COVID-19 testing is imperfect. Just like testing for any other medical ailment, there is a chance you’ll get a false positive or a false negative when testing for COVID-19 antibodies or a current infection.
Coronavirus antibody tests look for signs of past COVID-19 infections, and diagnostic or screening tests look for current infections.
Ellen Peters, director of the University of Oregon’s Center for Science Communication Research, said it would be “amazing” if people could rely just on COVID-19 antibody test results — for example, to tell if we are likely to become infected by the virus in the future.
“Tests are imperfect,” Peters said. “And so, when you test positive, you just don’t know for sure what you think you know for sure.”
For example, Oregon head football coach Mario Cristobal announced Tuesday that five positive COVID-19 cases among players turned out to be false positives.
COVID-19 testing accuracy varies, according to Harvard Health Publishing. For one type of diagnostic testing — including qPCR tests, those used by University Health Services, for example — false negative rates have ranged from 2% to 37% and false positives “should be close to zero,” according to the Sept. 30 article.
On Oct. 7, Peters presented during an Ideas on Tap webinar hosted by the Museum of Natural and Cultural History. Peters used Bayes’ theorem, a mathematical formula used for probability, to determine whether her friend, Doug, should have a “get out of COVID-jail free card.”
Doug tested positive for COVID-19 antibodies, Peters said during the presentation, and he wanted to know if this meant he was “immune” to the virus and could finally travel. Peters used the word immune, she said, to say that Doug had tested positive for antibodies consistent with the virus.
Bayes’ theorem considers a test result, but also considers the likelihood the test was wrong and the prevalence of the disease in people similar to Doug. Testing is only one piece of information you need, she said. For example, it’s important to consider the rate of COVID-19 in Doug’s area.
Using the equation and additional information, Peters found that — even though Doug had tested positive for antibodies — his likelihood of actually having antibodies and being “immune” was only 49%.
“It’s like flipping a coin whether or not he’s immune, despite the fact that he tested positive,” Peters said during the presentation.
In an interview, Peters said she wanted to convey two points during that talk: tests are imperfect and it’s important to use numbers to make good decisions.
It’s human nature to rely on feelings and emotions rather than data when making decisions, Peters said. College students and others are all tired of the virus, she said, and of the restrictions that come with it.
Similarly, most people have not experienced the virus. After months of worrying but not getting it, it’s natural for fears to start going away and for the lure of activities we’ve been missing out on to come back, Peters said.
“It’s this weird situation where experience is informing us that this isn’t a big deal, but it is,” she said. “So it’s like we all kind of have to keep in mind that that risk is still out there.”
UO senior Alexey Didenko has gotten tested for COVID-19 twice in the past few months and gotten negative results both times.
Even months into the pandemic, Didenko said he goes between feeling fine about coronavirus restrictions and feeling disheartened by the effects on businesses and friends.
“I'm still aware of the risks but there is less fear and more just struggling to adapt to the new reality,” Didenko said in an email.
The human reactions to the virus are understandable, Peters said.
“It’s just that sometimes in life we have to step back a minute and look at the bigger picture of not just ourselves but other people,” she said.
Peters recommends that people always take their COVID-19 test results with a grain of salt.
“If there’s some chance that you actually think you’re infected, quarantine yourself,” she said. “Keep people safe.”
Peters said people should start quarantining the moment they think they need to get tested for COVID-19. If the result ends up positive, individuals could have been contagious and unintentionally spread the virus.
If you really want to know for sure about your results, get a second test, she said. Getting the same result twice is a good indication of it being correct.
Didenko said he would feel more comfortable trusting COVID-19 testing if it was faster and more accurate. People might be walking around for days after a test before they know if they’re actually sick, he said.
“Sure they tell us to quarantine, but do you really think everyone does that?” Didenko said.
During the pandemic, Peters and those in her lab have tracked a national panel of Americans to study how emotions toward something like the coronavirus guide the public’s thoughts and intentions toward it.
“People’s fears about the coronavirus end up informing their risk perceptions,” she said.