When will this shit end?

[stuartbert two hats]

5 minutes ago, efcfanwirral said:

true - I find it hard to know what a decent source is - it can't be these types of accounts who do have proper doctors etc following them but aren't "official" sources, but then Independent Sage are a bit of an agenda driven mess. But equally Daily Mail and Guardian etc are completely compromised by hard right or hard left idealogy. 

The Guardian is not hard left!

[danmarks]

The nhs trust i work for is a 9 -10week gap anyway not 12 weeks. This enables a bit of wriggle room cos Nhs almost never runs totally smoothly to be honest.

[JoeyT]

 

[Guest]

A cursory look at the gov dashboard data does indicate a plateauing of cases by specimen date. But this necessarily isn't unexpected, is it? Is this the effect of the form of lockdown we have, given more people are out and about, more kids are in school etc maybe this is getting to the natural level with these types of restrictions.

I'm not saying we should tighten the lockdown or want to get into the argument of is this lockdown lenient again, it's just a thought on why we might be see a plateau.

[tigger123]

52 minutes ago, Toilet Duck said:

Ultimately, case numbers will be immaterial. At the moment, they want to keep an eye on them as there is still a large susceptible population, but that’s dwindling. Once over 45s are sorted, not only have you a significantly lower risk in those left, but also a higher proportion of natural immunity as cases have been higher in that cohort anyway, so it’s a completely different proposition to opening up while you still have large numbers of unprotected, higher risk individuals. While I’m not overly worried about variants, hanging on til you have more people protected reduces the risk from them further. If the vaccines don’t suppress transmission (it looks like they will to a degree), the virus will seasonally circulate anyway, so cases become important when people turn up in hospital (we now have treatment regimens that we use for Covid that are different to other respiratory conditions, so knowing why someone has been admitted will help choose the correct course of treatment). It will also remain a notifiable disease for the foreseeable future, so the same kind of surveillance that goes on in the background for a whole host of other communicable disease will continue, but you won’t see it in the news every day. In short, once you have the higher risk categories protected through vaccination, case number are less informative than hospitalisations/deaths as they should no longer be a predictor of future morbidity/mortality...but the super safe option just in case a weird variant emerges is to make sure the susceptible population is a small as possible.

Great, thanks a lot for this, all makes sense!

[stuartbert two hats]

9 minutes ago, Ozanne said:

A cursory look at the gov dashboard data does indicate a plateauing of cases by specimen date. But this necessarily isn't unexpected, is it? Is this the effect of the form of lockdown we have, given more people are out and about, more kids are in school etc maybe this is getting to the natural level with these types of restrictions.

I'm not saying we should tighten the lockdown or want to get into the argument of is this lockdown lenient again, it's just a thought on why we might be see a plateau.

That's true, but it's disappointing that it's happening so strongly before the schools have even gone back properly.

[Guest]

2 minutes ago, stuartbert two hats said:

That's true, but it's disappointing that it's happening so strongly before the schools have even gone back properly.

It is but let's hope that schools don't lead to that sharp of an increase, it could also be that with schools open this is near the natural level for this type of restrictions.

[eFestivals]

1 hour ago, gizmoman said:

It's amazing how people on here can understand Toilet Duck's detailed and complex explanations of this virus but can't seem to grasp this basic point about the PCR test, every time it's raised it gets ignored or attempts are made to dismiss it. The true false positive rate I.E. lab error or contamination is probably very small, but the number of positives identified from currently uninfected people must be getting quite high given the number who have now had the virus and who would still have traces of the virus in them, the PCR test as it is being used is too sensitive for identifying infectious people.

That's mainly the consequence of PCRs being the best we have, and the govt not wanting to undermine trust in them.

In practice, we don't bother with attempting to get negative results. It's presumed that a positive test is early in the infection cycle and ten days later a person is non-infective.

It's also worth remembering that a high proportion of tests are being done weekly on the same people in govt establishments, so they should be identifying people soon after infection (they stop testing a person who gets a positive result for 90 days because of the chance of further positive results)

So while what you say is in the mix, it's in the mix much less than you're thinking.

[eFestivals]

29 minutes ago, Ozanne said:

A cursory look at the gov dashboard data does indicate a plateauing of cases by specimen date. But this necessarily isn't unexpected, is it? Is this the effect of the form of lockdown we have, given more people are out and about, more kids are in school etc maybe this is getting to the natural level with these types of restrictions.

I'm not saying we should tighten the lockdown or want to get into the argument of is this lockdown lenient again, it's just a thought on why we might be see a plateau.

There'll also be a natural plateau eventually, due to those who can't isolate passing covid between themselves. 

That will inevitably first show itself via a flattening of the curve.

[gizmoman]

19 minutes ago, eFestivals said:

That's mainly the consequence of PCRs being the best we have, and the govt not wanting to undermine trust in them.

In practice, we don't bother with attempting to get negative results. It's presumed that a positive test is early in the infection cycle and ten days later a person is non-infective.

It's also worth remembering that a high proportion of tests are being done weekly on the same people in govt establishments, so they should be identifying people soon after infection (they stop testing a person who gets a positive result for 90 days because of the chance of further positive results)

So while what you say is in the mix, it's in the mix much less than you're thinking.

