Testing for COVID-19

Welcome to the Australian Prescriber Podcast. Australian Prescriber, independent, peer-reviewed and free.

It seems a key struggle we have with COVID-19 is knowing which patients have it, despite the implications of a diagnosis being so substantial. It's therefore understandable that testing for this infection has been in the spotlight possibly more than any other type of testing before. So what testing is out there? What are the strengths and limitations? And how do we do this in practice? I'm David Liew, your host for today's podcast, and I'm joined today by someone well-equipped to help us navigate the complexities of this issue. Anna Brischetto is a microbiologist and infectious diseases physician in Brisbane, and she's written an excellent early release article, now out in Australian Prescriber on testing for COVID 19. Anna, welcome to the program.

Hello, thank you for having me.

So Anna, tell me this. And I think this is a question that gets asked a lot, especially in the community. Why can't we have a test that does it all for COVID-19?

Now I understand this because there are so many tests for COVID-19. It can be really confusing working out which one you need to use. So I guess the first thing to understand is why you're using your test. If you're using a test to diagnose acute infection, somebody who's symptomatic in front of you, or perhaps you're trying to look to see whether a patient might have had COVID-19 in the past. So the main test that we use to diagnose acute infection is a nucleic acid test. And the most common one that's utilised is the real-time PCR test. Now, this is used to detect the actual RNA of the SARS-CoV-2 virus. Initially, when a patient is infected with the virus, the virus will replicate. And the point at which a viral replication reaches a high enough level, that's the point at which a sample that you take will be able to detect the RNA.

So at this point it looks like the highest viral loads of SARS-CoV-2 occur about 48 hours prior to symptoms occurring or on onset of symptoms. And this is the point at which our nucleic acid tests or our real-time PCR tests are going to be the most sensitive. After that there'll be gradual decline of virus RNA in the respiratory tract. And this will also affect your ability to detect the virus. Now, at some point, usually around 10 to 14 days, sometimes a little bit longer than that, there won't be any detectable virus left in the respiratory tract. So to be able to diagnose people who have had COVID-19 previously, that's where the serological tests come into it. So it's really about using the right test for the right stage of illness.

I mean, that makes a lot of sense. And I think that's a subtlety, which is probably not easy to convey in the community but really kind of underpins what kind of tests we would use at what point in time.

Yeah, that's right. The sample type preferred for this test is the combined deep nasal and throat swab. And it can be either a deep nasal, which is where the swab goes in about two to three centimetres, or a nasal pharyngeal, which is where the swab goes all the way back to the throat, and studies have actually been done comparing these two techniques and they've been found to be comparable. So that's why we will be accepting the deep nasal and throat swabs, as well as the nasal pharyngeal. As anyone who's had a nasal pharyngeal swab can attest to, they're not the most comfortable of tests. So there was real interest in looking to see whether or not the less invasive, deep nasal swab was comparable and it was found to be so.

Yes, well, as someone who's had five tests I can definitely attest that that’s definitely the case. But there are a few different variations, aren't there? I mean, there's obviously the real… the RT-PCR, which we use in common practice. And then there are some other tests as well, like point-of-care tests, like the GeneXpert tests.

Yeah, that's right. So you can separate them out into the laboratory-based nucleic acid tests and the point-of-care tests. And the difference is where it's performed. So the point-of-care tests are designed to be able to be performed at the patient's bedside. They rarely are, but can often be performed quite near, say in emergency departments or clinics. And then there's the laboratory-based nucleic acid tests. These tests are really the cornerstone of diagnosis here in Australia. This is where samples will arrive in the lab. There'll be batched. So we'll do multiple samples at once. We extract the RNA, so that's one step, and then go on to have the real-time PCR step. Now all up from when it arrives in the lab to getting a result, the actual testing time is about six hours, but by the time you log in samples, batch them together and also have to think about transport of that swab getting to the lab. We usually say turnaround times going to be about 24 to 48 hours, obviously, depending on how many samples were receiving.

Now, the advantage of these tests is that you can perform large quantities of swabs at the same time. And that's obviously very efficient. The point-of-care tests, including the GeneXpert, are a little bit different. It is actually a real-time PCR, but instead of having to have a separate step to extract the RNA, you can actually put the sample straight into a little self-contained cartridge and the whole process takes about 45 minutes. So the advantage of this test is obviously the speed at which you can get a result. However, you can only do one test at a time and you can imagine, Oh gosh, peak in Melbourne, I'm sure you were getting, certainly in Brisbane here, we were getting up to 10,000 swabs a day. You were getting more than that.

