Skip to main content
How Coronavirus Tests Actually Work

The coronavirus crisis is in large part a testing crisis. We are reading about tests. Arguing about tests. And, in many cases, struggling mightily to obtain tests for ourselves. But while test shortages are making headlines, there’s a lot about the technology behind these tests that isn’t as clear to the public. It’s like we’re all on one of those reality shows where you agree to marry a total stranger, entering into a high-stakes relationship with someone you couldn’t pick out of a crowd.

So let’s get a little more acquainted with the tests we can’t stop talking about. At the very least, learning a little more about the tests for COVID-19 will help you understand why it’s so hard to get one when you need it.

What happens when you get a COVID-19 test?

First, a medical professional puts a swab — an extra long, one-headed Q-tip — way, way, way up your nose. That captures viral particles, along with a bunch of other stuff that Dr. Davey Smith, a research virologist at the University of California, San Diego, referred to as “biological gunk,” like mucus and random cells. The next step, Smith said, is isolating the viral RNA — the genetic material that the virus uses to replicate itself. The RNA in a virus like the one that causes COVID-19 is similar to DNA, but instead of the twisted ladder of a double helix, it’s half that because it’s split down the middle. Some viruses carry their genetic code as DNA, but RNA viruses mutate a lot faster. That feature helps them jump species and evade both natural and medical efforts to kill them. Influenza, for instance, mutates so quickly that we need a new vaccine for it every year.

Just like DNA can identify a person, RNA can identify the virus that causes COVID-19. Isolating it requires a series of steps — adding different chemicals and repeatedly spinning the sample in a centrifuge — that aim to separate the sample into layers like a fancy cocktail shot, with the layer containing the RNA floating on the top. Then the RNA has to be further purified. There’s more than one way to separate out RNA, and companies sell kits that include the chemicals you need to make it work (called reagents, because they’re used to induce a chemical reaction).

Why The U.S. Can’t Process Coronavirus Tests As Fast As South Korea

From there, the RNA is mixed with short segments of DNA called primers. The primers and RNA get combined with loose building blocks of DNA, enzymes that work like genetic construction crews, and more reagents. Mix it all up, and your RNA turns into DNA.

Finally, the new DNA needs to be replicated until you have enough of it to actually study. That’s another chemistry kit — more primers, building blocks and reagents — doing what basically amounts to biological copy-paste, over and over. This is called a polymerase chain reaction (PCR) and the primers used here are especially important. These replication primers are basically fragments of the virus you’re looking for, Smith said, that will bind to the genetic material of that specific virus and nothing else If there’s no COVID-19 in the sample, then COVID-19 primers won’t replicate any DNA.

“We put in some special dyes so when it builds the right DNA we’re looking for, we can see the color light up on special machines,” Smith said. If you had COVID-19, your sample will now show up with freshly built, brightly colored DNA to prove it.

Wow, that’s a lot. Is that why people can’t get their COVID-19 test results for several days?

Well, no. This process is not particularly special or time-consuming and involves techniques used all the time in genetic research. PCR, in particular, is so crucial to the entire field that the guy who invented it won a Nobel Prize back in 1993. We use this same technique to test for all kinds of other viruses, from influenza to Zika, said Dr. Mary Jo Trepka, an infectious disease epidemiologist at Florida International University.

PCR doesn’t actually take that long to do — you can get results from an influenza PCR test in 30 minutes. Meanwhile, the chemistry set-style kits are something you can order from a website, and other needed machines can be as cheap as a used sedan. Any lab that works regularly with DNA has the ability, in theory, to do this testing.

In other words, the bottleneck slowing down test results is not a technical one. It’s about logistics and supplies, Smith said. In the earliest days, the Centers for Disease Control and Prevention was conducting all the COVID-19 testing in-house so they could monitor quality control, and they’ve been slow to authorize new labs to use their test kits. As more labs came online, though, they began running low on everything from the swabs they stick up people’s noses to the reagents that power the PCR’s chemical reactions to the human lab techs who actually do the tests, Smith and Trepka told me. It’s like a traffic jam.

Compare that to South Korea, where people can get their results in about a day. There, the government had been stockpiling the necessary chemicals for years after COVID-19’s cousin MERS briefly hit that country in 2015. That helped the country move quickly to approve and decentralize testing as soon as COVID-19 arrived.

So, wait. Is there no difference between South Korea’s test and ours, then?

There is a difference, but that doesn’t have much to do with why ours takes so much longer, Smith said. The differences are really just about which primers are being used. Researchers choose which chunks of a virus’s genetic code to use as a primer with an eye to making sure those primers are a) unique to the specific virus they’re looking for and b) likely to get preserved in any random bit of viral RNA scraped off the back of somebody’s nasal cavity. If the primer isn’t unique enough, you can get false positives. If the sequence your primer is looking for doesn’t make it from nose to test tube, then you’ll get false negatives.

There’s not really anything wrong with the primers the CDC test uses, Smith told me. Even when the CDC test was malfunctioning, that was about an inconsistent reagent, not the unique primers. South Korea is faster than us because it can conduct the tests in more places, and could right from the start. Plus, it has more supplies to do the tests with.

Is there no other way to test for COVID-19?

For right now, nope. That’s changing fast, though there are trade-offs. PCR is a great way to test for viruses because it’s both specific (unlikely to produce false positives) and sensitive (unlikely to produce false negatives), Trepka said. There are other tools that doctors use to test for more familiar viruses. For instance, if you think you have the flu, there’s a test you can take right in your doctor’s office that looks for antigens — the substances on a virus that stimulate your immune system into action. It produces results in as little as 10 minutes, but it’s not very sensitive and might tell you that you have the flu even if you don’t. “They’re helpful in clinical care but we don’t use them to monitor outbreaks,” Trepka said.

Meanwhile, new tests — like one from Swiss pharmaceutical company Roche that the Food and Drug Administration authorized on March 13 — are under development. The Roche test is 10 times faster at producing results than the standard PCR system. But unlike the PCR test, it requires the use of a rare (and proprietary) instrument. There are only 110 of these machines in the whole country. And that test still requires reagents, Smith said. As big companies like Roche bring these kinds of testing systems online, they’re sucking up the already limited quantities of reagent that smaller labs need for PCR tests. These big companies, such as Roche, are also making more reagent, Smith said. “But it’s very much profit-driven and they can corner the market very quickly,” he told me. “That’s not the end of the world, but it’s hard in the middle of the epidemic.”

CORRECTION (March 24, 2020, 9:43 a.m.): An earlier version of this article misstated the scientific definitions of sensitive and specific. When a test is sensitive it is unlikely to produce false negatives, and when a test is specific it is unlikely to produce false positives.

Maggie Koerth is a senior science writer for FiveThirtyEight.