COVID-19 Community Incidence and Associated Neighborhood-Level Characteristics in Houston, Texas, USA

Abstract

Central to developing effective control measures for the COVID-19 pandemic is understanding the epidemiology of transmission in the community. Geospatial analysis of neighborhood-level data could provide insight into drivers of infection. In the current analysis of Harris County, Texas, we used custom interpolation tools in GIS to disaggregate COVID-19 incidence estimates from the zip code to census tract estimates—a better representation of neighborhood-level estimates. We assessed the associations between 29 neighborhood-level characteristics and COVID-19 incidence using a series of aspatial and spatial models. The variables that maintained significant and positive associations with COVID-19 incidence in our final aspatial model and later represented in a geographically weighted regression model were the percentage of the Black/African American population, percentage of the foreign-born population, area derivation index (ADI), percentage of households with no vehicle, and percentage of people over 65 years old inside each census tract. Conversely, we observed negative and significant association with the percentage employed in education. Notably, the spatial models indicated that the impact of ADI was homogeneous across the study area, but other risk factors varied by neighborhood. The current findings could enhance decision making by local public health officials in responding to the COVID-19 pandemic. By understanding factors that drive community transmission, we can better target disease control measures.

Census Tract Patterns and Contextual Social Determinants of Health Associated With COVID-19 in a Hispanic Population From South Texas: A Spatiotemporal Perspective

“…Moreover, the high variability of SDOH within a county population was not able to accurately examine the impact of SDOH on COVID-19 disparities in populations [ 5 ]. Studies that investigate the SDOH and COVID-19 incidence and mortality at a geographical scale smaller than the US county are limited [ 7 , 8 ]. The lack of studies on a granular spatial scale is largely due to insufficiently detailed COVID-19 surveillance data, particularly data that are publicly available.…”

Examining Social Vulnerability and the Association With COVID-19 Incidence in Harris County, Texas

“…A negative binomial regression (NBR) model was used to analyze variations in SARS-CoV-2 incidence counts across census tracts. This model was used to handle the overdispersion of count-based data containing only non-negative integer values and consider the independent explanatory neighborhood-level characteristics ( 15 , 21 ). Since case count data was used instead of incident rates, all models controlled for the population size.…”

“…A growing number of studies have investigated the association between social vulnerability and SARS-CoV-2 incidence (13)(14)(15)(16). However, these studies have four important limitations: (1) few studies examine smaller geographic units such as census tracts;…”

Results of COVID-19 Surveillance in a Large United States Pediatric Healthcare System over One Year

“…Reports by geographic locations indicate that pediatric COVID-19 cases are substantially greater in economically disadvantaged census tracts [ 5 , 6 , 7 ]. One of the main challenges, however, has been the absence of a denominator to estimate true incidence of COVID-19 in the pediatric population, in either community or healthcare based settings.…”