There’s been a lot of talk about testing folk for COVID-19. It would seem blindingly obvious testing is important. It’s not just about people knowing their status so they can get the right care and reduce the possibility of their infecting others, it’s also crucial to inform an evidence-based response to the pandemic by the authorities.

It also allows critical workers like those in healthcare to know if it’s safe to go back to work.

But as ever with biology, it’s complicated. I don’t want to get into the politics of mass testing; for this post I just want to do the nerdy bit about how the different tests actually work. It’s wankingly cool so put the kettle on, pull up a chair and let’s do some biochemistry porn…


There’s a number of methods available to diagnose COVID-19 but each has its own strengths and weaknesses.

  • Clinical. There are signs that will lead to high clinical suspicion it’s COVID-19 rather than something else: fever, some typical chest CT features – but not everyone who’s infected has symptoms. But for those who do, confirmation with a pathological rather than purely clinical diagnosis is desirable. But that’s for the medics. I’ll stick to the microbiology.

  • Genetic. This is where you confirm the little bastards are present by taking a sample such as a throat swab and finding actual genetic material from the virus. This method is highly specific but has its drawbacks too.

  • Serology. This is the fascinating bit. Long after symptoms have subsided and you can’t recover any virions you can still find their footprints, fossilised in your blood. It’s these sorts of tests that are currently being rolled out to healthcare workers and they are very, very cool…

Genetic: PCR

This stands for Polymerase Chain Reaction. This is where you get a nose or throat swab and see if you can find the viral genetic material (RNA) in it. The PCR bit amplifies the signal by (usually) converting the RNA into DNA and then copying it many times until there’s enough to actually do the test.

Without getting into the fine detail of all the different PCR techniques now available it’s the same process used to amplify DNA recovered from crime scenes – and its inventor, Kary Mullis, got a Nobel for it.

But even though PCR exhibits both sensitivity and specificity it has its drawbacks; in the early stage of the disease the viral load might be too low to be detected or the subject might not have have major respiratory symptoms – and so little detectable virus in the nose and throat. You also need really good technique; it’s a very sensitive assay and samples are easily contaminated.

So, in many respects PCR is the ‘gold standard’ but it’s also time consuming, finicky and expensive.


This is the test that’s being rolled out now. It involves looking for evidence of the body’s reaction to the virus found in blood serum (or lymph or other extracellular fluid) – and it just needs a finger-prick of blood, something that looks like a pregnancy piss-stick and about 10 minutes. No labs. No mucking about.

Here’s how it works.

When our bodies are challenged by an invader, the body can mount both specific and non-specific defences. I want to concentrate on one specific response for now, antibodies.

An antibody or immunoglobulin (Ig) is a large protein used by the immune system to neutralise pathogens like bacteria and viruses. Individual antibodies are specific to a particular challenge and there are tons of them, one for each specific antigen you’ve ever been exposed to.

Antibodies recognise a molecule unique to each different pathogen (an antigen) and lock on to it. It’s like they recognise the fingerprint of a known offender. By attaching it tags a microbe or an infected cell for attack by other parts of the immune system. In some cases antibodies can neutralise the target directly by binding to a part of it that it needs to do its stuff and thus blocking it.

One of the first antibodies to be produced in an infection is IgM. It appears early in the course of an infection, then disappears but can reappear after subsequent exposure to the antibody if the person is reinfected.

The other class of antibody of interest here is IgG. It’s the most common type of antibody found in the body and does lots of cool stuff – but what’s of interest here is it stays in the blood and extracellular fluids after the infection has passed and after both IgM and any viral RNA have long gone.

You can see what I mean here:

So, viral RNA (from virus particles) is there from Day 1 of infection and if you can find it there’s a very specific test for it. Three days in IgM turns up and (in the case of COVID-19) is detectable until the end if the third week.

IgG doesn’t arrive until the end of Week 1 – but after that it’s around for good. It’s only in small – but detectable – quantities: it’s one of the parts of the immune system that’s keeping a watching brief for future infection. If the IgG comes across the specific antigen it was raised against it alerts the other wings of the immune system and starts a new response – and rapidly.

The Test

So, a couple of days or so after onset of symptoms the presence of antibodies tells us whether someone has been exposed to the virus and gives us a steer as to their likely status.

The test itself looks like a pregnancy test piss-stick sort of affair. There are several on the market but all work in the same way. Because blood is thicker than pee there’s a well to put a pinprick of blood into and another for a buffer solution – the buffer carries the blood plasma into the bit of the cassette where the clever stuff happens. Remember doing chromatography at school? Just like that.

The sample of blood goes in the well (S), the buffer is added to (B). A few minutes later a number of things might happen:

  • A coloured line next to C shows the buffer has wicked all the way to the top. No line here means the test cannot be relied on, irrespective of any other lines you can see in the cassette. This is the control.

  • A line next to IgM means it’s present – so likely there is an active infection;

  • A line next to IgG means the sample is from later in the infection as it takes a week or so for IgG to appear.

  • IgM without IgG means it’s likely the patient is in the earlier stages of the illness. Remember some have asymptomatic or mild illness.

  • IgG without IgM means it’s likely the patient has had the illness and recovered.

Assuming you do PCR and serology, here’s the permutations:

PCR IgM IgG Interpretation
+ Early infection – window period
+ + Early stage of infection
+ + + Active phase of infection
+ + Late or recurrent infection
+ Probably early stage and PCR false-negative
+ Past infection and has recovered
+ + Recovery stage and PCR false-negative

There is some other stuff you need to do when designing these like ensuring nothing else in the sample – blood in this case – interferes with the result.

So, antibody tests are quick to perform (10mins max), unlike PCR they don’t need any special kit or trained technicians and are pretty accurate (87% for IgM, 97% for IgG).

Even better, they can rapidly give an indication of what stage in any infectious process an individual is and can tell if someone has immunity long after the virus has cleared. What’s not to like?