The Department of Health’s daily update showed four more Covid deaths in the UK yesterday, making May 2nd and 3rd the lowest two-day total for eight months. MailOnline has more.
The Department of Health’s daily update showed infections are also down more than a quarter compared to last week after 1,946 positive tests were processed in the past 24 hours.
It comes after just one Covid fatality was recorded yesterday. It is the first time there have been single-digit deaths for two days running since September 14th.
The small number of deaths may be partly explained by the bank holiday, when the figures are often lower to due the way fatalities are logged. The seven-day rolling average number of daily Covid deaths is now 13.
People have also been less likely to come forward for tests on weekends or public holidays and many of the daily swabs are now conducted in schools and workplaces.
Meanwhile, latest figures show another 208,362 second vaccine doses were dished out across the UK on Sunday and 79,304 people were given their first injection.
It means 34.6million Britons — more than half — have been jabbed at least once and 15.6million — nearly a quarter of the population — have been fully vaccinated.
The promising data will be seized upon by MPs, pubs and restaurants, who have called for England’s lockdown to end sooner. One restaurant boss claimed reservations have been cancelled because of the ‘terrible weather’ and hospitality chiefs said it was essential that trading goes back to being “unrestricted” on June 21.
Sir Robert Syms, Tory MP for Poole in Dorset, yesterday said: “We need to push the Government to get on with it. A lot of normal life could be returned.” He said the country would “lose another summer” if rules aren’t eased soon.
The PM has so far refused to budge in the face of calls for more freedom, with restrictions set to stay in place until June 21st — touted as England’s independence day.
Worth reading in full.
To join in with the discussion please make a donation to The Daily Sceptic.
Profanity and abuse will be removed and may lead to a permanent ban.
As the false positive rate for the PCR test is unknown the headline should read
‘For the 407th day in a row the dictatorship produces bogus covid death figures’
It is almost certain that the probability of getting a false positive if you are negative is less than 0.2% for a PCR test i.e. there is a 99.8% chance it will get it right. This follows from simple mathematics. There were about a million tests yesterday and less than 2000 were positive. So even if everyone who was tested was negative then less than 0.2 % of them had a positive result. Of course the 99.8% is a minimum – it is very likely much better.
Some of the tests may not have been PCR tests in which case the figure is for that mix of tests. But as it is widely acknowledged that PCR tests are better than the alternative this would have only made the numbers for PCR tests better.
MFT, PCR tests cannot detect ‘live virus’ and cannot determine whether people are infectious or not.
Your demonstration of your ability to count is pointless, your point is pointless.
“not able to distinguish whether infectious virus is present” quote from
https://www.google.com/url?sa=t&source=web&rct=j&url=https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/926410/Understanding_Cycle_Threshold__Ct__in_SARS-CoV-2_RT-PCR_.pdf&ved=2ahUKEwjb2fGQz7DwAhXvQUEAHR0AAkoQFjAAegQIBBAC&usg=AOvVaw2Pe4mtjh6-hJr7UrjcjRLH
Even Public Health England knows that the tests are pointless, and yet the whole lockdown agenda is driven by these ‘gold standard diagnostic tests’.
Stop getting tested.
Stop wearing masks.
Don’t accept the experimental drugs.
Ridicule the compliant.
Ignore our corrupt ‘leaders’.
When the labs are set up in a hurry with poor processes and massive ramping up of capacity, a single positive will contaminate many others. In those circumstances it may be as high as 4% false positives. The BBC panorama episode showed what they are up to in lighthouse labs (the only episode of panorama worth watching). Now there is no virus, there is no contamination.
You are confusing the rate of false positives with the probability that if you are negative you will get a positive result. The rate of false positives is highly dependent on the proportion of positives in the population and therefore varies very quickly as virus numbers go up and down. The best test in the world will give a high percentage of false positives if the entire population is negative! The probability that if you are negative you will get a positive result is more or less constant provided the same test and processes are used.
Possibly poor processes were used at some stage – but they certainly aren’t now.
Why don’t you watch the hidden camera footage on that panorama investigation and then comment on “possibly poor processes”, and then tell me how you are certain that they aren’t now.
And no, I am not confusing different measures. I fully understand how prevalence affects the chance that your positive result is bullshit. However, the risk of contamination in piss-poorly managed labs seems likely to me to have more of an impact on false-positives than would be expected, had the labs been competent. So the PCR specificity may be variable according to prevalence, which is not good news for meaningful analysis (but then the ludicrous CT levels were already succeeding in obfuscation).
To all those replying above. I am sorry but this follows from simple mathematics. If you disagree show where the maths is wrong.
Have you read the document? Can you read?
