False positive tests

422 Replies

28th September 2020

(This post contains an erratum regarding a technical issue, at the end)

There has been a lot of noise about false positive COVID19 tests in the news. So, I thought I would try to explain what it all means. Or do my best anyway.

There are two measures in most medical screening tests which are usually defined as the ‘sensitivity’ and the ‘specificity’ of a test. In my opinion, these two words are far too close together in sound, so they are very easy to mix up in your brain.

I find it easier to think of the accuracy of test results in this way.

  • False negatives
  • False positives

A false negative is a result which informs someone that they do not have a disease, when in fact they do.

A false positive is a result which informs someone they do have a disease, when they don’t.

Ideally a test should never give a false negative (100% sensitivity) nor give a false positive (100% specificity). There is no known test that does this. In general, there is a trade-off going on between these two measures.

By which I mean, if you aim for 100% sensitivity, the specificity often falls away – and vice-versa

For example, in cancer screening the primary objective is you must never miss a case. So, the sensitivity rate is set very high. By definition the rate of false negatives is very low.

A shadow on the breast, a few strange cells here, a few strange cells there – ‘that might be cancer, better to be safe than sorry. Don’t take the risk’. Positive cancer test.

To put this another way. The fear of missing any cases of cancer results in the number of false positives being high. This raises the question with COVID19. Is it better to underdiagnose – many false negatives. Or over diagnose – many false positives?

Note I am talking here primarily about the naso-pharyngeal swab tests (i.e., antigen tests) which are used to see if you have the virus NOW and not the blood (antibody) test done which may be done later to see if you have ever had the virus.

This issue does not seem to have been discussed. If you want to prevent spread of COVID19, you would presumably want very few false negatives in these swab tests. Otherwise people will be told they don’t have the disease – when they do – and happily go off spreading it around. Equally, you would be relaxed about false positives. People would isolate when they don’t need to, but not a great health issue.

Weirdly, however, this does not seem to be the case.

COVID19 false negatives

With COVID19, there are a lot of false negatives, with some studies quoting figures as high as 50%. That is, half of those told they are not infected with COVID19, are probably infected1. A systematic review got figures between 2% and 29%. So, we could call that an average of 16%?

As you can see, these figures are clearly all over the place. This is in major part because there is no ‘gold-standard’ COVID19 test. By which I mean that we do not have a ‘test of tests.’ Namely, the expensive and time-consuming test by which we absolutely can know if someone truly is infected. The test against which your ‘field tests’ can be calibrated/verified.

Indeed, currently, there is no current agreement as to what ‘infected’ means with COVID19. Does it mean finding viral particles in the nose, sputum, or throat – or all three? Does it mean finding viral particles in these places, and also isolating it in the bloodstream, or lungs? Does it mean finding evidence of antibodies specific to COVID19 two to three weeks following ‘infection?’ Or what? It would be nice to know.

COVID19 false positives

More troubling, right now, than the very poor sensitivity of COVID19 testing (high number of false negatives) is the knotty question of how many false positive tests there are? This is important, because we are told that cases are rising and rising as we suffer a ‘second wave’ of COVID19.

However, if we have a high rate of false positives, then the rise in ‘cases’ could be driven by a rise in testing – and nothing else. And you don’t need a high percentage of false positive tests to do this. If the false positive rate is as little as just one per cent (1%) this means the majority of people told they are positive for COVID19, do not have COVID19!

I know that most people find this a difficult one. It goes like this.

First, you have to know the estimated prevalence of the disease in the community. That is, the total number currently infected. Last time I looked it was one in nine hundred. For the sake of this calculation I shall call it one in a thousand. [Or, to put it another way, sixty-seven thousand people in the UK (population 67 million) are currently infected with COVID19].

Using this one in a thousand figure. This means, if you randomly tested ten thousand people, you would expect to find ten COVID19 cases [forgetting the false negatives for now].

On the other side of the coin. If the false positive rate is one per cent, you would have an additional one hundred false positives cases.

10,000 x.01(1%) = 100

Putting this another way. With a prevalence of one in a thousand, and a false positive rate of one per-cent you would have ten true COVID19 positive cases, and ninety false positives. Ergo, the vast majority of people told that they have COVID19, do not. Is this actually happening?

There is heated debate. As in much heat and little light.

