There’s a lot of talk right now about PCR testing, and with good reason — PCR tests are what’s being used to determine whether someone is “positive” or “negative” for the novel coronavirus, SARS-CoV-2, that is believed to cause COVID-19. But most of the general public doesn’t understand what a PCR test is, and most of the explanations are pretty technical. So I want, in this post, to skip past most of the technical specifications and explain in practical terms what PCR testing is, what it can tell us, and what its limitations are.
DNA & RNA
Before we can get to that, we have to back up just a little and talk about DNA and RNA, because these are what the PCR test works with. For those of you who don’t remember biology class, DNA (deoxyribonucleic acid) is kind of like the “computer code” for living things, that resides within each cell. It’s comprised of chains of “nucleotides” which match up in pairs kind of like the teeth of a zipper.
When the DNA in an organism (living thing) is ready to be copied, it first has to “unzip” itself. RNA (ribonucleic acid) is a similar type of code, but it’s already “unzipped.”
What is a PCR Test?
A PCR test is designed to look for a certain set of DNA — a certain “zipper,” if you will. Or, if what we have is RNA, to build the “rezipped” version, and look for that. (PCR stands for polymerase chain reaction, and RT-PCR stands for reverse transcriptase polymerase chain reaction. The “RT” part refers to the “rezipping” process.)
But, as you can probably imagine, when you’re looking for something that small, it can be like looking for a needle in a haystack, so what this test actually does is double the material being tested, over and over, by copying the DNA repeatedly, so there’s more of whatever you’re trying to find.
For example, if we were able to do this with M&M’s™, imagine you had a pile of M&M’s™ with a single green one in the mix. You double the whole thing, so now you have 2 green…again and you have 4 green…then 8, 16, 32… At this point, it’s much easier to determine whether there’s a green one there, because it’s easier to see 32 than 1.
What Are the Tests Looking For?
However…the Covid PCR tests are not looking for the entire SARS-CoV-2 genetic sequence (the whole “zipper”). They’re looking for smaller snippets. (Different tests use different snippets.) 1 This is kind of like when you’re looking for a quote in Google and instead of typing the whole quote into the search bar, you type just a phrase that makes up part of the quote.
The quality of your results will depend on which phrase you choose, right? For intance, let’s say I’m looking for “You are the salt of the earth. But if the salt loses its saltiness, how can it be made salty again?”
“You are” would obviously be a ridiculous phrase to search for, because it would turn up a lot of completely-unrelated things! It’s very non-specific.
“Salt loses its saltiness” might be a better choice. But if I put that into Google, not only do I get some links to the verse I’m looking for, I also get a few links to chemistry pages. This, then, is much more specific than “you are,” but can still turn up some “false positives.”
I have to look for “You are the salt of the earth” before I start to see results that are pretty consistently limited to the actual biblical quote I was searching for.
We see, then, that the reliability of this type of test is limited, in large part, by the specificity of the snippets the test’s creators chose to look for. We want to know that whatever genetic material they’re looking for, it isn’t found in anything else — only in what they’re looking for.
How Many Cycles?
Another limitation of PCR testing the number of cycles used — that is, how many times they double the starting material. Copying DNA isn’t really as simple as multiplying M&M’s™, so errors and “noise” can be introduced along the way.2
That means scientists have to carefully balance making enough copies so what they’re looking for will show up, with making few enough copies that the errors and noise don’t make them thing they’ve found what they’re looking for when they really haven’t.
There isn’t a single agreed-upon number of cycles, either in science in general or for Covid testing, specifically. Usually, labs use somewhere between 25 and 35 cycles. More than 45 cycles are not recommended and, according to some sources, it’s best to keep it under 40.
If one lab doing Covid testing is using 25 cycles and another is using 35 cycles, you might get different results at each lab, even while testing the same sample from the same person. This makes it difficult to compare data from one state to another, or even, sometimes, within a state.
Is it Infectious?
The final limitation to PCR testing is that it can’t differentiate between an active virus and an inactive one. Think of it kind of like this: if someone tested my DNA, they would be able to identify that DNA as being from me. But they would have no way of knowing, on the basis of the test, whether the DNA was taken from me as a living person, or whether I’d died and someone had collected the DNA from my corpse.
PCR testing for a virus is very similar to that. Viruses aren’t actually living things, but there are factors that can make them either “active” or simply inert genetic material, and the test is unable to differentiate between these. All it can do is tell us (assuming the test is well-designed) whether the snippet is one that comes from the virus they’re looking for.
In Summary
PCR tests are very good at what they’re designed to do, but this type of testing was not originally designed as a diagnostic tool and, when used that way, has certain shortcomings.3 (Most diagnostic tests have limitations. It’s just the nature of reality, but we should be aware of what those limitations are.) Be educated so you know what the test can and cannot effectively do.
- https://jamanetwork.com/journals/jama/fullarticle/2765837
- https://www.thermofisher.com/us/en/home/life-science/cloning/cloning-learning-center/invitrogen-school-of-molecular-biology/pcr-education/pcr-reagents-enzymes/pcr-cycling-considerations.html
- https://www.reuters.com/article/uk-factcheck-pcr/fact-check-inventor-of-method-used-to-test-for-covid-19-didnt-say-it-cant-detect-infectious-viruses-idUSKBN24420X