon the findings of this study by including updated data as well as consider more spatial and temporal scales. For instance, studies could consider whether communities at risk suffered during specific periods of reopening or lockdown and if other spatial indicators are also indicative of higher risks of mortality, including distance to healthcare facilities.

Nevertheless, the present study is the first to attempt to include a wider range of indicators to improve the prioritization of COVID-19 vaccines, demonstrating that future policies need to incorporate additional factors to identify populations at risk of COVID-19 related mortality.