In order to shed a little light, I have been communicating with a senior scientist in a COVID19 facility who feels things have gone very wrong. Below is his take on the false positive situation, from a couple of weeks ago. It is highly technical, but for those who can follow it, I think the author makes some critical points. I have not named him for, were I to do so, he would almost certainly land in very hot water. However, the references are verifiable.

1: https://www.bmj.com/content/bmj/369/bmj.m1808.full.pdf
What do positive SARS-CoV-2 RT-PCR tests mean? (Absolutely Nothing!)
The Cepheid Xpert Xpress SARS-CoV-2 RT-PCR test is the “Gold Standard” COVID-19 antigen test used in our laboratory. The specificity of this test from the manufacturer’s package insert1. [Here referred to as negative percentage agreement or NPA) is 95.6% or 0.956 when expressed as a fraction].

I don’t know about other RT-PCR tests, but I imagine the specificity will be similar for all widely used commercially available kits.

The specificity of a test is defined by the equation:

SP = TN / (TN + FP)

Where SP = specificity, TN = number of true negatives, FP = number of false positives.  TN + FP = the total number of tests carried out.

Now the latest Government figures from Monday 7th September state that 350,100 tests were carried out and 2,948 people tested positive 2. So, if we apply the above equation to our PCR test and the Government’s figures, we get:

0.956 = TN / 350,100

Therefore, the number of true negatives is:

TN = 350,100 * 0.956 = 334696

Therefore, the number of false positives, FP we would expect from 350,100 tests is:

FP = 350,100 – 334,696 = 15,404

This is more than five times the number of positive tests reported, which means we cannot have any confidence that any one of those positive tests represents a genuine case.

What these figures show is that it is totally inappropriate to use RT-PCR as a screening test for a virus in an asymptomatic population when the prevalence of the infection is very low.

Even if there were a test with 99% specificity, you would still expect to get 3500 false positives from performing 350,000 tests – which is still greater than the number of “cases” reported. When the number of “cases” is lower than your rate of false positives, then a positive result on its own is virtually meaningless.

The PCR test is best utilized as a diagnostic test to confirm the diagnosis of an infection based on clinical signs and symptoms. It certainly should not be used as a screening test when there is low prevalence of disease and should NEVER be used as the sole determinant in the diagnosis of a case.

One source of false positives is the persistence of fragments of viral RNA long after a patient may have recovered and is no longer infective. These fragments will be amplified by PCR and will give a positive result that is indistinguishable from a genuine case.  We’ve had a patient whose swabs have been testing positive in our lab every week for over 3 months!

Non-specific amplification is another possible source of false positives. The nasopharyngeal swab samples are “dirty” samples: they are full of bacterial, fungal, other viral, and host DNA and RNA. Some of these will have high percentage sequence homology [NB homology basically means a similar sequence of base pairs- my words] to the gene sequences targeted by the PCR assay and these can also be amplified. The risk that this may have occurred is higher if the positive test has a very high Cycle Threshold (Ct) value – say 35 or above.

Recently, it has come to my attention that one of the primers – an 18-base primer for a region of the RdRP gene – has exact sequence homology with a region on human chromosome 8 3,4.

So, if any laboratory uses a PCR assay with that particular primer, they’re likely to get a lot of false positives!

Politicians and Health Officials are basing their numbers of cases entirely on the results of these tests, which are not fit for this purpose.

They are then using these figures to terrorise the population, and to justify decisions to impose local lockdowns, and increase nonsensical general restrictions which are having a massive impact on people’s lives and their health, and also on the economy, particularly hitting small businesses hard.

  1. https://www.cepheid.com/Package%20Insert%20Files/Xpress-SARS-CoV-2/Xpert%20Xpress%20SARS-CoV-2%20EUA%20PI%20GX%20System%20rev%20D.pdf
  2. https://coronavirus.data.gov.uk/
  3. https://pieceofmindful.com/2020/04/06/bombshell-who-coronavirus-pcr-test-primer-sequence-is-found-in-all-human-dna/
  4. https://www.ncbi.nlm.nih.gov/nucleotide/NC_000008.11?report=genbank&log%24=nuclalign&from=63648346&to=63648363

ERRATUM

In this blog I included a piece on false positives from a senior laboratory scientist. A number of people wrote in suggesting that the calculation was wrong. I contacted the scientist on this matter, and he has written:

In performing my calculation, I was unable to calculate the number of true positives (TP) because I did not have a figure for the prevalence of COVID-19. Since the prevalence seemed to be close to zero from the results obtained in the laboratory where I work, I assumed that TP would be negligible compared to the total number of tests carried out, and therefore did not include this in the equation I used. I acknowledge that the number of false positives (FP) calculated was thus an approximation.

