PCR testing – Science or Politics?

(now updated, to reflect comments from interested parties – thanks!)

PCR skeptics suggest that at the high Ct (amplification cycle threshold) values used for COV-2 testing (above 35, say) the test could result in positive results even in the absence of any meaningful virus presence.  PCR supporters say that the PCR test is definitive.

The only real way of showing beyond doubt that a test sample contains the virus is to grow a viable culture.  If you can’t culture a positive sample, and no disease develops in the patient, how can you demonstrate that the PCR test is correctly detecting a virus? This article (referenced by Colin Bannon’s very pro-PCR blog post) states that at Ct = 35, a full 97% of samples can’t be cultured.  That’s not a very good proof of the efficacy of the test.

(From the article: “It can be observed that at Ct = 25, up to 70% of patients remain positive in culture and that at Ct = 30 this value drops to 20%. At Ct = 35, the value we used to report a positive result for PCR, <3% of cultures are positive.“)

Furthermore, the generated polynomial suggests that by Ct = 38, no samples at all will be able to be cultured.  What does that say about the validity of the PCR test at Ct > 40?


Here’s another view – this article (from 2003) is about developing PCR testing for smallpox DNA.  It has the advantage that the samples being used are cloned, and have precisely known quantities of viral DNA.  It’s very long and complicated, but the bottom line is that the ultimate limit of detection was found to be about 12 gene copies, at a Ct of 40.  What, then, would be the meaning of a positive test a Ct=45, which is 32 times more sensitive? There is only a 0.375 probability that there is even a single genome to be found in that sample. 

And so the idea of believing that results of PCR testing at Ct > 40 does look highly suspect – as many experts in the field agree (including Kary Mullis, inventor of the PCR test). 

And yet the WHO were recommending testing at Ct = 50 (over 1000 times as sensitive!) at the outset of the COV-2 pandemic.  One has to wonder why.

But wait – the WHO changed their minds last December! https://www.who.int/news/item/14-12-2020-who-information-notice-for-ivd-users

So now we are surely all good, aren’t we?  There may be a few false positives due to poor test technique, but better safe than sorry, eh?  Alas, not quite. The WHO, while acknowledging the problem, go on to fudge the issue by saying it’s all very complicated, so they can’t make any actual recommendations.

Meanwhile, the actual false positive rate depends hugely on the presence of the virus in the population being tested.  If you only test people with symptoms, your positive results have a high chance of being right.  If you test everybody, and very few people are actually infected, your positive results have a high chance of being wrong.   It’s counter-intuitive, but true.  Here’s a video explaining the maths.

One can tell which situation applies from the test results themselves.  If, as this paper from June and Dr. Bannon’s blog suggests, the vast majority of test results (even in pillar 1) are negative, that strongly suggests that there is low prevalence in the population being tested, and thus that most of the positive results are indeed false positives.

But does any of this matter to patients?  No.  Patients have symptoms and feel ill.  My neighbour’s family has been sick for 2 weeks now, and are still far from well.  If everybody who had a positive test result had been whisked away and quarantined somewhere, would they still have caught the virus?  I don’t know, but it’s obviously wise to try to avoid catching this disease, and if false positives encourage safer behaviour, doesn’t the end justify the means?  Well, maybe – or maybe not, if the lack of honesty on which their advice rests undermines trust in so-called “authority figures”.  Or maybe that trust is misplaced to such a large degree that it needs undermining.

And that’s politics!

Outside in the real world, it’s undeniable* that there are a lot of hospitalisations going on. Hospitals don’t take people in just because they’ve had a positive test result – you need to be properly ill, with symptoms. People in ITU with COVID-19 symptoms is another fairly dependable measure – and even more so, excess deaths, which isn’t dependent on any kind of test results. Excess deaths are up now too. All in all, I think it’s best to believe that whatever the true meaning of PCR test results, there probaby really is a pandemic of some sort going on out there, and take whatever precautions you can, to avoid being a casualty. See my friend Dr. Bannon’s many excellent blog posts on how best to do that.

So to answer the question posed in the title – it’s whatever you want it to be. PCR testing is jolly clever, and certainly useful, whilst not being totally reliable. If you want to look at the science, there’s plenty to talk about.

But if you want to talk about politics, and whether the story reported by mainstream media is being manipulated with sinister intent – fill your boots! Just not with me – for now, I’m going back to excessive drinking.

*It has been pointed out to me that it’s not actually undeniable – just reported everywhere. That’s true. It’s hard to get solid, dependable information: most sources, even if they have no specific editorial agenda, pick a side – either “same as the mainstream” or “rebellious” (often labelled “conspiracy theory”, whether there is a suggested conspiracy or not). Is this years’s hospitalisation rate actually higher than a bad flu year? I’m not sure….

4 Replies to “PCR testing – Science or Politics?”

  1. Hi Pete, PCR positivity in Pillar 2 seems to be running at about 10%. As you say, with symptoms or close contact a test is really useful. Clinical input reduces the chance for error and in a way, the problem is not the test at all, but the lack of clinical context to the results. Mass screening is another matter.

    PCR tests are now, it seems using 3 gene probes, and need 2 to be pos for a positive test which must reduce the false positive rate. Interestingly, B117 seems not to react with the spike probe.

    I do think the dishonesty is coming from those who claim 90% of the tests are wrong and the pandemic is an invention. The usual genuine fleck of concern in a giant bun of nonsense.

    1. There’s dishonesty, and inflammatory behaviour, in both camps IMHO. The government have destroyed many people’s trust in themselves and their “official” scientists by lying, and trying to provoke people into changing their behaviour by using distortions and propaganda instead of honesty; and the “join the dots” people sometimes dispute actual facts to maintain their story, or even make up their own ludicrous “facts” like the “vaccine will change your DNA”.

      So people pick a side according to which side they more want to vilify, and nobody ever changes sides…

  2. Much to agree with here Pete – unfortunately right now it looks pretty awful around the country and with PCR testing, as you say, false positive are now less of an issue when there are lots of real positives around. If positive tests are routinely retested it helps (PHE suggests this at the limit of detection). Back to heavy drinking – schurlee not!! x

    1. Hi Colin – thanks, but I don’t agree that false positives are less of an issue while the test positive rate is as low as 2% (as in the article I referenced). What’s your number? Does anybody report it?

Leave a comment