You have no way of knowing exactly what the false positive rate is though. I'm not suggesting it's a large percentage, if we are doing 800,000 tests a day an error rate of only 0.25% would still give 1000 "cases" , that was the point Bobwillis was making and the WHO have updated their guidance to reflect this problem,

https://www.who.int/news/item/20-01-2021-who-information-notice-for-ivd-users-2020-05

WHO reminds IVD users that disease prevalence alters the predictive value of test results; as disease prevalence decreases, the risk of false positive increases (2). This means that the probability that a person who has a positive result (SARS-CoV-2 detected) is truly infected with SARS-CoV-2 decreases as prevalence decreases, irrespective of the claimed specificity.

 

Is the PCR test 99.75% accurate? If it is I would say that it's incredibly good, We know it's not 100% accurate otherwise the WHO wouldn't have to warn about false positives.

[Mr.Tease]

1 hour ago, danmarks said:

The nhs trust i work for is a 9 -10week gap anyway not 12 weeks. This enables a bit of wriggle room cos Nhs almost never runs totally smoothly to be honest.

Same here, when I booked my first Pfizer shot, the system automatically booked my second appointment for 10 weeks later, so hopefully this disruption shouldn't really take 2nd doses past 12 weeks.

Edited February 17, 2021 by Mr.Tease

[eFestivals]

22 minutes ago, gizmoman said:

You have no way of knowing exactly what the false positive rate is though. I'm not suggesting it's a large percentage, if we are doing 800,000 tests a day an error rate of only 0.25% would still give 1000 "cases" , that was the point Bobwillis was making and the WHO have updated their guidance to reflect this problem,

https://www.who.int/news/item/20-01-2021-who-information-notice-for-ivd-users-2020-05

WHO reminds IVD users that disease prevalence alters the predictive value of test results; as disease prevalence decreases, the risk of false positive increases (2). This means that the probability that a person who has a positive result (SARS-CoV-2 detected) is truly infected with SARS-CoV-2 decreases as prevalence decreases, irrespective of the claimed specificity.

 

Is the PCR test 99.75% accurate? If it is I would say that it's incredibly good, We know it's not 100% accurate otherwise the WHO wouldn't have to warn about false positives.

Even with false positives, they remain the best we have for detecting case levels in the community.

I remember seeing a fair bit about false positives early in the pandemic, so I suspect that it's factored into the modelling that the govt use. 

Just because there's not a song and dance about it doesn't mean it's been forgotten about. The situation deteriorates if the public lose trust in the tests.

[BobWillis]

7 minutes ago, eFestivals said:

Even with false positives, they remain the best we have for detecting case levels in the community.

That’s not the point I’m making in the slightest. I have no issue with the PCR test. 
 

My point is that as there has been talk of cases needing to be under 1,000 for restrictions to be lifted...if the base rate fallacy is above 1,000 cases then we’re going to be stuck in limbo for no reason. 
 

Last year Hancock was quoted as saying the FPR was “less than 1%” some people quoted it as being higher than 1%. We could be using different tests now with higher accuracy but the point stays the same. If cases can never drop below “x” due to the false positive rate then how can we base policy on these numbers. 
 

 

Edited February 17, 2021 by BobWillis

[BobWillis]

Obviously as the prevalence gets lower and lower then the government could mitigate the false positive rate by retesting every positive result, the chances of two false positives is incredibly small. 
 

They may already be doing this, I don’t know. 

[eFestivals]

3 minutes ago, BobWillis said:

That’s not the point I’m making in the slightest. I have no issue with the PCR test. 
 

My point is that as there has been talk of cases needing to be under 1,000 for restrictions to be lifted...if the base rate fallacy is above 1,000 cases then we’re going to be stuck in limbo for no reason. 
 

Last year Hancock was quoted as saying the FPR was “less than 1%” some people quoted it as being higher than 1%. We could be using different tests now with higher accuracy but the point stays the same. If cases can never drop below “x” due to the false positive rate then how can we base policy on these numbers. 

The relevance of case numbers ends as the vaccine effect starts.

You can find people saying every possibility for how we should exit. It doesn't mean they're correct.

[eFestivals]

6 minutes ago, BobWillis said:

Obviously as the prevalence gets lower and lower then the government could mitigate the false positive rate by retesting every positive result, the chances of two false positives is incredibly small. 
 

They may already be doing this, I don’t know. 

I know they follow up a positive LF test with a PCR test.

And you might be onto something for where testing goes in the short term, with double testing. In the longer term we're unlikely to care about the virus circulating.

[Toilet Duck]

23 minutes ago, BobWillis said:

Obviously as the prevalence gets lower and lower then the government could mitigate the false positive rate by retesting every positive result, the chances of two false positives is incredibly small. 
 

They may already be doing this, I don’t know. 

Yep, samples are run in duplicate. At least, they are here and that’s the standard CDC/WHO protocol that everyone is following. 

[BobWillis]

19 minutes ago, eFestivals said:

I know they follow up a positive LF test with a PCR test.

