Excess Mortality Rates Early in Pandemic Highest Among Blacks
The impact of the COVID-19 pandemic has been starkly uneven across race, ethnicity and geography, according to a new study led by SHP's Maria Polyakova.
Unusually high numbers of racial minorities and people of Hispanic origin nationwide died of all causes in the early days of the pandemic, according to a new study led by SHP's Maria Polyakova.
After adjusting for age, sex and state of residence, researchers found that an additional 6.8 per 10,000 Black people died of all causes last April, compared with the average number who died since 2011 in April. For Hispanics, that figure was 4.3; for Asians, 2.7; and for whites, 1.5.
Excess mortality rates were greater in some areas than others.
“The overall rates of excess mortality and differences by race and ethnicity in New York and New Jersey are absolutely staggering,” said Polyakova, PhD, assistant professor of medicine at Stanford Health Policy and a co-author of the study, published Feb. 1 in Health Affairs.
In New York and New Jersey, the study estimates that 30 additional Black people per 10,000 died last April compared with the 10-year average for that month. Hispanic people in the two states fared only slightly better, with an estimated 27.2 more deaths per 10,000 in New York and 20.5 more per 10,000 in New Jersey. By contrast, the estimated rates of white excess mortality were much lower: 7.1 per 10,000 in New York and 8.6 per 10,000 in New Jersey.
The study is the first to use data from the Social Security Administration and Census Bureau for the entire U.S. population. The April 2020 database comprised information on 241.5 million people nationwide ages 11 to 99 and registered 276,000 deaths of all causes.
Polyakova, an economist who is also a faculty fellow at the Stanford Institute for Economic Policy Research, said the study has several policy-relevant takeaways:
* Excess mortality associated with the pandemic is higher in the nonwhite population in general, so health officials may want to take these differences into account when planning vaccine distribution.
* Excess mortality in minority populations was found in some states even before there was a spike in COVID-19 infections. This could imply that the indirect effects of the pandemic on mortality were important.
* Several hypotheses about the underlying causes of racial and ethnic disparities in the mortality impact of the virus have to do with occupations, education, residential density, food availability, pollution, access to health care and comorbidities. Understanding which one played the key role could help inform policymaking.
* Public health officials need to better understand why the impact of the pandemic varied by geography, as that may shed light on the drivers of health disparities.
Geographic Variations
The researchers found that nationwide averages mask the substantial geographic variations in excess mortality in April 2020. In Wisconsin, for example, they estimate statistically insignificant white excess mortality — 0.27 per 10,000 — but significant Black excess mortality — 4.6 per 10,000 — and Hispanic excess mortality of 1.4 per 10,000.
“Further work understanding the causes of geographic variation in racial and ethnic disparities — the relevant roles of social and environmental factors relative to comorbidities and of the direct and indirect health effects of the pandemic — is crucial for effective policymaking,” the authors wrote.
Polyakova is now working on a study of the specific economic factors that may have played a role in these mortality disparities. She said mortality spikes among people of color could have been driven by economic distress and higher economic vulnerability, leading to more rapid loss in access to health care, in addition to the direct effects of COVID-19 infections.
Other co-authors of the study are U.S. Census Bureau research economists Victoria Udalova, PhD, Katie Genadek, PhD, and Keith Finlay, PhD; Amy Finkelstein, PhD, professor of economics at MIT; and Geoffrey Kocks, a graduate student in economics at MIT.
Funding for the work was provided by the National Institute on Aging (grant R01-AG032449) and the National Institute for Health Care Management.
