More Boris lockdown breaches

[Scott777]

a false positive rate of 1% (implied by a specificity of 99%) would yield at least 2,000 positive results from a pool of 207730 PCR tests.  It didn't so the specificity cannot be as low as 99%

Not correct.  You would expect, as an estimate, to see 2000 positive results.  It doesn't mean that's what you would get.  It just means you have to subtract 2000 from the actual result, because they cannot be taken at face value, so none of the 159 positives can be taken as true positives.

[Scott777]

The grammatical nonsense is in your head.

First let's.look at the entire paragraph.
The first word " evaluation" is a link to the lancet study of 5 tests. (The ones I detailed earlier)

It then says "including the ELISA assay used in this study".

So it is now referring to an ELISA study, which is in the study linked.

So the "this" will (according to the rules of grammar you are pinning your hopes on) refer to them last thing mentioned - in this case the "ELISA assay used in this study" - which we know is the Oxford Study in the lancet study.  And to back this up, the figures given match the figures for the Oxford Study.

So we can be pretty sure that the study "this" refers to.is the Oxford Study.

They then say:
And those figures happen to match the Siemens results given in the lancet study (the one linked to in the first word of the paragraph) - which also happens to be the best of the 4 commercial assays tested in that study.

The two scenarios were run to study the impact of differing selectivity and specificity on the results. This is standard practice.

In both scenarios they use "high specificity", defined as 99.9% in scenario 1.
And scenario 2 is also "high specificity"

Here are the problems with your interpretation:

If you were correct, using correct grammar, the article SHOULD say "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology has shown that the ELISA assay, as included in our study, has sensitivity and specificity...".  That way, we know exactly what they are referring to.  In the actual sentence, the Elisa assay is within a clause, and so not the main subject of the sentence.

"In both scenarios they use "high specificity", defined as 99.9% in scenario 1."  This is nonsense, because the 99.9% would not be contained ONLY within scenario 1 if it applied to both scenarios.

As it stands, you are suggesting NO specificity has been given for scenario 2, and that the specificity in scenario 1 is just a single number, without any range of variability, or margin of error.  Despite the five immunoassays ALL showing a range of variability, AND this "To allow for the fact that individuals are self-swabbing...  provides a lower bound."  So scenario 2 is explicitly intended to provide a lower bound, and yet you claim there is just a single number for the specificity.

Finally, "We have used Bayesian analysis to calculate what prevalence would be in two different scenarios".  Do you understand, if you apply the exact same specifity to a quantity of positive test results, then you get the same number of false positives, and therefore the same number of true positives, then the prevalence would be the same, contradicting the purpose of the 2 scenarios.

So, pray tell, why is the specificity not stated for scenario 2, why is the specificity without any range of error, and why would both scenarios have the same specificity if they are for measuring the possible range of prevalence?

[BeElBeeBub]

What 207730 PCR tests are you referring to?

Read the link I provided.

Here it is again.

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/methodologies/covid19infectionsurveypilotmethodsandfurtherinformation#test-sensitivity-and-specificity

[BeElBeeBub]

So you have found an interesting point, that the figures for 2 of the blood assays are identical to the figures in the ONS study.  It's interesting, but doesn't prove a thing.  There are still a number of flaws in your conclusion.  The ONS does not refer to the Siemens study when talking about a best performing commercial assay.  So although you may ASSUME it is referring to that, it isn't necessarily the case, just because the figures are the same.

Then the other figures, you say refer to the Elisa assay, despite it not stating that at all.  As I have said, you are claiming the ONS has made grammatical / linguistic errors, by referring to the Elisa assay without saying so, but only saying "this assay".  Looking at the sentence, there is a clause, which is grammatically not essential as part of the sentence.  Removing that clause, we have "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology has shown that this assay has sensitivity and specificity....".  So you claim this sentence is referring to an assay ("this assay") which we have no way to know about, unless it is referring to the PCR assay which is the main subject of the study.  It is grammatical NONSENSE to refer to an assay which is not the focus of the sentence, or an essential part, or part of the subject matter, by using the words "this assay".

