Background: In December 2021, Paxlovid received FDA emergency use authorization for treating patients with mild-to-moderate COVID-19 who are at high risk for progression to severe COVID-19. Paxlovid has become a preferred treatment for eligible COVID-19 patients. However, there is lack of research assessing whether racial disparities exist in Paxlovid use in the US after controlling for confounding.
Objectives: To assess the association between Paxlovid use (dispensing as a proxy) and race or ethnicity with and without confounding adjustment and explore the impact of other determinants of health (DOH) on the use.
Methods: A retrospective cohort study was conducted using US Optum Claims with Social Economic Status data (12/22/2021 - 11/15/2022). Patients aged ≥12 years having a COVID-19 diagnosis or Paxlovid dispensed and with ≥365 days of continuous enrollment prior to COVID-19 diagnosis or Paxlovid dispensing date were stratified into 2 groups: Paxlovid vs. Non-Paxlovid users. Race or ethnicity were classified as: White, African American (AA), Asian, Hispanic, and Unknown. Baseline (BL) covariates, which encompassed DOH, included demographic (age, sex, region, insurance), social economic (SE: education, housing status, household income), clinical characteristics (17 comorbidities, COVID-19 vaccination (≥1 dose), past COVID-19 infection), and healthcare utilization (HCU). Four logistic regression models were used to estimate Paxlovid use across racial or ethnic groups and assess the relationship between other DOH and Paxlovid use: Model 1 (race or ethnicity unadjusted), Model 2 (M1 plus Demographics), Model 3 (M2 plus SE), Model 4 (M3 plus comorbidities and HCU).
Results: A total of 276,190 Paxlovid and 909,322 Non-Paxlovid patients were identified. Comparing with AA patients, the unadjusted (M1) and fully adjusted (M4) Odds Ratios and 95% CI were: White: 1.54 [1.52,1.57], 1.20 [1.18, 1.22]; Asian: 1.79 [1.74,1.84], 1.30 [1.26, 1.34]; and Hispanic: 1.16 [1.14, 1.18], 1.18 [1.16, 1.21]. Overall, patients who were older, male, had higher education, income, home ownership, comorbidities, more physician office visits, and received COVID-19 vaccine were associated with more Paxlovid use while patients with previous COVID-19 infections and more hospitalization and ER visits at BL were less likely to use Paxlovid (M4).
Conclusions: This is the first study using large real-world data to assess racial disparity and DOH in Paxlovid use in the US controlling for known confounding. Demographic and social economic factors had strong impact on the disparity in use. Racial disparities in Paxlovid use exist: White and Asian had ≥20% more use than AA after full adjustment. Hispanics also had more use than AA even after adjusting for other DOH, which warrants further research.
