IE 11 is not supported. For an optimal experience visit our site on another browser.

Rapid Covid tests give many false negatives, but that might mean you're not contagious

The tests seem to be most accurate when viral loads are high, so experts think a negative could mean you’re less likely to spread the virus.
A person processes a rapid at-home Covid-19 test in Easton, N.H., U.S., on Dec. 7, 2021.
A person processes a rapid at-home Covid test in Easton, N.H., on Dec. 7, 2021.John Tully / Bloomberg via Getty Images file

In early January, more than 700 Stanford University athletes took rapid Covid tests upon returning to the campus. Those who tested negative — and some who tested positive — were also given PCR tests.

The students were all participants in an experiment run by the Stanford University School of Medicine, which assessed the accuracy of BinaxNOW rapid tests.

The results, published Wednesday in JAMA Network Open, showed that the rapid tests caught 63% of positive cases, meaning they produced quite a few false negatives. The accuracy varied significantly depending on whether the infected athletes had symptoms. The tests caught 78% of symptomatic cases but 39% among asymptomatic athletes.

False positives weren't a problem though: The tests were accurate in nearly 100% of cases in which athletes did not have Covid.

"I do think that the viral load is the big issue here," said Dr. Calvin Hwang, the study’s lead author and a clinical assistant professor at Stanford. Viral load refers to the amount of virus in a person’s body.

"These asymptomatic patients or pre-symptomatic patients, their viral load may not be high enough to be seen on a rapid test, and thus you get the problem of false negatives," he said.

But Hwang and two other disease experts said rapid tests (also known as antigen tests) are good indicators of when a person might be contagious.

"Only the people shedding the most virus are going to be positive with a rapid test, but those are the people you especially want to identify because they’re the most infectious," said Dr. Sheldon Campbell, an associate professor of laboratory medicine at the Yale School of Medicine who wasn't involved in the research.

Dr. Michael Mina, the chief science officer for the at-home testing company eMed, said people who are negative on a rapid test might still be mildly infectious, but "they’re not going to be super-spreaders."

"They’re not going to go off into a locker room and infect 15 of their teammates," added Mina, a former Harvard University epidemiologist who was not involved in the new study.

More false negatives than false positives

When the Food and Drug Administration authorized BinaxNOW, which is made by Abbott, in December 2020, the company said the test picked up 92% of positives and 100% of negatives seven days or less after symptoms started.

But a study in May in San Francisco found that BinaxNOW detected 65% of positive cases during the initial omicron surge. The test did, however, pick up almost all cases among people with the highest viral loads.

"Abbott’s research, as well as third-party research, continue to demonstrate BinaxNOW’s ability to detect people with Covid, including omicron, when they are infectious and likely to spread the virus. The test shows performance of 95.6% positive agreement (sensitivity) in people seven days or less post-symptom onset with high viral loads," said John Koval, director of public affairs for Abbott’s rapid diagnostics business.

Rapid tests from other manufacturers have comparable accuracy levels to that of Abbott. An analysis in February of 10 commercially available rapid tests found similar sensitivities among the tests when they were used to detect the delta and the omicron variants.

"They all perform pretty similarly because they’re all using a similar technology," Hwang said.

Studies have also found rapid tests to be reliably good at identifying negative cases. Out of more than 900,000 rapid tests administered across Canada, just 0.05% produced false positives, according to a research letter in January.

"If you’re positive on an antigen and you’re symptomatic, case closed. You don’t even need to get a PCR," Hwang said.

Why do some people who likely have Covid consistently test negative?

Some people may have symptoms and known exposure to Covid but never test positive on either a rapid or a PCR test.

Experts said that's more likely to happen among vaccinated people, since Covid vaccines can prompt a swift immune response that stops the virus from replicating to detectable levels.

These people may still develop symptoms, Mina said: "There’s no free lunch, so when your immune system is working, you feel it and that causes symptoms."

In the Stanford study, the vast majority of athletes were vaccinated.

"This should be viewed, in many ways, as a success of your immune system," Mina said, referring to vaccinated people testing negative. "It’s been very bizarre to me that people keep seeing this as a failure of rapid tests."

When should you get a PCR test?

Abbott instructs people to test themselves twice over three days, 24 to 48 hours apart.

"If you test repeatedly, a rapid antigen test will detect you as soon as your viral load gets high enough that you actually become a risk to other people," Mina said.

He added that rapid tests are useful in answering a common question: "Am I safe to leave isolation?"

"If I’m still positive, I should just not leave isolation. If I become negative, then I’m probably safe," he said, estimating that rapid tests "generally turn negative within a 24-hour window or so of when people are no longer infectious."

Hwang said he’d still recommend a PCR test for people who have Covid symptoms but are negative on a rapid test, especially those who are immunocompromised.

"If you’re a person at risk for severe Covid disease, I don’t think you should stop with the antigen test, because we can treat you," Campbell said.

Do variants change test accuracy?

So far, omicron doesn't seem to have changed the accuracy of rapid tests, experts said, since the tests look for parts of the virus that are less prone to mutations.

But Hwang noted that because the athletes in his study got an early version of the omicron variant, BA.1, which is no longer dominant in the U.S., "there may need to be further studies done on BA.2 and future variants."

Campbell, too, said it's possible a future variant could render diagnostic tests less effective.

"We're probably not out of the woods on this silly thing," he said.