I have since learned that the prevalence is approximately 0.1% according to the ONS, which means that my value for FP is actually a very good approximation, and this validates my argument that the number of false positives far outnumbers the number of true positives.

I hope that clarifies matters

422 thoughts on “False positive tests”

  1. andrea

    I would have thought that the fact you are testing people WITH symptoms makes all the difference.
    If you are testing people at random (like at the airport) is one thing.
    If you are testing a self selected group it is quite another.
    How do you take into account this difference?

    Reply ↓
    1. Cheezilla

      They are not testing people with symptoms!

      Most people rocking up for tests are asymptomatic and have been sent from work, school, need a test to fly somewhere, or is the panicking parent of a snotty-nosed kid.

      Anyone voluntarily turning up for a test has been self-selected for idiocy, not covid.

      Besides, those with symptoms should be isolating at home and not out spreading their cold or flu germs around the rest of the population.

      Reply ↓
        1. dcfl51

          The tests carried out by or on behalf of the ONS are of volunteers who are intended to represent a random sample of the population. The idea is to establish how prevalent the disease is in the population as a whole. Most of the volunteers will be free of the virus, reflecting the situation in the population. I would imagine that the ONS, being statistical experts, will adjust the figures according to the demographics of the sample v the general population.

          I believe that the samples are taken using home testing kits and I would guess that they are then posted to a lab for the PCR test. It does make me wonder, if the PCR test produces so many false positives, doesn’t this mean that the baseline figure for prevalence produced by the ONS could be way off the mark ?

          Reply ↓
        2. Dan Anderson

          ONS is I’m guessing an agency that manages statistical information and involved in testing orders. I was surprised at the £450 cost of testing, which seems high. Also, others are asking about the reliability of testing of the dead and how false negatives or positives might make our data unreliable. I’d like to see this issue, those numbers, examined factually.

          Reply ↓
          1. dcfl51

            Yes, this is the Office for National Statistics which is a government agency independent of political influence. The £450 is a payment made to the volunteers to take part in the survey which runs continuously with repeated testing of the volunteers.

    2. Ian Roselman

      Yes, that is exactly the point. “The PCR test is best utilized as a diagnostic test to confirm the diagnosis of an infection based on clinical signs and symptoms.” As I understand it, in England only people with symptoms can book a test (if they can get a test slot). The prevalence of infection in the tested group is higher (can anyone say how much higher?) than that in the general population. It follows that the ratio of false positives to true positives is lower, maybe much lower, than for the population at large.

      Reply ↓
        1. Dr Tim

          Professor Philip Nolan is chair of the National Public Health Emergency Team (NPHET) Irish Epidemiological Modelling Advisory Group. He stated that “If you have symptoms, or are a close contact, the prior probability of having SARS-CoV-2 infection is much higher than the population prevalence. A close contact has about a 1 in 10 chance of being infected.”

          Reply ↓
          1. KJE

            But is that infected with no symptoms (ie just a positive test) or infected with symptoms (ie actually ill)? If we keep on like this, we’ll never get herd immunity.

          2. Jerome Savage

            In a lot of employment situationss it’s an opportunity to get away for a break. A cough, a sneeze, sore throat is never ever questioned, off you go, get tested and see you when we see you.
            Naive if you think otherwise

        2. Aileen

          Also, I believe anyone having an invasive procedure on the NHS has to self-isolate and then have a Covid swab three days before the procedure is due. At least that is what happens in our Trust. No need for any symptoms.

          Reply ↓
        3. ricksanchez769

          “We’ve had a patient whose swabs have been testing positive in our lab every week for over 3 months!” Heavens! do we have a living day Typhoid Mary? Alas I think not – all this terrorizing nonsense of asymptomatic spreaders, asymptomatic ‘super’ spreaders is anecdotal or based more on an n=2 wives’ tale. Run the PCR test for 20 cycles nil, run it for 25 nil, run it for 35 and poof – there we have it…silly (and this is called science)

          Reply ↓
      2. andrea

        Yes, that would certainly be the case. This is why I don’t understand the point that the author wants to make. Besides, the more real cases you get, the less relevant the false positives are as they become just background noise.