Finally, the study makes it clear they have provided 2 scenarios, and each has a different specificity, so why have you picked the scenario 1 outcome, and simply ignored scenario 2?  That is your entire basis for concluding the specificity is 99.9%.  I shall assume the ONS are not grammatical and linguistic morons, and that it is coincidence that "this assay" happens to have the same figures as the Elisa assay.  You however, assume it is the same assay, and only scenario 1 is correct.  I'm afraid my conclusion is the only plausible one, unless you can explain the grammatical nonsense.

The grammatical nonsense is in your head.

First let's.look at the entire paragraph.

Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study,..

The first word " evaluation" is a link to the lancet study of 5 tests. (The ones I detailed earlier)

It then says "including the ELISA assay used in this study".

So it is now referring to an ELISA study, which is in the study linked.

...has shown that this assay has sensitivity and specificity (95% confidence interval) of 99.1% (97.8 to 99.7%)....

So the "this" will (according to the rules of grammar you are pinning your hopes on) refer to them last thing mentioned - in this case the "ELISA assay used in this study" - which we know is the Oxford Study in the lancet study.  And to back this up, the figures given match the figures for the Oxford Study.

So we can be pretty sure that the study "this" refers to.is the Oxford Study.

They then say:

....and 99.0% (98.1 to 99.5%) respectively; compared with 98.1% (96.6 to 99.1%) and 99.9% (99.4 to 100%) respectively for the best performing commercial assay.

And those figures happen to match the Siemens results given in the lancet study (the one linked to in the first word of the paragraph) - which also happens to be the best of the 4 commercial assays tested in that study.

The two scenarios were run to study the impact of differing selectivity and specificity on the results. This is standard practice. 

In both scenarios they use "high specificity", defined as 99.9% in scenario 1.

Scenario 1 (medium sensitivity, high specificity)

Based on similar studies, the sensitivity of the test used is plausibly between 85% and 95% (with around 95% probability) and, as noted earlier, the specificity of the test is above 99.9%

And scenario 2 is also "high specificity"

Scenario 2 (low sensitivity, high specificity)

To allow for the fact that individuals are self-swabbing, Scenario 2 assumes a lower overall sensitivity rate of on average 60% (or between 45% and 75% with 95% probability), incorporating the performance of both the test itself and the self-swabbing. This is lower than we expect the true value to be for overall performance but provides a lower bound.

So also using 99.9%

They then discuss the effects of these scenarios (basically not much).

Then in the paragraph you have picked in isolation and employed a skewed and unsupported interpretation of, they discuss the selectivity and specificity of the blood tests (not the PCR tests)

Time and time again, you grasp at straws and try and torture the interpretation of a straightforward text to fit your preconceived view.

The ONS state the specificity of the PCR tests is very high, at least 99.9%

Test specificity
Test specificity measures how often the test correctly identifies those who do not have the virus, so a test with high specificity will not have many false-positive results.
We know the specificity of our test must be very close to 100% as the low number of positive tests in our study over the summer of 2020 means that specificity would be very high even if all positives were false. For example, in the six-week period from 31 July to 10 September 2020, 159 of the 208,730 total samples tested positive. Even if all these positives were false, specificity would still be above 99.9%.

We know that the virus was still circulating at this time, so it is extremely unlikely that all these positives are false. However, it is important to consider whether many of the small number of positive tests we do have might be false. There are two main reasons we do not think that is the case.

Symptoms are an indication that someone has the virus; but are reported in a minority of participants at each visit. We might expect that false-positives would not report symptoms or might report fewer symptoms (because the positive is false). Overall, therefore, of the positives we find, we would expect to see most of the false-positives would occur among those not reporting symptoms. If that were the case, then risk factors would be more strongly associated with symptomatic infections than without reported symptoms infections. However, in our data the risk factors for testing positive are equally strong for both symptomatic and asymptomatic infections.

Assuming that false-positives do not report symptoms, but occur at a roughly similar rate over time, and that among true-positives the ratio with and without symptoms is approximately constant, then high rates of false-positives would mean that, the percentage of individuals not reporting symptoms among those testing positive would increase when the true prevalence is declining because the total prevalence is the sum of a constant rate of false-positives (all without symptoms) and a declining rate of true-positives (with a constant proportion with and without symptoms). However, this is not what our data shows 

[Scott777]

In addition, a false positive rate of 1% (implied by a specificity of 99%) would yield at least 2,000 positive results from a pool of 207730 PCR tests.  It didn't so the specificity cannot be as low as 99%

What 207730 PCR tests are you referring to?

