Almost 17 million U.S. coronavirus cases were not detected during first half of 2020, study led by UMBC graduate finds

For every documented COVID-19 infection during the first six months of the pandemic last year, five cases slipped by undiagnosed — roughly 16.8 million — according to a federal study led by a University of Maryland, Baltimore County graduate.

The results of the National Institutes of Health study led by immunologist Kaitlyn Sadtler “suggest a much larger spread of the [COVID-19] pandemic than originally thought,” wrote Sadtler, who heads the section for immunoengineering at NIH’s National Institute of Biomedical Imaging and Bioengineering. The study reviewed data from January through July of last year as the pandemic ramped up across the country.


The study also indicates that Black and Hispanic communities were hit hardest by the virus from its outset.

The 16.8 million figure falls in line with estimates from the Centers for Disease Control and Prevention regarding undiagnosed Americans, which said just one in 4.6 cases were reported in the first half of 2020.


“I don’t want to say it surprised me,” Sadtler said, but “putting that number in front of you and knowing that it’s based on laboratory research … I can trust this number. That says a lot.”

Solid numbers on the true toll of the coronavirus have eluded health officials because of the shoddy rollout of diagnostic coronavirus testing across the U.S. during the first wave of infections and the prevalence of asymptomatic carriers who may not know to get tested, said Gigi Gronvall, an immunologist and professor at Johns Hopkins Bloomberg School of Public Health.

“This type of study is really important for public health decision-making,” Gronvall said. “It’s really important for officials to know if there are particular groups that are not getting the resources they need, or care or preventative measures.”

Researchers analyzed blood samples sent by mail to Sadtler’s Bethesda lab from a representative sample of more than 9,000 volunteers who had never been diagnosed with COVID-19.

Using an enzyme-linked immunosorbent assay, a test that detects antibodies, researchers determined volunteers’ “seropositivity” — whether or not their blood serum contained anti-spike protein antibodies that appear in those who have carried the virus.

Of Black study participants, 14.2% were found to have natural coronavirus antibodies. That figure dropped to 6.8% among Native American and Pacific Islander volunteers and 6.1% among Hispanic participants.

And among white volunteers, the study showed 2.5% were undiagnosed carriers. Among Asians, it was 2%.

“Clearly there’s something — or in the case of this pandemic, many things — wrong with the way that we approach ... public health resources and medical care,” Gronvall said.


In addition to factors such as race, age, ethnicity, sex and location, researchers also gathered information on socioeconomic and health factors — data Sadtler wants to dig into more to confirm if higher coronavirus infection rates among people of color track along with those factors, which in the U.S. often differ across racial lines.

“If we’re only looking at white, suburban individuals in a certain area, that tells us nothing about our neighbors in a different state, or in the inner city,” Sadtler said.

Through funding from the National Cancer Institute’s Serological Sciences Network, Sadtler’s team is repeating the study, which has yet to be peer reviewed, with the same participants in the second half of 2020 and again throughout 2021.

In the next rounds, Sadtler wants to assess antibody duration to see how long infection-induced antibody immunity may last in asymptomatic cases.

Repeating the study also may address the uncertainty surrounding reinfection rates, but researchers will have to account for inoculation-induced antibodies and cross-check the original coronavirus variant detected in volunteers in the first half of 2020 with new coronavirus variants, Sadtler said.

Breaking News Alerts

Breaking News Alerts

As it happens

Be informed of breaking news as it happens and notified about other don't-miss content with our free news alerts.

“I’m really looking forward to the next time-point and the time-point after that and understanding the immunology behind it,” Sadtler said. “I mean that 16.8 million number; how does that not sit with you?”


Gronvall cautions against viewing the findings as an indicator of the country’s progress toward herd immunity, when the majority of a given population develops antibodies that prevent infection spread.

“Herd immunity has never been achieved without vaccination in modern times,” she said.

Natural coronavirus antibodies are detectable at least eight months after infection, Sadtler said. And there is evidence, she added, “that people with viral infection-induced antibodies have protection, though reinfections do occur sometimes.”

One of the goals, Sadtler said, was to create a protocol that could be easily repeated “in a very simple laboratory” in countries where researchers may face scarce resources and have a harder time getting financial backing.

Her model already is being replicated by researchers worldwide in Mali, Cambodia and several other countries.

“It’s a global pandemic,” she said. “So we have to think about it globally.”