Spatio-temporal model for mapping COVID-19 risk
The COVID-19 was a major threat to public health around the world from the beginning of COVID-19 pandemic. The U.S. was one of the countries with the most COVID-19 cases. Despite the mitigation efforts to control the disease at both local and national levels, the number of COVID-19 cases in the U.S. remained high throughout the pandemic. This study focused on Cook County in Illinois. During the COVID-19 pandemic, Cook County was one of the counties with the highest COVID-19 cases in the U.S. This study described the spatial and temporal dynamics of COVID-19 risk in two-week periods from August 2020 to December 2020 in Cook County. This study also assessed the impact of neighborhood socioeconomic and demographic on COVID-19 incidence. The Bayesian spatio-temporal model was used to produce COVID-19 risk maps and to evaluate covariates' effects. The results show the spatial heterogeneity in COVID-19 risk from time to time, with the risk peaked in the first weeks of November. Over different time points, some parts of the county exhibited constant COVID-19 high-risk levels. Among these high-risk areas, many of them were majority-Hispanic neighborhoods in Chicago (i.e., Chicago west side) and Cook County suburbs (i.e., Franklin Park and Elgin). The model summary shows that the percentage of Hispanic population, health insurance coverage, and public transit commuters were associated with COVID-19 incidence. The posterior median and the 95% credible interval for the relative risk of a 1% increase in the percentage of Hispanic population was 1.009 (1.007, 1.011), indicating that a 1% increase in the percentage of Hispanic population corresponds to an increase in COVID-19 risk of 0.9%. The corresponding relative risk for a 1% increase in health insurance was 1.015 (1.006, 1.025), while for a 1% increase in the percentage of public transit commuters, the relative risk was 0.991 (0.987, 0.995). This study's findings highlight the importance of integrating the geographical information system into disease routine surveillance programs and transforming routinely collected health data into critical information. This information can be used to identify risk factors that could be addressed by allocating resources or implementing health policies.