clock menu more-arrow no yes mobile

Filed under:

A negative Covid-19 test isn’t an all-clear

What is the probability of testing positive for Covid-19 before you get symptoms? Scientists really don’t know. 

A person sitting in a chair on a lawn gets instructions from a medical professional wearing protective gear and standing about 6 feet away.
If you don’t have symptoms, a negative Covid-19 test may not tell you very much.
Leonard Ortiz/MediaNews Group/Orange County Register via Getty Images
Brian Resnick is Vox’s science and health editor, and is the co-creator of Unexplainable, Vox's podcast about unanswered questions in science. Previously, Brian was a reporter at Vox and at National Journal.

Vice President Mike Pence has not tested positive for SARS-CoV-2, the virus that causes Covid-19. But he’s been in close contact, in recent days, with people with confirmed Covid-19 infections. And he debated Sen. Kamala Harris last night. Was this a good idea?

In a word: No. Per the Centers for Disease Control and Prevention’s own guidelines, Pence should have been quarantining, not debating, even if he’s testing negative. (That said, CDC director Robert Redfield had cleared Pence to debate.)

Since the Trump administration’s actions over the last 11 days have muddied the message from public health experts, let’s be clear: A negative test is not an all-clear for doing risky activities during the pandemic.

Scientists don’t yet understand exactly when a person who is infected with the coronavirus will start testing positive for the virus. There are situations when a person could test negative, actually be infected, and also be contagious. It’s also possible — since this virus multiplies itself exponentially in the body very, very quickly — that someone could test negative in the morning (and not be contagious), but by the afternoon test positive (and be very contagious).

Confusing? Yes, it is. But the bottom line is that Covid-19 diagnostic tests (both the slower, more common, viral genetic test — called RT-PCR— and the more rapid viral protein test, called an antigen test) are most useful, and most accurate, when used on people experiencing symptoms.

“One of the huge gaps now in the data is: What is the probability of testing positive before you get symptoms?” Benny Borremans, a disease ecologist at UCLA, says. Right now, scientists just don’t know for sure.

Why testing is less accurate before symptoms begin

There are several reasons scientists are unsure about when in an infection people will start testing positive for SARS-CoV-2. To understand why, and to make this less confusing, it’s helpful to think through all the things that have to happen for a Covid-19 test to come back positive.

First, the virus needs to take its time to establish itself in a person’s body. This is called the incubation period, and it can take upward of two weeks. On average, this happens in about five or six days. During the incubation period, a person might not test positive for the virus because there’s not enough virus in their body to detect in a test.

“The virus particles, day by day, will multiply,” Muge Cevik, a virologist and physician at the University of St. Andrews, says. “The virus needs to reach a threshold for the PCR [i.e. viral genetic] tests to pick it up.” PCR is the more common Covid-19 diagnostic test because it requires a lower threshold of the virus to test positive; rapid antigen tests would require a higher level of virus to register a positive test.

Testing positive should coincide with being contagious. But not always.

Generally, a person can start being infectious for the virus around two days before they start to show symptoms, in what’s known as the presymptomatic phase.

And, generally — but not always — scientists would expect that if a person is contagious, they’d test positive. After all, if they’re spewing enough virus out to get another person sick, they’re probably spewing enough virus out for a diagnostic test to pick up on.

But there are a few wrinkles here: When exactly a person makes the jump from testing negative and being non-infectious to testing positive and being infectious is hard to predict.

“If everything works as it should, the test should be positive if you are infectious at the very moment of the test, as there must be virus present then,” Justin Lessler, an epidemiologist at Johns Hopkins University, says. “However, you could easily test negative then become infectious a day, or even hours, after the test.” Unless you’re testing every hour, it’s impossible to get a fine-grained view on when the infectious period truly begins. (Also possible, but probably rarer: A person tests positive before they start to be contagious.)

Even if a person is contagious, they may not test positive. It could come down to where the sample for testing was taken from.

In general, “we consider the gold standard to be the nasopharyngeal swab,” Bobbi Pritt, the director of clinical microbiology at the Mayo Clinic, says. “That’s the deep nasal swab that goes all the way back into the back of your nose. Whereas other specimens — like a throat swab or just the very outer edge of your nose, like right inside your nostril — that’s not going to contain as much virus.”

Early on in the infection, a person who is incubating the virus is expected to test negative. Over the summer, Johns Hopkins researchers — including Lessler — published a paper estimating the likelihood of a false negative test in the first few days after being exposed to the virus. On the first day, they found the chance of a false negative is near 100 percent. No test is going to find the virus so early. Through the first four days, that rate drops to 67 percent on day four, on average, but with a very large range of error. On the day people first reported symptoms, there’s still a significant false negative rate, at 38 percent.

What does this all add up to? “What we’re saying is don’t test anyone in less than four days after exposure,” Cevik says. It’s not going to tell you much about the person’s status. Or if a person is tested in that time, they ought to be retested a few days later.

“In general, five to eight days after exposure is the best time to test,” Cevik says. “Or day three after symptom onset.” That’s when the genetic RT-PCR tests are most likely to reveal a true positive.

Because nothing about Covid-19 can be simple, here’s another thing to consider: The antigen tests that produce quick results have a shorter window in which you’d expect a person would test positive.

They are also slightly less accurate. But if used correctly, they can be very useful: They’ll test positive in the window when a person is most likely to be contagious. With repeated use, scientists hope these quick tests could help stop outbreaks from growing out of control.

The White House, on the other hand, has been using another rapid test, Abbott’s ID Now, to screen asymptomatic people. We just don’t know how good these tests — or any tests, for that matter — are at screening asymptomatic, or presymptomatic, people. “The FDA would be the first ones to tell you that they don’t know how the test is going to perform in that population,” Pritt says.

A negative test without symptoms might not mean much. Keep your mask on.

Here’s the bottom line: “We don’t know when one will test positive pre-symptom onset for PCR or antigen tests,” epidemiologist A. Marm Kilpatrick writes in an email. If you have symptoms, you’re likely to test positive the day you start feeling ill, but not guaranteed. The first few days after starting to feel sick, you have a very high probability of testing positive.

We could learn more in the months ahead about testing asymptomatic and presymptomatic people with studies following people after they have been exposed to the virus, and testing them repeatedly over a few weeks to determine the likelihood of testing positive before symptoms begin. “We have a lot of data from symptom onset onwards, but we don’t have data in terms of pre-symptoms,” Cevik says.

This is why testing is no replacement for other Covid-19 mitigation measures, like quarantining people exposed to the virus, mask-wearing, and social distancing.

“Testing negative is not like a passport for people to go out and do whatever they want to do,” Cevik says. If you might have been exposed to the coronavirus, like Vice President Pence was, you should quarantine for two weeks, regardless of what your test says.

CORRECTION: This post originally misstated A. Marm Kilpatrick’s name. It also mischaracterized Abbott’s ID Now’s test as an antigen test. The test looks for viral RNA.