Inclusion of Race and Ethnicity With Neighborhood Socioeconomic Deprivation When Assessing COVID-19 Hospitalization Risk Among California Veterans Health Administration Users

Key Points Question How is exclusion of racial and ethnic segregation from California's Healthy Places Index (HPI), a composite measure of the neighborhood environment, associated with race and ethnicity-stratified COVID-19-related hospitalization? Findings In this cohort study of 19495 veterans with COVID-19 residing in California, COVID-19-related hospitalization was higher among Black and White veterans living in neighborhoods with lower HPI scores. However, accounting for Black segregation eliminated HPI's hospitalization association for White veterans, but not for Black veterans. Meaning These findings suggest that care should be taken when using composite neighborhood deprivation indices that do not account for racial and ethnic segregation, as these associations may differ by race and ethnicity. This cohort study examines the associations among the Healthy Places Index, Black and Hispanic segregation, and the Social Vulnerability Index with COVID-19-related hospitalization by race and ethnicity group among veterans with COVID-19 in California. Importance Despite complexities of racial and ethnic residential segregation (hereinafter referred to as segregation) and neighborhood socioeconomic deprivation, public health studies, including those on COVID-19 racial and ethnic disparities, often rely on composite neighborhood indices that do not account for residential segregation. Objective To examine the associations by race and ethnicity among California's Healthy Places Index (HPI), Black and Hispanic segregation, Social Vulnerability Index (SVI), and COVID-19-related hospitalization. Design, Setting, and Participants This cohort study included veterans with positive test results for COVID-19 living in California who used Veterans Health Administration services between March 1, 2020, and October 31, 2021. Main Outcomes and Measures Rates of COVID-19-related hospitalization among veterans with COVID-19. Results The sample available for analysis included 19 495 veterans with COVID-19 (mean [SD] age, 57.21 [17.68] years), of whom 91.0% were men, 27.7% were Hispanic, 16.1% were non-Hispanic Black, and 45.0% were non-Hispanic White. For Black veterans, living in lower-HPI (ie, less healthy) neighborhoods was associated with higher rates of hospitalization (odds ratio [OR],1.07 [95% CI, 1.03-1.12]), even after accounting for Black segregation (OR, 1.06 [95% CI, 1.02-1.11]). Among Hispanic veterans, living in lower-HPI neighborhoods was not associated with hospitalization with (OR, 1.04 [95% CI, 0.99-1.09]) and without (OR, 1.03 [95% CI, 1.00-1.08]) Hispanic segregation adjustment. For non-Hispanic White veterans, lower HPI was associated with more frequent hospitalization (OR, 1.03 [95% CI, 1.00-1.06]). The HPI was no longer associated with hospitalization after accounting for Black (OR, 1.02 [95% CI, 0.99-1.05]) or Hispanic (OR, 0.98 [95% CI, 0.95-1.02]) segregation. Hospitalization was higher for White (OR, 4.42 [95% CI, 1.62-12.08]) and Hispanic (OR, 2.90 [95% CI, 1.02-8.23]) veterans living in neighborhoods with greater Black segregation and for White veterans in more Hispanic-segregated neighborhoods (OR, 2.81 [95% CI, 1.96-4.03]), adjusting for HPI. Living in higher SVI (ie, more vulnerable) neighborhoods was associated with greater hospitalization for Black (OR, 1.06 [95% CI, 1.02-1.10]) and non-Hispanic White (OR, 1.04 [95% CI, 1.01-1.06]) veterans. Conclusions and Relevance In this cohort study of US veterans with COVID-19, HPI captured neighborhood-level risk for COVID-19-related hospitalization for Black, Hispanic, and White veterans comparably with SVI. These findings have implications for the use of HPI and other composite neighborhood deprivation indices that do not explicitly account for segregation. Understanding associations between place and health requires ensuring composite measures accurately account for multiple aspects of neighborhood deprivation and, importantly, variation by race and ethnicity.