Yes I read it. It is not relevant to this discussion. All it is saying that a PCR test does not directly measure whether infectious virus is present. It looks for genetic material that is unique to the virus.
This can rarely give a false positive because e.g. the patient may have had the virus in the past and some genetic material remains. No test is perfect. The question is how often does this happen. My maths gives an upper limit – about 0.2% of the time.
Why the personal insult?
Ah, you can read, you’re just stupid then.
A PCR test, conducted properly and without any cross-contamination, will very accurately detect viral material, but as a diagnostic test, it’s useless unless being used to confirm a diagnosis based on symptoms.
Merely detecting a whole virus or viral fragments without providing the number of cycles needed to provide a positive result, and without symptoms is meaningless.
PCR testing is routinely used for surveillance and diagnostic testing for flu (https://www.who.int/influenza/gisrs_laboratory/final_who_pcr__meeting_report_aug_2011_en.pdf). Is this also an error or is there some key difference?
That’s not a scientific paper replete with references to other research papers, it’s a report of a working group.
It means nothing, robust evidence is the foundation of of good scientific debate, not quasi-religous dogma or political rhetoric.
It almost seems that you spent 10 minutes on Google, that’s not research.
The simple mathenatics of 100% – 0.2% = 99.8% is correct, but meaningless on two counts. Firstly, the 0,2% percentage is very low and depends on keeping within valid cT values and in a very clean lab environment. Neither of these are the case in the UK’s mass PCR testing. Secondly, even if 0,2% were a valid figure, in a population prevalence of 0.2%, there is a 50% chance of a positive being true. That is the meaningful method of calculation.
No – the 0.2% is not a guess or an estimate. It is measured from mass testing that is being done at the moment. It doesn’t matter what the environment is or how a PCR test works. This is the measured result.
“ in a population prevalence of 0.2%, there is a 50% chance of a positive being true”
I am sorry but that is just wrong. The chances of a positive result being true is known as the positive predictive value and depends on a lot more than the prevalence. The formula is:
(sensitivity * prevalence)/((sensitivity * prevalence) + (1 -specificity)*(1-prevalence)
In case you don’t know, specificity is the technical name for probability that the result is correct ifs the person being tested is negative (the 99.8% in my example). Sensitivity we haven’t met yet. It is the probability the result is correct if the person being tested is positive.
If you don’t believe me google PPV.
I am very familiar with thus analysis. Here is my more accurate downloadable Excel version of the BMJ calculator https://drive.google.com/file/d/1f7X5sJvGM4fAfnDBuWZVb8Tu_aY7_Y40/view?usp=sharing
You are saying, if I understand you correctly, that the specificity is 99,8%. If the true prevalence is 0.2%, then around 1 in 2 of positive tests is false (exactly true if 100% sensitivity). The measured prevalence will be 0.4% = sum of true and false.
The 0,2% that you refer to is simply 100% – 99.8% and is NOT derived from mass testing.
I am sorry I didn’t realise how you came to the conclusion. I don’t disagree with you. It seems to me that at the low level of virus we have at the moment we are running at about 50% false positives. My point was simply that the specificity of the test, whatever the failings in processes, cannot be lower than about 99.8% which is pretty good.
But MTF, however somebody does that maths, whether or not the sums are right or wrong, the fact remains that the PCR tests DO NOT DETECT INFECTIONS!
The fundamental point is that all arguments and measures based on PCR tests ARE POINTLESS!
Winston
PCR tests for strings of DNA that are unique to the virus. This is a proxy for the infection which seems to work out well practice. Similar techniques are used for testing for a wide range of infections including Flu and AIDS. Are you going to dismiss all of these tests as pointless?
“seems to work out well practice”
By what measure?
“which seems to work out well practice.”
Evidence?
PCR works well in concert with a well taken history of symptoms and contact tracing.
PCR testing doesn’t detect LIVE INFECTIOUS VIRUS.
Yes, mass testing of asymptomatic members of the community at large.
Completely pointless.
99.8% is not very likely in practice for mass testing. Have a look here https://forums.lockdownsceptics.org/viewtopic.php?p=6863#p6863 where I compared the probable effective specificities from various locations. Around 95% seems to be the specificity from Lighthouse Labs in practice. This multiplies the positive results by a factor of 8!
Things really went pear-shaped at the beginning of September when they ramped up the test rate 3x and got really sloppy.
Anyway, false positives aside, even true positives are meaningless. This is for a variety of reasons. The test picks up viral fragments, not live infections. The period during which these remain after an infection has passed is several times longer that the infectious period. Paradoxically, so-called ‘cases’ are mostly healthy recoveries. Then also, the viral fragments picked up are not necessarily sars-cov-2. The Drosten PCR test formulation was based on a simulation and not an isolate.
