COVID testing FAQ: Types, accuracy, and how much you should trust them

This week LA Mayor Eric Garcetti blamed small gatherings and parties for yet another spike in Covid cases in the region. With holidays coming up and a lot of people planning to use testing as a barometer of whether it’s safe to get together, KCRW turned to Omai Garner, director of clinical microbiology for the UCLA Health System who answered frequently asked questions about COVID-19 testing. 

KCRW: Let's get an overview. How many different types of tests are there out there and available to most people?

Omai Garner: “There's two tests for what I'll call acute disease, and that's either the PCR [polymerase chain reaction] tests or the antigen tests. Then we have a third test for whether or not somebody has been exposed to the disease or caught it in the past, and that's called the antibody test.”

Let's begin with a PCR test. And this is the test I understand that you would get if you went to Dodger Stadium, for example, or any of the other county sites, right, you'd get a PCR test?  It’s the most accurate. How does it work?

“I think this is a little bit of a challenge in how we talk about the testing. When people say a testing center, what they actually mean is a collection center. So what happens, let's say at Dodger Stadium, is a sample is taken, whether it's the cheek swab or the deep nasal swab, and then that sample is sent to a centralized laboratory, and the laboratory then performs the PCR test. 

Now the PCR test works in two steps. The first step extracts all of the DNA and RNA or nucleic acid out of the sample, the second step looks specifically for the viral RNA from COVID-19. It's exquisitely sensitive, meaning there can be a very low amount of virus there, and the test will find it and call it positive. So if there's a very low level of the virus, that means that you could not be showing symptoms, and it would still pick that up.”

If you've been infected. What is the ideal time period to get the most accurate PCR test? Can you get it? For example, if you have just been infected within the last 24 hours? Can you get an accurate test? Or do you have to wait a little bit longer? 

“If you look at the typical infection course, for COVID-19, if they're going to progress the symptoms, you typically do that between five to eight days after encountering the virus. That really ends up being the best window to test to see whether or not somebody has been infected with the virus. Now some people, it takes all the way out to 14 days. So you can't guarantee if somebody is negative in that time that they aren’t going to become symptomatic. But that five to eight day window post, encountering the virus is really the best time to be able to get the most accurate tests. Now what's nice about PCR is that in the two to three days before symptomatic illness, that's when people are potentially contagious, asymptomatic, and where a PCR test can be very valuable in asymptomatic screening.”

How many false negatives or false positives are associated with this test?

“So we've looked at that at UCLA health, and we find that our false-negative rate is below 2%. And we see very, very few false positives. Now false positives [are] a little bit difficult because if people are late in the infectious process, let's say they're on day 10, of having symptoms. They could be positive by PCR, but not necessarily be infectious or contagious. That's something we're still trying to figure out.”

This is all assuming that the collection process is good. But if you have people self-testing, which is often the case. They are taking their own swabs or sticking the cotton swab up their nose or swabbing their mouths or whatever, and doing it themselves. Doesn't that introduce a level of uncertainty in terms of the accuracy?

“It certainly does. And the accuracy numbers that I'm quoting to you are accuracy based on the deep nasal swab collected by a healthcare worker. So the best possible collection, anything else, either a cheek swab, or a self-collected swab or saliva all hinders the accuracy. Because this best specimen has not been taken.” 

Let's move on to the antigen test. This is often called the rapid test, right? Because you can get the results within minutes, but it's not as accurate?

“Yeah, that is correct. So one of the advantages of the antigen test is it doesn't require the centralized laboratory that I spoke about for the PCR test. And so as soon as the sample is collected, it can be tested on-site, giving you a rapid result in typically 10 to 15 minutes of sample collection. You're more likely to get a false negative on an antigen test than you are on the PCR test. A better way to think about it is it requires a much higher level of virus to be able to call a positive than something like a PCR test, which is why antigen tests are only FDA approved for symptomatic testing within the first five days of illness.”

Let's say I want to have a Thanksgiving gathering with a few of my family members with whom I don't live, or friends. And we say, okay, everyone's got to get tested before we get together. Is there a way to do that, and be assured that the test you get is accurate, that your gathering would be safe if everyone tested negative? 

“There is not. I think the emphasis from public health entities is ‘please don't do that.’ While I understand the stress of it, the recommendation is to not have those gatherings in that way, because there isn't a test that can be performed that same day that can guarantee that you don't have the virus, and thus won't spread it at a Thanksgiving gathering the next day. We don't have a test that can do that that is available. And so the suggestion is to not have those gatherings as we're moving forward into one of the highest peaks that we've seen with COVID-19 across the country.”



  • Omai Garner - Director of Clinical Microbiology at UCLA Health