And you might be onto something for where testing goes in the short term, with double testing. In the longer term we're unlikely to care about the virus circulating.

I think they believe they’ve got a supply of very accurate LFT to replace the innova ones so the retesting with a pcr test might be ending. 
 

I completely agree that cases become irrelevant once the vaccine effect takes hold, I’ve been saying it for a while but I don’t believe they’ll stop focusing on case numbers for quite a while yet which is why I worry about the false positive rate. Especially as they look to bring in even higher levels of mass testing. 

[gizmoman]

10 minutes ago, Toilet Duck said:

Yep, samples are run in duplicate. At least, they are here and that’s the standard CDC/WHO protocol that everyone is following. 

But we're not talking about lab error, it's the fact the test picks up tiny fragments of possibly old covid, if someone has no symptoms and tests positive you would run a separate test from a new swab to confirm. Is that done?

[eFestivals]

14 minutes ago, gizmoman said:

 it's the fact the test picks up tiny fragments of possibly old covid

In real life, that wouldn't happen very much.

The people being tested are generally being weekly tested where they will only pick up new covid, or people are presenting themselves for testing because of symptoms.

[MrBarry465]

Reading through the complaints of people who are in hotel quarantine, some of them were oblivious to the fact they even needed to do it and are kicking off.

What planet are some of these on? Do they realise it's a pandemic currently, or should they just be able to nip over from Brazil?

 

[Toilet Duck]

48 minutes ago, gizmoman said:

But we're not talking about lab error, it's the fact the test picks up tiny fragments of possibly old covid, if someone has no symptoms and tests positive you would run a separate test from a new swab to confirm. Is that done?

There wouldn't be much point as they would likely get the same result (sampling error has more impact on false negatives). What's being conflated here (not by you) is the error rate of the test for detecting nucleic acids (very low) and the difference between infected and infectious (which is hard to tell!...CT values aren't entirely quantitative, but even when we use those, samples with very high CTs can still have sufficient virus in them to infect cells in the lab...how that translates to an ability infect others is another matter altogether!). We've been through this before, so you know I agree that the test can't necessarily tell you if you are infectious at the moment the test was conducted. If you have a low apparent viral load, it could be because you are that start of an infection and may be infectious already or may be in a couple of days (another test after a few days will help decipher that one). It could also be because you did have an infection, but are now over it and what is being detected are fragments of left over dead virus (these can hang around for a bit, but don't usually show up on the test where those with a very high viral load show up). Or it could be because of a sampling error, in which case, a repeat test will help rule that out (you would probably need to be symptomatic to justify that though and as noted, false negatives would be the more likely outcome of a sampling error). So there's two errors (technical, the inherent error in the test, which is corrected for by running samples in duplicate (and is very low); and biological, which is due to sampling and could only be corrected for by a repeat test). Correcting for both will improve the accuracy of detection but still won't answer the infected vs infectious question (properly conducted and validated antigen tests are better for this question). It remains important to note that the false positive rate for detecting viral nucleic acids is incredibly low and a positive result in and of itself is still useful information. If the patient is in hospital, then we can treat them accordingly, and it's also useful for contract tracing and cluster identification (even if you are no longer infectious, those you were recently contact with might be so identifying them helps control outbreaks and breaks transmission chains). As things progress this year, case numbers will become less of an issue, especially once we reach a critical mass of vaccinations (coupled with some natural immunity) and the susceptible population gets smaller and smaller. 
 

edit: meant to say high CTs, low viral load (the higher they are, the fewer copies of viral nucleus acid is present).. 

Edited February 17, 2021 by Toilet Duck

[gizmoman]

6 minutes ago, eFestivals said:

In real life, that wouldn't happen very much.

The people being tested are generally being weekly tested where they will only pick up new covid, or people are presenting themselves for testing because of symptoms.

That's simply not true, in my area the council are going door to door to encourage people to get a test (and have been for months), they have also done the mass testing in Liverpool and the recent "surge" testing for the new variants. They seem very keen to find as many cases as possible. This is why this is an issue, it makes no difference as far as healthcare goes if there are extra false positives as long as genuine cases are identified. The problem is if these inflated case numbers are used to justify damaging restrictions.

[fraybentos1]

6 minutes ago, gizmoman said:

That's simply not true, in my area the council are going door to door to encourage people to get a test (and have been for months), they have also done the mass testing in Liverpool and the recent "surge" testing for the new variants. They seem very keen to find as many cases as possible. This is why this is an issue, it makes no difference as far as healthcare goes if there are extra false positives as long as genuine cases are identified. The problem is if these inflated case numbers are used to justify damaging restrictions.

Why must you constantly suggest there is some sort of conspiracy going on to keep us locked down? Why would the gov do that? The same gov who like 6 months ago did the eat out to help out scheme.

[stuartbert two hats]

4 minutes ago, fraybentos1 said:

Why must you constantly suggest there is some sort of conspiracy going on to keep us locked down? Why would the gov do that? The same gov who like 6 months ago did the eat out to help out scheme.

And then catastrophically finished the second lockdown too soon, costing thousands of lives.