Exacerbation of COVID-19 mortality by the fragmented United States healthcare system: A retrospective observational study

BACKGROUND: Before widespread vaccination, the United States was disproportionately affected by COVID-19 with a mortality rate several times that of other affluent societies. Comparing regions with different rates of health insurance, we assess how much of this excess mortality may be due to the relatively large population without health insurance. METHODS: We use daily surveillance data from the US Centers for Disease Control and Prevention (CDC) stratified by region, age group, gender, and race in regression analysis of daily COVID-19 cases, hospitalization, and mortality. COVID-19 data have been matched with structural characteristics of the region including average proportion with health insurance. As checks, we have estimated regressions for different time periods, different groups of states, and by comparing adjacent counties between states with and without Medicaid expansion. FINDINGS: Groups with lower health insurance coverage had significantly higher mortality as well as greater case counts and hospitalization. Early in the pandemic, they were also less likely to be tested for COVID-19. Applying our regression estimates, we estimate that had there been full health insurance coverage of the population, there would have been 60,000 fewer deaths, 26% of the total death toll in the period of this analysis. INTERPRETATION: Our study demonstrates that a significant share of COVID-19 mortality in the United States, and much of the excess mortality in the United States compared with other countries, is due to our reliance on a system of market-driven healthcare. Providing universal insurance coverage should be part of our campaign to reduce COVID-19 mortality. It also suggests that these concerns should not be restricted to COVID-19 but apply across all diseases, contributing to many unnecessary deaths in the United States each year even apart from the COVID-19 pandemic. FUNDING: This study was supported by grants from the NSF (Expeditions grant 1918784), and the NIH (1R01AI151176-01 and 5K01AI141576).