The number of coronavirus infections in many parts of the United States is more than 10 times greater than the reported rate, according to data released Friday by the Centers for Disease Control and Prevention.
The analysis is part of a comprehensive set of surveys initiated by CDC to estimate how widely the virus has spread. Similar studies, sponsored by universities, national governments, and the World Health Organization, continue worldwide.
The CDC study found, for example, that in South Florida just under 2% of the population had been exposed to the virus as of April 10, but the proportion is likely to be higher now due to increased infections. in the state. The prevalence was highest in New York City, at nearly 7% as of April 1.
“This study underscores that there are probably many people infected without knowing it, probably because they have a mild or asymptomatic infection,” said Dr. Fiona Havers, who led the CDC study. “But those people could still infect others.”
She emphasized the importance of washing hands, wearing cloth masks, and social distancing to stop the spread of the virus from people without symptoms.
The numbers indicate that even in areas affected by the virus, an overwhelming majority of people have not yet been infected, said Scott Hensley, a viral immunologist at the University of Pennsylvania who was not involved in the research.
“Many of us are sitting ducks that are still susceptible to second waves,” he said.
The difference between reported infections and actual prevalence in the data was greatest in Missouri, where about 2.65% of the population was infected with the virus as of April 26, although many people may not have felt sick. This number is approximately 24 times greater than the reported rate: almost 162,000 compared to the 6,800 believed to be already infected.
The results confirm what some scientists have warned for months: that without more extensive evidence, dozens of infected people go unnoticed and circulate the virus.
“Our politicians may say that our evidence is incredible, but the fact is that our evidence is inadequate,” Hensley said. “These are exactly the type of studies we need right now.”
CDC Director Dr. Robert Redfield hinted at this trend Thursday during a call with reporters.
“Our best estimate at the moment is for each reported case, there were actually 10 other infections,” said Redfield.
The source of his claim was unclear at the time. Subsequently, the CDC published the data on its website and in MedRxiv, a repository of scientific results that have not yet been reviewed by a peer review.
CDC researchers analyzed samples from 11,933 people in six US regions during discrete periods from March 23 to May 3: Washington’s Puget Sound region, where the country’s first COVID-19 case was diagnosed. as well as New York City, South Florida, Missouri, Utah, and Connecticut.
The samples were collected in commercial laboratories from people who attended routine examinations, such as cholesterol tests, and the presence of antibodies against the virus was evaluated, which would indicate a previous infection even in the absence of symptoms.
The researchers then estimated the number of infections in each area. New York City, for example, reported 53,803 cases as of April 1, but the actual number of infections was 12 times higher, almost 642,000.
The city’s prevalence of 6.93% in the CDC study is well below the 21% estimated by the state survey in April. That number was based on people recruited from supermarkets, so the results would have been biased toward people who would go shopping during a pandemic: young people or those who had already had the virus and felt safe, experts said.
Havers also notes that when New York conducted that study, April 19-28, a jump in prevalence would be consistent with the increase in infections in the city at the time. She said CDC plans to repeat the surveys across all regions to see how prevalence changes over time. Supplemental CDC studies will test how well this approach captures true prevalence.
Saskia Popescu, an epidemiologist at the University of Arizona, said the CDC survey may also be biased by people with chronic conditions who are more likely to visit commercial laboratories. Still, it is more representative of the general population than other surveys because it included everyone who came to the labs for a variety of purposes, rather than limiting it to specific groups, such as healthcare workers who felt sick from the coronavirus. .
“A lot of the serology testing we’ve been looking at has really focused on that: people who thought they were exposed or felt sick at some point,” he said. “This approach is much more representative, ultimately.”
He also praised the researchers for not drawing inferences from the participants’ immune status study, because it is still unclear how the presence of antibodies relates to protection against the virus.
The analysis also highlights the large disparities between different parts of the country, and the importance not only of sufficient testing but also of laboratory capacity, Popescu said. In Arizona, he added, the delay is taking test results by five to six days.
Hensley said he was concerned that New York and other northeastern states might falsely believe they had passed the danger point and would reopen too soon.
“We need to turn south to see what a debacle things have been there,” he said. “If we open like Florida or Texas did, you can almost bet that we will be in the same position they are in now.”