Empiric evidence of ethnic disparities in coronavirus positivity in Washington State

Abstract

ABSTRACT Objectives Early reports from the initial months of the coronavirus pandemic reveal ethnic disparities in coronavirus incidence, severity, and mortality. This study aimed to evaluate the relationship between ethnicity and outcomes of coronavirus positivity and hospitalization. Design An observational cohort study using electronic health record (EHR) data from a large community healthcare system in Washington State across the first phase of the pandemic (March 5 – June 7, 2020). Results A total of 18,667 patients (65.9% of all tested) with EHR-documented ethnicity were included. Overall, 6.4% of patients tested positive for coronavirus. Among Latinx patients, 18.6% of those tested were positive, compared to only 4.0% of tested White patients. Multivariable logistic regression revealed significantly higher odds of positivity for Latinxs (aOR\u2009=\u20094.96, 95% CI 4.19–5.87), Asians (aOR\u2009=\u20092.33, 95% CI 1.74–3.08), Blacks (aOR\u2009=\u20091.82, 95% CI 1.43–2.31), and members of other ethnic minority groups (aOR\u2009=\u20092.34, 95% CI 1.80–2.95), compared to Whites in models adjusting for relevant confounders. Latinxs had a higher percentage of self-pay insurance (22.2%) compared to other ethnic groups (7.9–15.8%) and, among those who tested positive, were the only ethnic subpopulation with significantly higher odds than Whites to be hospitalized for COVID-19 (aOR\u2009=\u20092.19, 95% CI 1.45–3.33). We observed a positive correlation between infection and the percentage of Latinxs (r\u2009=\u20090.61, 95% CI 0.45–0.74), Blacks (r\u2009=\u20090.51, 95% CI 0.32–0.66), or Asians (r\u2009=\u20090.64, 95% CI 0.49–0.76) in a given zip-code. This correlationwas negative for Whites (r\u2009=\u2009−0.63, 95% CI −0.75, −0.45). Conclusions We present empirical evidence of higher rates of coronavirus positivity among People of Color compared to White people in Washington State. Social determinants of health, such as occupation, housing, healthcare access, and community structure, may contribute to health disparities in the coronavirus pandemic. Targeted capture of these variables in electronic health records is warranted to inform health equity analyses.