The COVID-19 pandemic has been marked by its disparate impacts on residents in lower income neighborhoods. To better understand reasons for disparities in disease severity, we asked: To what extent does the neighborhood built environment independently predict COVID-19 hospitalizations for people with SARS-CoV-2?
Retrospective analysis of all cases of SARS-CoV-2 diagnosed within the University of Colorado Health and Denver Health healthcare systems from 3/1/2020 – 12/15/2020. Electronic health records were queried for positive SARS-COV-2 PCR results for individuals aged ?18 years during the study period. Demographic and health variables were extracted. Home addresses were matched with social vulnerability indices and built environment variables including population density and crowding, environmental hazards and amenities, and mobility options. Logistic regression was used to identify factors of the neighborhood built environment associated with hospitalization after a positive SARS-CoV-2 result.
Among the two systems, 39,304 individuals had positive SARS-CoV-2 tests; 14,604 (37.16%) had full demographic data; 4,101 (28%) were hospitalized. Odds of hospitalization were higher for individuals living in apartments and in census blocks with lower residential density and higher percentage of multi-family housing units. See Table. Higher particulate matter (PM2.5) levels were associated with higher odds of being hospitalized but living within ¼ mile of a highway was not. Living within 1/2 mile of a park and more park acreage in the neighborhood were associated with lower odds of hospitalization. Odds of being hospitalized were higher for individuals in neighborhoods with a lower Walk Score ®, lower Bike Score ®, and higher Transit Score ®. Effects were more pronounced for Latinx individuals.
Among those with SARS-CoV-2 infection in Denver, living in areas with high levels of PM2.5, less park access, and lower Walk® and Bike Scores® were found to be independent risk factors for hospitalization when controlling for income and medical comorbidities.