[Scott777]

The lancet study looked at 5 blood tests.for antibodies

Abbott assay sensitivity was 92·7% (95% CI 90·2–94·8) and specificity was 99·9%(99·4–100%)
DiaSorin assay sensitivity was 96·2% (94·2–97·7) and specificity was 98·9% (98·0–99·4)
Oxford immunoassay sensitivity was 99·1% (97·8–99·7) and specificity was 99·0% (98·1–99·5)
Roche assay sensitivity was 97·2% (95·4–98·4) and specificity was 99·8% (99·3–100)
Siemens assay sensitivity was 98·1% (96·6–99·1) and specificity was 99·9% (99·4–100%).

The "Oxford" test was a "novel 384-well ELISA" test.
I've broken the para into lines.

Look at the ELISA assay they used (the Oxford one) on the 2nd line.  It matches the figures for the Oxford assy in the lancet study.

The figures on the 3rd line match the figures from the Siemens study, which was the "best performing commercial assay"

So we have the source for both sets of specificity figures you gave and in particular the 99% you referenced here

In addition, a false positive rate of 1% (implied by a specificity of 99%) would yield at least 2,000 positive results from a pool of 207730 PCR tests.  It didn't so the specificity cannot be as low as 99%

Again you have come to the wrong conclusion based on an erroneous reading of studies and reports

Stop and have a think that if you did it here, how many other times might you have done it and how many other conclusions that you have come to may be flawed. Having looked at your previous posts on other topics - I'd say several.

So you have found an interesting point, that the figures for 2 of the blood assays are identical to the figures in the ONS study.  It's interesting, but doesn't prove a thing.  There are still a number of flaws in your conclusion.  The ONS does not refer to the Siemens study when talking about a best performing commercial assay.  So although you may ASSUME it is referring to that, it isn't necessarily the case, just because the figures are the same.

Then the other figures, you say refer to the Elisa assay, despite it not stating that at all.  As I have said, you are claiming the ONS has made grammatical / linguistic errors, by referring to the Elisa assay without saying so, but only saying "this assay".  Looking at the sentence, there is a clause, which is grammatically not essential as part of the sentence.  Removing that clause, we have "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology has shown that this assay has sensitivity and specificity....".  So you claim this sentence is referring to an assay ("this assay") which we have no way to know about, unless it is referring to the PCR assay which is the main subject of the study.  It is grammatical NONSENSE to refer to an assay which is not the focus of the sentence, or an essential part, or part of the subject matter, by using the words "this assay".

Finally, the study makes it clear they have provided 2 scenarios, and each has a different specificity, so why have you picked the scenario 1 outcome, and simply ignored scenario 2?  That is your entire basis for concluding the specificity is 99.9%.  I shall assume the ONS are not grammatical and linguistic morons, and that it is coincidence that "this assay" happens to have the same figures as the Elisa assay.  You however, assume it is the same assay, and only scenario 1 is correct.  I'm afraid my conclusion is the only plausible one, unless you can explain the grammatical nonsense.

[BeElBeeBub]

To help you out a bit, let me break it down for you.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study," - this refers to serology, then the next part "has shown that this assay has sensitivity and specificity..." is now referring to a single assay (cannot be the 5 assays again) there is a single PCR assay.

The lancet study looked at 5 blood tests.for antibodies

Abbott assay sensitivity was 92·7% (95% CI 90·2–94·8) and specificity was 99·9%(99·4–100%)
DiaSorin assay sensitivity was 96·2% (94·2–97·7) and specificity was 98·9% (98·0–99·4)
Oxford immunoassay sensitivity was 99·1% (97·8–99·7) and specificity was 99·0% (98·1–99·5)
Roche assay sensitivity was 97·2% (95·4–98·4) and specificity was 99·8% (99·3–100)
Siemens assay sensitivity was 98·1% (96·6–99·1) and specificity was 99·9% (99·4–100%). 

The "Oxford" test was a "novel 384-well ELISA" test.

including the ELISA assay used in our study, has shown that this assay has

sensitivity and specificity (95% confidence interval) of 99.1% (97.8 to 99.7%) and 99.0% (98.1 to 99.5%) respectively;

compared with 98.1% (96.6 to 99.1%) and 99.9% (99.4 to 100%) 

I've broken the para into lines.

