The US may have seen 8.7 million coronavirus cases in March alone, new research shows — meaning millions went undiagnosed

The US may never know the true tally of its coronavirus infections but scientists are getting closer to an accurate count. New research suggests that many millions of infections went undiagnosed in March, even as testing ramped up.

The nation's current case count hovers at around 2.4 million. But a study published Monday in the journal Science estimated that the US saw 8.7 million coronavirus cases from March 8 to 28. The country's official tally indicates there were around 100,000 infections during that time period, suggesting that the real-life outbreak was more than 80 times larger than the one reported.

To arrive at this estimate, the researchers looked at reports of flu-like illnesses among outpatients using the CDC's influenza surveillance system . Many states, including Colorado, Louisiana, Maryland, New Jersey, New York, Oregon, Pennsylvania, and Washington, reported a surge of illnesses in March around 2.8 million that weren't attributable to the flu or seasonal respiratory diseases.

The researchers think most of those were coronavirus cases that went undiagnosed due to limited testing capacity or false-negative test results. They also estimated that only one-third of coronavirus patients in the US sought medical care in the first place, since many coronavirus cases are asymptomatic or mild. So the researchers suggest the actual surge of coronavirus cases was likely three times larger than those extra cases of flu-like illness: 8.7 million cases.

The findings support what epidemiologists have long suspected: that the coronavirus is far more prevalent and less deadly than current figures suggest.

Around 5% of US coronavirus cases reported so far have been been fatal. Accounting for a surge of undiagnosed cases in March could lower that fatality rate to between 0.07% and 1.4%, the researchers found. Recent estimates suggest that the global fatality rate could be around 1%.

US infections may have doubled every 3 days from January to March

Previous models predicted that the actual number of US coronavirus cases could be anywhere from 5 to 20 times the current number. The new study suggests that estimate isn't far off.

By the end of March, the researchers estimated, the US had confirmed less than 13% of that month's "flu-like illnesses" as coronavirus infections. Many of these undiagnosed cases were located in New York, the original epicenter of the US outbreak.

More than 8% of New York's population or 1.6 million people had been infected by March 28, the researchers found. (The state's antibody testing later determined that nearly 14% of New Yorkers, or 2.7 million people, had been infected by March 29.) But data at the time suggested only 0.3% of the state, or 58,000 people, had been infected.

That means New York may have failed to diagnose up to 2.6 million cases in March.

New York coronavirus subway Alexi Rosenfeld/Getty Images

"At first I couldn't believe our estimates were correct," Justin Silverman, an assistant professor at Penn State University who co-authored the study, said in a statement.

But the researchers cross-referenced their model with the rise of coronavirus deaths in the US. Using data from the New York Times, they found that deaths were doubling every three days in March. That same doubling rate explained the rise of coronavirus cases from a single infection in January to 8.7 million infections in March (assuming that the majority of cases didn't seek medical care or get official diagnoses).

By the third week of March, the study found, coronavirus infections identified as "flu-like illnesses" reached their peak nationally. Many states then saw a decline in these infections over the following week. That could mean that increased testing or social distancing measures worked or that fewer people presented themselves to healthcare providers.

But infections continued in New York and New Jersey.

"Our results suggest that the overwhelming effects of COVID-19 may have less to do with the virus' lethality and more to do with how quickly it was able to spread through communities initially," Silverman said. "A lower fatality rate, coupled with a higher prevalence of disease and rapid growth of regional epidemics, provides an alternative explanation of the large number of deaths and overcrowding of hospitals we have seen in certain areas of the world."

NOW WATCH: How we know the COVID-19 coronavirus wasn't made in a lab