Meet the Other Phone. Flexible and made to last.

Meet the Other Phone.
Flexible and made to last.

Buy now

Please or to access all these features

Covid

Mumsnet doesn't verify the qualifications of users. If you have medical concerns, please consult a healthcare professional.

Patrick Vallance clip

56 replies

fungussingstheblues · 21/02/2021 16:38

Patrick Vallance is admitting in a clip I can't seem to attach that PCR tests can return positive results when a person isn't infectious.

It's from the daily briefing on 10 February.

What proportion of tests does this? Pretty crucial to know the answer, isn't it, considering the entire lockdown policy of the country has been based on these tests for more than a year now??

Hold on and I'll write the text verbatim.

OP posts:
fungussingstheblues · 21/02/2021 16:41

"If you compare lateral flow tests to PCR, you'll get more positives with PCR, cos PCR are very, very sensitive, they can pick up very low levels of virus, which may not even be infectious. It may just be low levels there."

OP posts:
covetingthepreciousthings · 21/02/2021 16:43

I thought this was quite well known? Which is why they advice you shouldn't have a second PCR test however many days after a first positive one (I think 90?) as it could still show positive.

This is why the isolation period is for 10 days, and not until you get a negative test as it likely wouldn't happen.

SonnetForSpring · 21/02/2021 16:43

What is your point? The scientists and government are of course aware there is a percentage of false positives. No test is 100% accurate.

fungussingstheblues · 21/02/2021 16:54

What's the percentage?

OP posts:
SonnetForSpring · 21/02/2021 16:56

I don't know but believe or not, I'm not one of the people who need to know in order to make decisions!!! Seriously, get a grip. These people are educated to an extremely high level and test accuracy is a basic concept!

sleepwouldbenice · 21/02/2021 16:58

@SonnetForSpring

I don't know but believe or not, I'm not one of the people who need to know in order to make decisions!!! Seriously, get a grip. These people are educated to an extremely high level and test accuracy is a basic concept!
Yep, this

Do you really think that they don’t take this into account

Useruseruserusee · 21/02/2021 16:59

@fungussingstheblues

"If you compare lateral flow tests to PCR, you'll get more positives with PCR, cos PCR are very, very sensitive, they can pick up very low levels of virus, which may not even be infectious. It may just be low levels there."
So the person has the virus, so it’s not a false positive?
doublehalo · 21/02/2021 17:03

This is old news. The PCR test gives lots of false positives.

Blaggingit123 · 21/02/2021 17:04

Exactly as pp says, he is referring to PCR tests picking up dead virus cells as a positive test when the person isn’t infectious any more.

If you’re reading into it that he thinks there’s a chance that some people who have it aren’t infectious, that’s pie in the sky. Viruses work by being infectious in all people who catch it. A virus that isn’t infectious isn’t a virus at all. The 10 day isolation period is an average period of how long people are infectious for and is nothing to do with how long a PCR test will identify the virus.

RosieLemonade · 21/02/2021 17:05

Thanks you have just rememinded me I need to test tonight.

OP posts:
SonnetForSpring · 21/02/2021 17:37

The point is not the number of positive tests....that is merely an indicator of the potential number of hospitalisations and resulting deaths. We are more concerned about changes in the number of positive tests... and the outcomes associated with that. This sounds to be like more conspiracy theory nonsense. We know that the figures are always going to have a level of inaccuracy. All figures do. But what we do know is the vague relationship between all the figures and that us what is being modeled both forwards and backwards. The are approx indicators. They give us information to help make decisions. I still do not understand your point.

titchy · 21/02/2021 17:44

The actual number of positive PCR tests doesn't matter. What matters is the relationship between positive tests and hospitalisation, and the time lag between those two factors.

If we know that 10,000 positives a day will lead to 1,000 admissions in two weeks time (numbers made up) then it won't matter how many of those positives were false. We can extrapolate that once we're down to 1,000 positives that will only lead to 100 admissions a fortnight later.

Although most positives from the community will be genuine - people only get a test if they have symptoms.

donquixotedelamancha · 21/02/2021 17:46

Patrick Vallance is admitting in a clip I can't seem to attach that PCR tests can return positive results when a person isn't infectious.

I'm willing to bet that isn't what he says. I'd bet my house what he actually says is around the fact that the PCR tests for Covid DNA and so theoretically someone might not be shedding at that precise moment.

What proportion of tests does this? Pretty crucial to know the answer, isn't it, considering the entire lockdown policy of the country has been based on these tests for more than a year now??