Look at the ELISA assay they used (the Oxford one) on the 2nd line.  It matches the figures for the Oxford assy in the lancet study.

The figures on the 3rd line match the figures from the Siemens study, which was the "best performing commercial assay"

So we have the source for both sets of specificity figures you gave and in particular the 99% you referenced here

That's not how specificity is actually calculated.  They explain their upper and lower bounds for specificity.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study, has shown that this assay has sensitivity and specificity (95% confidence interval) of 99.1% (97.8 to 99.7%) and 99.0% (98.1 to 99.5%) respectively;"

So it's quite possible the specificity was 98.1%, but most likely estimate was 99%.  At 99%, your 500m tests would produce an expected 5m false positives out of the 20m.  However, it's POSSIBLE the specificity was only 98.1%, so the expected false positives would be 9.5m out of 20m, so about half.  We are of course assuming this was the same test and same CT used in your 500m tests, but you didn't specify how you know it was.

In addition, a false positive rate of 1% (implied by a specificity of 99%) would yield at least 2,000 positive results from a pool of 207730 PCR tests.  It didn't so the specificity cannot be as low as 99%

Again you have come to the wrong conclusion based on an erroneous reading of studies and reports 

Stop and have a think that if you did it here, how many other times might you have done it and how many other conclusions that you have come to may be flawed. Having looked at your previous posts on other topics - I'd say several.

[Scott777]

"Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study, has shown that this assay has sensitivity and specificity..."  This means 5 assays were used in blood tests to determine the specificity of THIS assay (in the singular), which is the PCR assay.  I.e., serology is part of the process to find the PCR specificity.

To help you out a bit, let me break it down for you.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study," - this refers to serology, then the next part "has shown that this assay has sensitivity and specificity..." is now referring to a single assay (cannot be the 5 assays again) there is a single PCR assay.

[Scott777]

You are really grasping at straws here.

Assay can refer to any test.  However in the context of the above para it refers to blood tests for antibodies.

How can I say this?

The word "evaluation" in the paragraph is a link which is to this Lancet paper.

https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30634-4/fulltext

That's the study to which the rest of the para refers to.

And the summary says
Serology meaning blood tests and not nasal swabs. Immunoassay meaning antibody not DNA fragment.


And to really put the nail in the coffin


Blood samples meaning, well, blood samples and not to be an any way confused with nasal samples for PCR tests.

Look you've obviously brought a butter knife to an intellectual gunfight so quit before you embarrass yourself any further.

The false positive rate of PCR tests is far to small to have any effect on the overall figures. The classification of "due to covid" deaths *only* includes deaths a doctor has certified the underlying cause as being COVID.

Will you look at the data provided, and reassess your positions in light of new evidence? I doubt it. You'll either ignore it or come up with yet more spurious *evidence" based on your misunderstanding of science. Anything to avoid thinking "F@@@, maybe I was.wrong and all these experts might have been right"

Still wrong.  Try to read it carefully.  In the context, it is referring to TWO different kinds of assays.  As I said (and which you did't seem to understand), blood tests for antibodies were used to help find the accuracy of the PCR assay.  The specificity at the end is NOT referring to the 5 serology assays.  That's obvious because it says "has shown that this assay has sensitivity and specificity...", where "this assay" is singular, and therefore NOT referring to the 5 blood-test assays, which would be plural.  Really do pay attention, before you nail the coffin.

[BeElBeeBub]

Assay can refer also to PCR.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study, has shown that this assay has sensitivity and specificity..."  This means 5 assays were used in blood tests to determine the specificity of THIS assay (in the singular), which is the PCR assay.  I.e., serology is part of the process to find the PCR specificity.

You are really grasping at straws here.

Assay can refer to any test.  However in the context of the above para it refers to blood tests for antibodies.

How can I say this?

The word "evaluation" in the paragraph is a link which is to this Lancet paper.

https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30634-4/fulltext

That's the study to which the rest of the para refers to.

And the summary says 

....Serology is considered key to population-level surveillance and potentially individual-level risk assessment. However, immunoassay performance has not been compared on large, identical sample sets. We aimed to investigate the performance of four high-throughput commercial SARS-CoV-2 antibody immunoassays and a novel 384-well ELISA......