Just would not be sustainable to do all of your testing like that. So really should be saving the point-of-care tests for those situations in which you would need to return a result very quickly. The other thing to consider for these point-of-care tests is that there is limited reagents available and each lab is allocated a certain number of these cartridges. And once we use them up, they're gone. So we're really trying to decide which tests really need to be returned in that quick timeframe and which can go through the usual throughput of the lab.

Yes, because I mean, I guess that kind of technology in the past has really been used for relatively low-volume things. Hasn't really, I mean, it's been used in selective situations where we need that quick turnaround.

They actually have been in use during flu seasons and have been used in a similar way, in that we've used them for those patients in which you want a quick result. Also, obviously very handy to have when you need a result quickly in areas in Australia where there's going to be long transport times because there's a lot of centralisation of laboratories.

Yeah. So really your detecting parts of the virus. And I guess that probably gives us an idea as to why people can swab positive even when they're clinically well, and it probably gives us an idea about the limitations and what might dictate sensitivity or specificity in that kind of situation.

Yeah, absolutely. And it's really tricky to be able to give you one quoted sensitivity and specificity of these assays. We know if viral RNA is present or very good at detecting it, however, there's a lot of other factors that will affect the sensitivity and specificity, mainly the sensitivity. Depending on the point of the illness that the patient’s in, the sensitivity is going to be very different. Obviously at the beginning of the illness, we know with SARS-CoV-2 you have very high viral loads, much more likely to return a positive result.

However, if you're later in your illness, or even if you sometimes, I guess when people are testing asymptomatic close contacts who don't have symptoms yet not detecting the virus doesn't necessarily mean that you either haven't had COVID-19 or are going to develop it. It just means at that point, the level of virus in your respiratory tract isn’t high enough to detect. When somebody has the disease COVID-19 but test negative for SARS-CoV-2, we call that a false negative of a test, but often it's more factors relating to specimen collection and things like that that can affect that. False positives are pretty rare, especially to do with the assay. There's a few ways in which you can get false positive tests in the lab. The most common to be honest is what we call pre-analytical factors, which is where there's issues that occur before it even comes to the lab. So, if the wrong swab types been collected, labelling of the swab might have somebody else's details on it. Things like that.

Yeah, absolutely. I mean, I think a lot of people would be conscious as well that there are other corona viruses that SARS-CoV-2’s obviously only one of many. What are the chances of picking up another corona virus as a positive on an RT-PCR SARS-CoV-2?

Yeah. So, so it's pretty low and certainly a lot is done in the design of these assays to minimise that. They certainly would look at the genome of all of the corona viruses and make sure that you're picking a very specific part either of a different protein or part of that protein. So yeah, the chance is very low.

Yeah. Okay, great. So lately there's been a lot of discussion about antigen testing, so that's different to the RT-PCR isn't it? But I guess it's picking up, we're still looking for active infection. It still plays a similar clinical role.

Yeah. So the nucleic acid tests are looking for the RNA of the virus, whereas the antigen test picks up proteins of the virus. So of course, RNA is the nucleic acid, which encodes proteins, and then the antigen test picks up the actual proteins that have been expressed by the virus. There's similarities in influenza testing in which real-time PCR is the cornerstone of diagnosis, but antigen tests are available and can be used. And like influenza testing, early reports seem to suggest that the antigen testing for SARS-CoV-2 is less sensitive than the nucleic acid real-time PCR tests. So I guess certainly I don’t think there'll be replacing the real-time PCR test and some antigen tests are just being approved on the market here in Australia.

We're really working out how best to use these tests. They're less sensitive. And so we think that any test would probably be a good idea to follow up with testing with a nucleic acid test, but they're often… these antigen tests can look like little pregnancy tests. So they're a lot less sophisticated, a lot easier to do with less technology and expertise. You can do them at satellite labs. They're pretty cheap compared as well to the real-time PCR tests. And so some people are looking into whether you could use them in population-based studies. So yeah. Watch this space. At the moment, I'm not aware of any labs offering them, but they're certainly people who are looking into whether or not we think that they are going to have a place in our testing algorithm.

Okay. Right. So I think that on top of all of that, like you said before, there's a separate process looking at antibody testing. And I think we're probably all a little bit familiar with antibody testing from things like testing for viral hepatitis. I guess this is really a different, it's really trying to target a different role in SARS-CoV-2 to testing in COVID-19.

Yeah, that's right. So initially there were reports of the antibody tests being used for acute infection because of course the body makes antibody in response to infection. It makes IgM usually first in acute infection and then IgG after the factor in chronic infection. However, with SARS-CoV-2 it turns out that you produce IgM and IgG often around the same time. And a lot of people will start producing antibody by day seven, but consistently people are producing it by day 14. And there are some people who don't produce it until up to 28 days post onset of symptoms. So really the use of antibody tests in Australia is in diagnosing past infection.