Finally, WHO specify four requirements that have to be met for PCR testing – and the UK fails all four! They are not to be used for surveys. They are to be used in conjunction with symptoms. The cT value must be recorded. Positive test results must be repeated.
However you look at these PCR tests, they are a complete disaster.
Typing words into Google isn’t research
Now I know you’re a teenager living with your mum
Bless
Of course, the positivity rate of any batch of tests serves as an upper limit on the operational false positive rate, for that batch. However, it does not constrain the operational FPR for any other batch. To take it as a universal operational FRP upper limit, we’d need to have established the repeatability of the testing process. (Perhaps ISO 2000 would be sufficient.)
Thanks for this comment. It gives me an opportunity to set out my argument in a hopefully clearer and calmer way. I apologise for underestimating the level of knowledge of medical statistics in the community. Knowing that you all understand the jargon allows me to be much more concise.
All I am claiming is that published testing data shows that the specificity of the PCR tests is at least 99.8%. I make no case for the sensitivity but this is widely accepted to be extremely good even by the sceptical community.
The core argument is extremely simple.
* For any test as the prevalence tends to zero, the percentage of positives in the test result tends to (100-specificity). This follows from the definition of specificity.
* I think we all agree that prevalence in the UK is extremely low at the moment.
* Therefore the current percent of positives gives a good guide to the (100-specificity) and therefore the specificity.
Possible Objections
1) The prevalence is not yet zero. Quite likely, that just means that the specificity is actually lower than indicated by percentage of positive tests.
2) The tests use poor processes etc. I doubt this, but even if it is true, it doesn’t affect the argument. The percent of positive test results is the results of the mixture of tests warts and all used on the day. It is what they were, for all their faults, able to achieve in practice. I believe they are a mixture of PCR and Lateral Flow – but the PCR (whatever we think of it) is widely accepted as having better specificity and sensitivity than Lateral Flow. So again, this can only mean the specificity of the PCR tests is better than the positive test proportion suggests.
3) That is only the result for one day. On other days the specificity may have been very different.
The percentage of positive tests during times of low prevalence is remarkably consistent. I attach a chart showing how the ration has changed since 1st April. The same is true of the period June to August last year but I don’t know how to attach two images.
Yawn….
The fundamental point is that all arguments and measures based on PCR tests ARE POINTLESS!
You entirely miss the point! It is the percentage of fale positives among only the positive PCR test results that is significant – not the percentage of the total tested population.
This is why populations of low prevalence should not be mass tested. That is, the true positives get lost in the noise.
Mike
To calculate the percentage of false positives among positive results (100-PPV) you need to know the specificity, sensitivity and prevalence – agreed? I am simply placing limits on the specificity in practice for the mix of tests and labs that the UK currently uses. I don’t see how that misses the point.
As far as the limits on specificity are concerned. Please point out where the logic of my argument is wrong. Do you deny the key point that for any test as the prevalence tends to zero, the percentage of positives in the test result tends to (100-specificity)?
PS. I am impressed by the work you have put into the subject on the forums you link to and wish I had time to follow it up.
PPS I absolutely accept that as we get close to zero cases the PPV will fall and of course if/when we get to zero cases it will 0% and all positives will be false.
On my spreadsheet, there is a ‘helper’ panel that allows you to enter number of tests and number of measured positives. That produces the measured prevalence. With the ONS prevalence estimate in the pre-test probability, an ‘effective’ real-life specificity is found. The sensitivity I take as 80%, but is not a significant driver for false positives. This is all the data that we have to work with at the moment.
What muddies the figures at the moment is that there are far more LFT tests than PCR tests (as of today there are 1,069,727 total, of which only 139,163 are PCR), so we cannot assess PCR tests from this data.
Going back to, say, 21st September (just because I conveniently have the data on my spreadsheet!) ONS prevalence was 0.21%, tests/day 228,139, ‘cases’ 4,964. This gives an effective specificity of 98.0% and a false discovery rate of 92.3%, or 1 in 13 positives were true. This is the true measure of the PCR test in practice at that time.
Mike
That is a good point about the proportion of LTF tests which means my calculation is about the specificity of LTF tests rather than PCR. What was the proportion of LTF back last summer? We also had almost no virus and a very low proportion of positive results.
I am not surprised that the FDR was so high in Sep of last year. With a prevalence of only 0.21% I would expect any test to give a very high FDR. Virtually all the population is negative so of course the majority of positive results will be from negative people. It doesn’t reflect on the quality of the PCR test. Set specificity and sensitivity to 99.5% and you still get an FDR of 70.5%. That is just what happens as you get very low prevalences.