No, to just about every part of that. Our lockdown policy has been based on all the people dying, we didn't even have mass testing when it began.

If there is a systematic error within the PCR test compared to those who are substantially contagious (and there is bound to be some) it doesn't matter- that error will be consistent and so the 'false' Covid rate data can still be used to predict death/hospitalisation rates.

It may be that some people have isolated who could have got away with not doing but since there isn't a more accurate test there is no way to avoid that.

PracticingPerson · 21/02/2021 17:50

[quote fungussingstheblues]What if it's about a 90% false positive rate?
www-nytimes-com.cdn.ampproject.org/v/s/www.nytimes.com/2020/08/29/health/coronavirus-testing.amp.html?amp_js_v=a6&amp_gsa=1&usqp=mq331AQHKAFQArABIA%3D%3D#aoh=16139283188074&referrer=https%3A%2F%2Fwww.google.com&amp_tf=From%20%251%24s&ampshare=https%3A%2F%2Fwww.nytimes.com%2F2020%2F08%2F29%2Fhealth%2Fcoronavirus-testing.html[/quote]
If it is a 90% false positive rate then the number of deaths we have had means covid is really deadly

It isn't a 90% false positive rate btw.

bobbiester · 21/02/2021 17:55

Look at New Zealand. They administer loads of PCR tests (1.6 million in total) and hardly ever get positive results.

If the false positive rate was a significant problem New Zealand would keep on erroneously detecting many cases.

ExcusesAndAccusations · 21/02/2021 17:59

They’re not false positives in the case that they pick up people who never had Covid, these are people who had the disease but might no longer be infectious. But our testing regime is still mostly driven by testing people with symptoms, who will almost certainly be infectious at the stage they test positive.

As people have cogently said above, the UK policy uses positive tests to show what’s coming in terms of hospitalisation and deaths, and responds accordingly. If the strategy was flawed in the way the OP implies then the huge surge in positive tests in December would not have been followed by an increase in very sick people in hospital with Covid symptoms and then deaths. But they were, and illness and deaths tracked positive tests just as predicted.

If we do move to a mass policy of completely random testing then the decision makers would need to think hard about the impact the change would have on the data and adjust assumptions accordingly. But as also pointed out upthread, they’re not idiots.

donquixotedelamancha · 21/02/2021 20:07

If the strategy was flawed in the way the OP implies then the huge surge in positive tests in December would not have been followed by an increase in very sick people in hospital with Covid symptoms and then deaths.

From another thread, OP doesn't believe there has been a second wave. She thinks everyone vulnerable died in spring and death rates have been normal since. How that would fit with false positives causing prevelence to be over estimated I can't fathom.

Nerdygirl · 21/02/2021 21:39

It does also depend on the number of cycles. . If you have higher number of cycles you basically can find a lot of very okd dead virus traces . The WHO recommended cycles were reduced because of this and this has caused lots of lower numbers in positive tests .

It’s right to question this just like it’s right to question why the death numbers are based on within 28 days of a positive test. Also it’s right to question why we are not seeing stats on deaths within 28 days of a positive test and whether they have been vaccinated or not

Matt Hancock again has been caught lying with the vitamin d analysis he didn’t actually do despite saying it . So it’s right to question the government , and the basis of lockdown on things like pcr tests as the op is doing

pinkhappy · 21/02/2021 22:14

Patrick Vallance is admitting in a clip I can't seem to attach that PCR tests can return positive results when a person isn't infectious.

This is not new. If you have recently had covid but are no longer infectious the PCR test may come back as positive. In that situation the lateral flow test is likely to come back as negative so is better. On the other hand if you are presymptomatic but infectious the PCR test is likely to come back positive. In that situation the lateral flow test may come back negative so is worse.

Having said this, you will probably know if you recently had covid so this doesn't affect many people.

CoffeeandCroissant · 21/02/2021 22:23

@Nerdygirl

It does also depend on the number of cycles. . If you have higher number of cycles you basically can find a lot of very okd dead virus traces . The WHO recommended cycles were reduced because of this and this has caused lots of lower numbers in positive tests .

It’s right to question this just like it’s right to question why the death numbers are based on within 28 days of a positive test. Also it’s right to question why we are not seeing stats on deaths within 28 days of a positive test and whether they have been vaccinated or not

Matt Hancock again has been caught lying with the vitamin d analysis he didn’t actually do despite saying it . So it’s right to question the government , and the basis of lockdown on things like pcr tests as the op is doing

This is incorrect. The WHO didn't recommend anything of the sort.

