AI Study Uncovers 155,000 Uncounted COVID-19 Deaths in U.S., Exposing Stark Racial Disparities
A groundbreaking new study published in the journal Science Advances has revealed that the early death toll of the COVID-19 pandemic in the United States was significantly higher than official counts suggest. Using artificial intelligence, researchers estimate that as many as 155,000 additional deaths likely occurred outside of hospitals in 2020 and 2021, representing approximately 16% of total COVID-19 fatalities during that period.
Disproportionate Impact on Marginalized Communities
The study highlights dramatic disparities in the uncounted deaths, with the undiagnosed dead more likely to be Hispanic individuals and other people of color. These deaths predominantly occurred in the first few months of the pandemic and were concentrated in states across the South and Southwest, including Alabama, Oklahoma, and South Carolina.
Steven Woolf, a researcher at Virginia Commonwealth University who was not involved in the study, emphasized the ongoing challenges faced by marginalized populations. "People on the margins continue to die at disproportionate rates because they can't access care," he stated in an email, underscoring the persistent inequities in healthcare access.
Barriers to Accurate Death Counts
Several factors contributed to the undercounting of COVID-19 deaths. While hospital patients were routinely tested for the virus, many individuals who grew sick and died outside of hospitals were not tested, often due to the limited availability of at-home testing early in the pandemic, according to study author Elizabeth Wrigley-Field of the University of Minnesota.
Additionally, the study points to flaws in the death investigation system. In some regions, elected coroners without specialized training handle death investigations, which may have affected the accuracy of cause-of-death determinations. Partisan opinions and family pressures also played a role, with some coroners reporting that families requested COVID-19 not be listed as a cause of death.
"Our antiquated death investigation system is one key reason why we fell short of accurate counts, particularly outside of big metropolitan areas," said Andrew Stokes of Boston University, the senior author of the paper.
Political Controversies and Broader Pandemic Impacts
The official death count, which exceeds 1.2 million COVID-19 deaths since early 2020, has long been a subject of debate. False claims on social media and political rhetoric, including statements from former President Donald Trump, fueled misinformation about the accuracy of the numbers.
It is important to note that the pandemic also led to other types of deaths, such as those from overdoses due to social isolation and loss of treatment access, or from untreated medical conditions in overwhelmed hospitals. However, this study specifically focused on deaths directly caused by coronavirus infections.
Methodology and Future Implications
The researchers employed machine learning to analyze death certificates, using patterns from hospital deaths to evaluate non-hospital fatalities attributed to conditions like pneumonia or diabetes. While the use of AI in such research is still evolving, Woolf described the team's approach as "intriguing" and potentially valuable for improving public health data accuracy.
This study underscores the critical need for enhanced death investigation systems and equitable healthcare access to better prepare for future public health crises.