Serology meaning blood tests and not nasal swabs. Immunoassay meaning antibody not DNA fragment.


And to really put the nail in the coffin 

We derived sensitivity and specificity from 976 pre-pandemic blood samples

Blood samples meaning, well, blood samples and not to be an any way confused with nasal samples for PCR tests.

Look you've obviously brought a butter knife to an intellectual gunfight so quit before you embarrass yourself any further.

The false positive rate of PCR tests is far to small to have any effect on the overall figures. The classification of "due to covid" deaths *only* includes deaths a doctor has certified the underlying cause as being COVID.

Will you look at the data provided, and reassess your positions in light of new evidence? I doubt it. You'll either ignore it or come up with yet more spurious *evidence" based on your misunderstanding of science. Anything to avoid thinking "F@@@, maybe I was.wrong and all these experts might have been right"

[Unlucky4Sum]

Seems Scotty can't remember what he posted

 . . . using PCR which has a pathetic level of accuracy . . .

The only thing that has a pathetic level of accuracy are Scotty's posts

[Scott777]

The quote you posted refers to the specificity of the blood tests for antibodies (the terms assay and serology are the giveaway - plus they directly describe the ELISA test as a blood antibody test) and NOT the PCR tests.

Assay can refer also to PCR.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study, has shown that this assay has sensitivity and specificity..."  This means 5 assays were used in blood tests to determine the specificity of THIS assay (in the singular), which is the PCR assay.  I.e., serology is part of the process to find the PCR specificity.

[BeElBeeBub]

That's not how specificity is actually calculated.  They explain their upper and lower bounds for specificity.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study, has shown that this assay has sensitivity and specificity (95% confidence interval) of 99.1% (97.8 to 99.7%) and 99.0% (98.1 to 99.5%) respectively;"

So it's quite possible the specificity was 98.1%, but most likely estimate was 99%.  At 99%, your 500m tests would produce an expected 5m false positives out of the 20m.  However, it's POSSIBLE the specificity was only 98.1%, so the expected false positives would be 9.5m out of 20m, so about half.  We are of course assuming this was the same test and same CT used in your 500m tests, but you didn't specify how you know it was.

It is. From Cochrane 

Specificity is the proportion of people WITHOUT Disease X that have a NEGATIVE blood test. A test that is 100% specific means all healthy individuals are correctly identified as healthy, i.e. there are no false positives. 

So if we assume nobody in the 207,730 had COVID, then there were 207,571 people correctly identified as negative (and 159 incorrectly identified as positive). Which works out as 99.9% as a worst case.

The quote you posted refers to the specificity of the blood tests for antibodies (the terms assay and serology are the giveaway - plus they directly describe the ELISA test as a blood antibody test) and NOT the PCR tests.

[Scott777]

They conducted 207,730 tests.

They got 159 positive results.

If *every single positive* was false (i.e. not one person in the 207,730 actually had covid) you ha e a.macimum false positive rate of 159/207,730.

That is a rate of 0.000765416647

The specificity is 1-0.000765416647

That's 0.999234583
Which is 99.9%
Ant that is the absolute worst case. They looked for evidence that those 159 cases were false, and didn't see any of the statistical "tells".
It's all in the link

That's not how specificity is actually calculated.  They explain their upper and lower bounds for specificity.  "Evaluation of the test sensitivity and specificity of five immunoassays for SARS-CoV-2 serology, including the ELISA assay used in our study, has shown that this assay has sensitivity and specificity (95% confidence interval) of 99.1% (97.8 to 99.7%) and 99.0% (98.1 to 99.5%) respectively;"

So it's quite possible the specificity was 98.1%, but most likely estimate was 99%.  At 99%, your 500m tests would produce an expected 5m false positives out of the 20m.  However, it's POSSIBLE the specificity was only 98.1%, so the expected false positives would be 9.5m out of 20m, so about half.  We are of course assuming this was the same test and same CT used in your 500m tests, but you didn't specify how you know it was.

[BeElBeeBub]

What link?

Reply 126, a few from this one.

But here it is again

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/methodologies/covid19infectionsurveypilotmethodsandfurtherinformation#test-sensitivity-and-specificity