NEW YORK — A recent study has revealed that the initial death toll from the Covid-19 pandemic in the U.S. was significantly underreported, highlighting substantial disparities in the number of uncounted deaths.
During 2020 and 2021, approximately 840,000 Covid-19 deaths were recorded on death certificates. However, researchers employing artificial intelligence techniques estimate that there were an additional 155,000 unrecognized deaths during that period, primarily occurring outside of hospitals. This suggests that roughly 16% of Covid-19 deaths were not counted.
The findings, published Wednesday in the journal Science Advances, align with estimates from other studies regarding pandemic deaths during that period. The authors focused on identifying which specific deaths were likely omitted from official records.
It was found that the unaccounted deaths were predominantly among Hispanic individuals and people of color, who died in the early months of the pandemic in states such as Alabama, Oklahoma, and South Carolina in the South and Southwest.
Steven Woolf, a researcher at Virginia Commonwealth University who was not part of the study, commented, “People on the margins continue to die at disproportionate rates because they can’t access care,” in an email. Six years after the coronavirus initially spread in the U.S., these barriers remain for many of the same communities.
Access to care wasn’t the only challenge
Elizabeth Wrigley-Field from the University of Minnesota, one of the study’s authors, noted that while hospital patients were routinely tested for Covid-19, many who fell ill and died outside of hospitals did not have access to testing, often due to the lack of readily available at-home testing early in the pandemic.
In some regions, death investigations are conducted by elected coroners who may lack the specialized training of medical examiners. Research suggests that political views may have influenced whether individuals or their families sought Covid-19 testing and whether coroners pursued postmortem testing. Some coroners reported being pressured by families not to cite Covid-19 as a cause of death.
Andrew Stokes of Boston University, the senior author of the study, mentioned, “Our antiquated death investigation system is one key reason why we fell short of accurate counts, particularly outside of big metropolitan areas.”
Death counts were swept up in Covid politics
The Centers for Disease Control and Prevention has recorded over 1.2 million Covid-19 deaths since the pandemic began in early 2020, with more than two-thirds of these occurring in 2020 and 2021.
The reported figures have been subject to debate, as misinformation on social media claimed the death toll was exaggerated. President Donald Trump contributed to this controversy by retweeting a post in August 2020, which falsely stated that only 6% of reported deaths were due to Covid-19, a post that Twitter later removed.
Other types of pandemic-related deaths included those of uninfected individuals who could not receive care for other medical conditions due to overwhelmed hospitals, as well as drug overdoses resulting from social isolation and lack of treatment access. Studies estimating the true number of pandemic deaths have considered these factors as well.
Stokes and his team focused on fatalities directly caused by the coronavirus. They utilized machine learning to analyze the death certificates of infected patients who died in hospitals and used patterns from these records to assess certificates of those who died outside hospitals, with deaths attributed to conditions like pneumonia or diabetes.
The scientific community is still learning about the strengths and limitations of research relying on machine learning, but Woolf found the team’s approach “intriguing.”
— Mike Stobbe