It was only PCR tests mentioned on Pillars 1 and 2 at the time. LFTs were later trialled in Liverpool from what I recall.
The ONS estimate,of prevalence today is less than in September: 0.12%. If one were to address the cases/tests as if they were all PCR tests (I am not proposing this!), about half the positives would be false.
There is a maximum specificity as a function of the test method itself. You will have seen from my forum link that Cambridge Labs approached a commendably high value. However, Lighthouse Labs were sloppy and liable to cross-contamination and excessive cT counts. The net result amounts to a ‘real world’ specificity that is much less. One can simply not use specificity from ideal test lab conditions.
Anyway, as I and others have said here, even true positives are not a measure of infections – and even less of ‘cases’.
Thanks. I was saved the effort of looking up test numbers and positive results for last summer as it was documented here: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/methodologies/covid19infectionsurveypilotmethodsandfurtherinformation#test-sensitivity-and-specificity.
To quote: “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 99.92%.”
I did not, of course, estimate specificity from ideal lab conditions – quite the reverse, I estimated the minimum specificity from the real world results.
Whether true positives measure infections is of course the subject of debate. We seem to have no problem taking true positives for flu PCR tests as a measure of infection. Should we abandon those as well?
In the summer, before they ramped up the rate of tests, NHS would have achieved no less than 99.5% specificity because the measured prevalence was down to 0.5%. But the ONS estimate for actual prevalence was down to about 0.1%, so there were still a high percentage of false positives. What they know about specificity would have applied to these NHS tests, not the Lighthouse Labs. But, paradoxically, the prevalence at that time meant that mass testing was totally inappropriate anyway – even with the high specificity.
Regarding the general use of PCR tests, I mentioned the four WHO requirements earlier. If there are symptoms, one PCR test plus another to confirm using appropriate cT values, then I see no problem using a suitably formulated PCR test for flu. But we would not lock down the whole. largely unaffected, population based on it!
I think this debate has run its course but thank you for sticking with it. I have learned a lot. And I apologise again for grossly underestimating your knowledge of the subject.
Patronising wanker
No normal with face knickers.
End of.
Yes – definitely the wrong place!
Or nappies.
Nappies cover the arseholes of babies and the faces of arseholes.
Brilliantly said, Annie.
(And other bedwetters)
Another bullseye from Annie
People wear knickers on their face?!
Some have.
Didn’t McAfee walk into a German airport with his wife’s panties on from memory (and promptly got taken aside by the German police)
Meanwhile there have been 1047 deaths where the vaccine was implicated, according to the Yellow Card adverse reporting scheme, and 147 days since the first vaccination was done in the UK on 8th December. An average of 7 deaths per day.
So there were probably around seven deaths today where the vaccine was implicated. But only one where the virus was implicated.
You’d think that when deaths from the vaccine overtook deaths from the virus, there might be a bit of media coverage. But no.
Those figures don’t include the individuals who caught Covid shortly after receiving the vaccine. There were mortality spikes in community and care homes settings which coincide with the vaccine rollout.
would those not have gone down as covid deaths seeing as that overrides everything else for 28 days
And that trend is replicated all over the world. Israel, UK, Ireland, Cambodia, Gibraltar, Bhutan, Chile, India.
Bhutan and Cambodia are especially striking recent cases of this trend.
This is being acknowledged with the pathetic explanation that these people are relaxing their mask use after the vaccine.
Masks are the gift that keeps giving. Everything can be hidden behind it.
(the killing fields of Cambodia?)
Maybe they haven’t seen those graphs of covid “cases” showing where mask wearing was introduced.
So the Covid-19 Testing Center in Farnborough that was formerly (until March 2021) the town’s Leisure Center has now been disbanded. It is now an empty building with a glorious 33m unused swimming pool which the community needs but cannot access. Rushmoor Council took the Covid bribery money from the government instead of re-opening. At least they have stopped the testing, as at 1 May.
And yet still they claim to care about our health. It is now a 15 minute drive to the nearest pool and thousands of people are denied local health promoting Leisure Center services. Nice on Rushmoor Council.
The BBC had said that the Nightingale Hospitals would be closed in April. If they have been, surely we are no longer expected to prioritise protection of the NHS above actually leading our lives?
That never ends. Lockdowns for flu incoming. They’ve already talked about keeping certain measures to deal with the massive waiting lists caused by Covid only NHS.
Yes, you can now get served at restaurants if you sit outside. Enjoy the hailstones!
“People have also been less likely to come forward for tests on weekends or public holidays and many of the daily swabs are now conducted in schools and workplaces.”
What sort of person goes and gets tested? I still can’t get my head around that. Doesn’t make any sense at all 8unless you’re really keen on self-isolation and need a reason to do it. Rhetorical.