Have we done much of that? Yeah.

Yeah, so there have been. So examples of where antibody tests might be useful. Patients who had symptoms that were consistent with COVID-19, but might not have had access to testing. And this really happened early on in the pandemic, especially overseas. So we certainly had a little spike in people who'd returned, say from the US or Europe around February or March, had had symptoms, but didn't have access to testing in those places. And so we're interested down the track as to whether or not they might have had COVID-19. Sometimes we're using them in people who have unexpected positive PCR results or inconsistent PCR results to see if it could just be one more piece in the puzzle in us trying to determine whether the patient actually has the disease. And that's often as well used in outbreak investigations to see if people might've been exposed.

What about false negative results, I think you're getting?

Sorry, so yeah. Look not everybody produces antibodies to SARS-CoV-2. We're estimating it at around 5 to 10% of people. And the other thing that we're not sure about is the rate at which people are losing their antibodies. So the rate at which antibodies wane and become undetectable, and certainly there's cases I can think of where we've had known PCR-positive patients who've come through the lab, but then done follow-up serological testing, and the assay may be negative. So yeah, I mean, look, sensitivity’s not 100%, you're certainly going to have to use it using clinical data and epidemiological data, putting it all together to decide whether a patient has had the disease.

Yeah. And I guess as we go forward and I mean, there's a lot of talk about immunity passports and about how we might go about vaccination obviously as well. How might those antibodies correlate to protection? Do we know that they are definitely protected?

We don't know. So we don't have the data to know that the antibodies that we are testing for whether or not that will correlate to protective immunity. So I think there's a lot of work to be done on what antibodies we're going to test for. At what point after your illness, are we going to test for these antibodies to be able to confidently say that you have developed them and you would have immunity, and I guess we're still gathering data on whether just how much immunity natural infection gives. Certainly it seems like your risk of developing SARS-CoV-2 after having the disease is low, if you're re-exposed, but there's case reports coming out of people with repeat infections. So again, I think it's really too early to tell.

Yeah. Well, let me ask you one more question. Where do you think that this field is going to evolve in the future? I mean, usually when I say the future, I'm talking the next 10 years, so somewhere between the next 10 days to 10 months. Where do you thing things are going to evolve?

When we were writing this article, actually, because, I wrote it, as you do a few months ago, and then it goes through review. And by the time it came back to us, we had to add whole sections of tests, which were now approved in Australia that we were now thinking really should be included. So it is just so fast moving. I think that diagnosis of acute infection, I think real-time PCR, is what we'll be using. I think labs are really having to work out how to cope with these surges. That's the really tricky part because we might be routinely doing 1000 tests a day. And then all of a sudden one weekend we might have to do 9,000 tests a day and just really working out how to have processes in place that we can cope with that and have capacity and really cutting down on some of the front-end work. So I think that's where a lot of work's going into, being able to cope with such high numbers of testing in the acute phase. But certainly as you were saying, I think as our understanding develops of antibody response and serological testing, I think as the world moves through the pandemic there’s going to be a lot of questions about what having that antibody means, whether you can be classified as immune and what it means for I guess opening up countries.

Yeah. I mean, I think it's interesting to hear these reports from China, where… they've tested a whole city, is that something that we are likely to be able to do?

I guess it's important to really, really think about who you are going to target testing on because once you start upscaling testing to such a degree, the efficiency and the turnaround time of results is of course going to blow out. And then it gets to the point where if you were turning results in four to five days, it's really not clinically meaningful and especially from a public health point of view, and especially if patients aren't isolating at home.

So I think the approach at the moment really is testing symptomatic people, but also I guess, extending that to asymptomatic people in things like outbreak situations and in nursing homes and things like that, where the consequences are high. Which I think is what we've been doing certainly up here in Brisbane and seems to be working quite well, because you really want to balance doing a whole lot of tests with the disadvantages you get as soon as you're testing so many people with a low pre-test probability of having the disease, you're going to have some positives that you might have to sort out and work out if they're real, but also you're going to have to deal with turnaround times that'll blow out.

Absolutely. And I guess that's where the National COVID-19 Expert Task Force and the Living Guidelines have just been so useful in Australia.

Yeah. And I've really, I've drawn heavily on the PHLN, the Public Health Laboratory Network, guidelines and the Communicable Diseases Network Australia, the CDNA, guidelines as well, in writing our article.

And it's an excellent article. I'd encourage everyone to have a look at it. Anna thank you so much for joining us on the podcast today.

No problem. Thank you for having me.

The views of the guests and the host on this podcast of their own and may not represent Australian Prescriber or NPS MedicineWise. I'm David Liew. Thanks for joining us once again, stay safe and we look forward to seeing you next time.