The WHO notice does not advise laboratories to reduce the cycle count in PCR tests. It advises laboratory professionals to use tests with the proper instructions to ensure accurate results.
www.reuters.com/article/uk-factcheck-who-instructions-pcr-guidan-idUSKBN2A429W

The advice boils down to: new to PCR? Read the instructions and understand the purpose of testing. But of course, this lab-focussed advice has been taken by those with malicious intent, or with too little understanding of the topic, and blown it up into something else entirely wrong.
virologydownunder.com/who-to-new-pcr-users-read-the-damned-manual/

TravellingTilbury · 21/02/2021 22:58

@doublehalo

This is old news. The PCR test gives lots of false positives.
Exactly, it has always been an issue - particularly with tests where the actual incidence of an 'event' (ie presence of disease) is low.

I love the topic of false positives - part of signal detection theory.

Signal Detection Theory says that there are four possible outcomes: Hits, Misses (ie false negatives), False Alarms (ie false positives) and Correct Rejections.

Think of it like a grid. The columns (going down) will represent an 'event' ie an actual thing that is present eg being ill. The first column would represent the people who are actually ill and the second column would represent the people who aren't actually ill.

Then you have the rows. These represent the test - the 'signal'. The first row includes the people who test 'positive' and the second row the people who test 'negative'. If someone 'has' something and a signal is detected (ie tests positive), that is a 'Hit'. Underneath that, if someone 'has' something but tests negative that is a 'Miss' (a 'false negative' if you like).

The second column includes someone who isn't ill ie there is no 'event' yet there is a 'signal' - ie they still tested positive - which is the false alarm (ie a 'false positive') and then under that you have people who were neither sick nor tested positive for the 'signal' - these are the 'correct rejections'.

I'm happy to demonstrate with figures but it can get confusing to explain without drawing it.

But say you've got 200 sick people out of 100,000 - a low incidence. Therefore 99,800 healthy people.

Imagine the false positive rate is 1% (we don't know it). That means 1% of the 99,800 healthy people will have been falsely told they were sick ie 998 people. We'll come back to this later.

Say the test was 70% effective (again we don't know how effective/sensitive the tests are but I'll be prudent and say 70%). So 70% of the 200 people were correctly told they were sick when they were actually sick (these are the 'hits') ie 140 people. Well, that's okay, better than nothing.

Sadly 30% of the sick people were 'missed' (ie false negative) so sadly there would be 60 sick people who were told they were fine. Not great but they could always retest. And 60 out of 100,000 isn't too much of a problem.

Back to the healthy people - the 99,800. Out of the healthy people, 99% out of 99,800 were 'correct rejections' ie they were healthy and no 'signal was detected' so they correctly tested negative = 98,802 people. So that's good.

But hang on, 998 people were incorrectly told they were sick when they weren't. And actually only 200 people were sick (because incidence is low)...

See the issue? Without clinical diagnosis there is potential for way more false positive 'cases' than actual 'cases'.

The PCR test was never intended to be used for diagnosis of a virus in the absence of symptoms. It's a scandal and everyone who has studied this knows this.

I should add that if the PCR was voluntary - and simply used as a screening aid - fine. Or, if each 'case' had a 50% death rate or something then again, perhaps it would be better to be safe than sorry.

But in this case, to use these inflated PCR 'case' numbers to decide government, health and economic policy is knowingly misleading.

Related to this is the issue of the fluctuating cycle threshold that will influence how sensitive the test is ...

Bah! Don't get me started on PCR tests Grin

NB The history signal detection theory is v interesting and I believe it has been used in eye witness testimony and to test signals in warfare. There is no use in having a test (or eyewitness) that is over sensitive and says 'yes' to everything - it has to be balanced against the noise. Imagine if pregnancy tests were over sensitive and were telling men they were pregnant etc when they weren't.

In my opinion, the PCR test should not be used without clinical diagnosis, unless it is being used merely as a screening tool.

TravellingTilbury · 21/02/2021 23:03

Kary Mullis talking about the limitations of using the PCR for diagnostics (before he died in 2019):

TravellingTilbury · 21/02/2021 23:17

@bobbiester

Look at New Zealand. They administer loads of PCR tests (1.6 million in total) and hardly ever get positive results.

If the false positive rate was a significant problem New Zealand would keep on erroneously detecting many cases.

It would depend on the cycle threshold - higher ct, more false positives, lower ct, fewer false positives.
Please create an account

To comment on this thread you need to create a Mumsnet account.

This thread is closed and is no longer accepting replies. Click here to start a new thread.

Swipe left for the next trending thread