A shortage in U.S. testing capacity has been a major bottleneck in containment efforts throughout the COVID-19 pandemic. At first, the Centers for Disease Control and Prevention developed its own test kits and distributed them nationwide. But problems with their chemical reagents and overly strict testing criteria caused major setbacks. The Food and Drug Administration later gave companies the green light to make their own assays, and testing capacity increased significantly. But the numbers of tests has declined worryingly in recent weeks, possibly because fewer people are seeking them as a result of long lines and wait times.

Now Yale University researchers have developed a saliva-based assay that is easier to administer and analyze than the standard nasal swabs and can be manufactured using a range of chemical suppliers. The FDA granted the COVID-19 diagnostic test, called SalivaDirect, an emergency use authorization last week.

Unlike the typical test for the disease—which infamously involves inserting a long swab up the nose to what has been described as a “brain-tickling” depth—SalivaDirect only requires a subject to spit into any sterile container. It also skips the lengthy extraction step used to isolate the novel coronavirus’s genetic material, or RNA, before amplifying it to a detectable amount using a technique called polymerase chain reaction (PCR). And while it is not the first saliva test the FDA has authorized, SalivaDirect has been shown to work with chemical reagents from a variety of companies, so it reduces the risk of the supply-chain issues that have plagued other methods. The assay can run 92 samples in about three hours, although the turnaround time depends on the lab conducting the test. Its components cost less than $5, so labs could charge about $10.

SalivaDirect was evaluated in a seemingly unlikely population—the National Basketball Association—which hopes to use it to detect players who have the virus but show no symptoms. Scientific American spoke with Chantal Vogels, a medical entomologist at the Yale School of Public Health, who helped develop the saliva test, about how accurate it is, how it compares with nasal swab assays and how easily it could be scaled up for use by other laboratories.

[An edited transcript of the conversation follows.]

How does the saliva test differ from a standard nasal swab assay for COVID-19?

Our SalivaDirect method has three main differences: the first one is we use saliva instead of a nasopharyngeal, [or nasal], swab. And that is mainly because the nasopharyngeal swab is very uncomfortable, and you need a very skilled person to take a swab. If you don't get deep enough and get enough virus, then you might not detect it. When you take the swab, there’s a chance it will prompt coughing and sneezing, which puts the person who takes the swab at risk. That’s why we figured if we can use saliva—which you can essentially just gently expel into a couple of tubes—that’s more comfortable, and it’s safer to collect. And we found that it actually works really well for detection of SARS-CoV-2.

The second major difference is that many of the other methods that are out there use an RNA extraction step in the sample processing. That is a step where you try to extract the genetic material of the virus: you break down the virus particles and everything that’s in the sample, you purify that, and that then goes into your final testing step (for instance, PCR). What we found is that you can actually skip, or at least simplify, that [extraction] step. Instead of needing this whole extraction with special equipment, if we simply treat the saliva with an enzyme called proteinase K and heat it, we can successfully break open the virus. We lose a little bit of sensitivity, but still, it’s very good for virus detection. It’s important to note that our method still uses PCR, and that means you do need to perform the test in a CLIA [Clinical Laboratory Improvement Amendments]–certified lab—a clinical lab that is equipped to run these tests. So it’s not a test that you can run at home. It’s not a point-of-care test.

Chantal Vogels. Credit: Isabel Ott

The third thing—and I think it’s kind of the most unique part about our method—is that it’s very flexible. Normally, most of the methods that are out there are based on a [specific test] kit. But if this kit is running into supply-chain issues, then that’s really going to negatively affect that testing. So we validated saliva directly with reagents and instruments from different vendors. Therefore, if one vendor is running into supply-chain issues, at least there are still other options out there that you can use. And this flexibility really minimizes the risk for supply-chain bottlenecks.

How accurate is your test?

As part of our FDA application for emergency use authorization, we had a couple of components that we had to perform in the lab. One is a “limit of detection” experiment, [which determines] the lowest concentration of virus that you can detect in a sample. What we found is that we can detect six to 12 copies of virus in a sample, which is pretty sensitive. The other component was a clinical evaluation. In that part, we had a selection of paired nasopharyngeal swabs and saliva from the same individuals, and we tested the swabs with an already authorized assay. Then, in parallel, we tested saliva with SalivaDirect just to show that both [were in] agreement in testing results. What we found there is that if you compare those nasopharyngeal swabs with saliva, there’s a 94 percent positive agreement. It’s important to note, though, that we had a couple of nasopharyngeal swabs that were testing positive with SalivaDirect but that were negative for the nasopharyngeal swabs. And that kind of gets to the point that if you compare these two different sample types, there’s always some variation.

If saliva works, why do we use nasal swabs at all?

Before the pandemic, the nasopharyngeal swab was used for detection of many other pathogens. So it has always been the gold standard. Early on in the pandemic, you had to rely on information that was previously known. And because the swab is used a lot on these respiratory pathogens, that’s the sort of the thing that you go to first. Along the way, we just started developing this more and looking into alternatives. Now it seems like saliva is indeed a good alternative. It doesn’t rule out the swab totally, but at least for certain contacts and certain scenarios, [saliva samples] may prove to be better.

Do you see SalivaDirect as replacing existing COVID-19 tests or complementing them?

I think it’s never like “One test is going to solve the whole problem.” I mean, we're dealing with a pandemic, and I think we just need options. There are many [other] saliva tests out there, and that’s really what we need. It’s not that our test is going to be, like, the answer. I think we have to be realistic. We still need that PCR test that needs to be performed in a highly complex laboratory. And for certain settings in certain contexts, it might be better to have an in-home test. But if you need to do lots of repeated testing, [our assay] might be better. So I think we just need a lot of different tests out there that can be used to tailor what is best in a certain context. Also, for instance, in low-resource settings, if you don’t have a constant supply of electricity or if you don't have access to one of those PCR machines, then you do need an alternative. I think having multiple options for testing out there is great.

How did the NBA come to work with your group to validate the test?

I was not involved in how that all got set up. That was my principal investigator, who the NBA reached out to. But essentially, to [authorize the test for] asymptomatic individuals, you would need a population of asymptomatic individuals who are still being tested regularly. Then the type of study that you need to do is, again, to compare results from the nasopharyngeal swab [with those from] saliva. It’s quite hard to set up a study with an asymptomatic population and then have them get repeatedly tested with both the swab and saliva. The NBA has been great because the players and the staff were going to be tested regularly with the swab: by also taking the saliva sample, we could just really nicely use them as our testing population to set up the study, to actually collect the data.

What are the next steps in scaling up this type of testing?

Even though now SalivaDirect has received emergency use authorization, so far it’s just for suspected COVID-19 cases. We’re now working in collaboration with the NBA to also get asymptomatic individuals authorized. We’re also looking into pooling: Can you pool multiple saliva [samples] together and still be able to detect the virus so that you can find ways to scale up testing? We’re looking into ways of automation, because now it’s all manual processing of the sample. But if you could use robots, that can be a way to really scale up the testing. We’re trying to team up with other universities and other groups as well to really help people push out their tests, because it’s not like our test is going to be the answer to everything. I think it’s a very good first step, but there’s still a lot more to do.

Read more about the coronavirus outbreak from Scientific American here. And read coverage from our international network of magazines here.