Stay-at-Home Orders, African American Population, Poverty and State-level Covid-19 Infections: Are there associations?

Importance: To cope with the continuing COVID-19 pandemic, state and local health officials need information on the effectiveness of policies aimed at curbing contagion, as well as area-specific socio-demographic characteristics that can portend vulnerability to the disease. Objective: To investigate whether state-imposed stay-at-home orders, African American population in the state, state poverty and other state socio-demographic characteristics, were associated with the state-level incidence of COVID-19 infection. Design, Setting, Participants: State-level, aggregated, publicly available data on positive COVID-19 cases and tests were used. The period considered was March 1st -May 4th. All U.S. states except Washington were included. Outcomes of interest were daily cumulative and daily incremental COVID-19 infection rates. Outcomes were log-transformed. Log-linear regression models with a quadratic time-trend and random intercepts for states were estimated. Covariates included log-transformed test-rates, a binary indicator for stay-at-home, percentage of African American, poverty, percentage elderly, state population and prevalence of selected comorbidities. Binary fixed effects for date each state first started reporting test data were included. Results: Stay-at-home orders were associated with decreases in cumulative (β:-1.23; T-stat: -6.84) and daily (β:-0.46; T-stat: -2.56) infection-rates. Predictive analyses indicated that expected cumulative infection rates would be 3 times higher and expected daily incremental rates about 60% higher in absence of stay-at-home orders. Higher African American population was associated with higher cumulative (β: 0.08; T-stat: 4.01) and daily (β: 0.06; T-stat: 3.50) rates. State poverty rates had mixed results and were not robust to model specifications. There was strong evidence of a quadratic daily trend for cumulative and daily rates. Results were largely robust to alternate specifications. Conclusions: We find evidence that stay-at-home orders, which were widely supported by public-health experts, helped to substantially curb COVID-19 infection-rates. As we move to a phased re-opening, continued precautions advised by public-health experts should be adhered to. Also, a larger African American population is strongly associated with incidence of COVID-19 infection. Policies and resources to help mitigate African American vulnerability to COVID-19 is an urgent public health and social justice issue, especially since the ongoing mass protests against police brutality may further exacerbate COVID-19 contagion in this community.