        Reply ↓
        3. Harry de Boer

          My feeling is that ‘the more cases you get’, the more cases you have *had*. I.e. the cumulative number of ‘previous cases’ is always increasing. This will lead to a growing part of the population walking around with covid particles from a destroyed virus.
          So the later you test a random sample in the population, the higher the supposed prevalence of ‘cases’ is that you will detect while in fact the increase comes from the recovered true cases.

          In the Netherlands however, the whole reasoning of this cited expert won’t hold as there people are told to (apply for a) test when they feel they are symptomatic only. So there actually the test is used to confirm already suspected cases, which is a better use of the test. That’s why there currently only 16.8 tests are performed per confirmed case[1], which is much less than the reciprocal false positive rate, which means that the contamination of the numbers with false positives is rather low.

          Last week 152 cases have been admitted to the hospitals[2] (last table: “Aantal meldingen ziekenhuisopname 152”), which in my view isn’t sensationally much, yet friends are already panicking, whipped up by the media of course, but could have been prevented if people were prescribed one of the highly successful ionophore/zinc++/macrolide cocktails, which is outright shameful, also in view of the possible long-term adverse effects of an infection.

          [1] https://ourworldindata.org/coronavirus-data-explorer?zoomToSelection=true&time=2020-03-01..latest&country=~NLD&region=World&testsPerCaseMetric=true&interval=smoothed&perCapita=true&smoothing=7&pickerMetric=total_deaths&pickerSort=desc

          [2] https://www.rivm.nl/nieuws/aantal-meldingen-met-ruim-60-gestegen-ook-meer-ziekenhuisopnames

          Reply ↓
          1. andrea

            True, but here it is happening the same. You are supposed to book a test only if you have symptoms and not on a whim. When school went back in Scotland there was a huge demand for tests, but that produced a VERY low set of positives, which probably represent the false positive rate. It was pretty much negligible.
            The increase in positives in the last few weeks cannot be explained by false positives and cannot be just “legacy” cases as the increase it too fast.
            One thing for sure, with this test we will NEVER be rid of Covid, but that’s a problema for another day.

          2. Jane in France

            I’m trying to reply to Andrea. When kids go back to school or university after the summer holidays they always catch cold or freshers’ flu. This year if they get tested some SARS Cov 2 is bound to be floating around at the back of some noses. In that sense these kids aren’t “false positives.” On the other hand it doesn’t mean they’re dangerously ill or are likely to make anyone else dangerously ill.

      4. David McCluskey

        Just for the record (although it doesn’t significantly alter the conclusions), there are errors in the calculation. TN + FP is not equal to the number of tests carried out but is equal to the total number of negative results. So the arithmetic that follows should use 350,100 – 2,948 = 347152 as the multiplicand, making TN = 0.956 x 347152 = 331,877. Thus FP = 347152 – 331877 = 15,275 not 15,404 as shown.

        Reply ↓
        1. Justine Kent

          But TN + FP is not equal to the total number of negative results either. Surely TN + FP is the total number of people tested who don’t actually have the virus? So Dr Kendrick is incorrect to use specificity as the true negatives as a proportion of all tests carried out. But it is also not correct to use it as the true negatives as a proportion of all negative results. (From what I can understand from the formulas, this latter would instead be “predictive negative value”). Specificity is the number of negatives picked up in tests as a proportion of all those tested who are actually negative, And we just don’t know this last figure. So I’m confused about how specificity is being used to calculate the number of true negatives (and therefore the number of false positives).

          Reply ↓
    3. KJE

      But what symptoms? Most so-called CV19 symptoms could just be the same as a cold or mild flu-like illness, that’s the problem

      Reply ↓
      2. Dr Tim

        Professor Nolan states that “If you have symptoms, or are a close contact, the prior probability of having SARS-CoV-2 infection is much higher than the population prevalence. A close contact has about a 1 in 10 chance of being infected. 2/10”

        I think that most close contacts are going to be asymptomatic.

        Reply ↓
        1. Ben Palmer

          “We mostly test people with symptoms and close contacts of known cases” This may be the case. But in Germany and France people randomly queue up at test centers in the city or in their car at gas stations for voluntary tests. They for certain have no symptoms.

          Reply ↓
      3. Harry de Boer

        Not really. If you have a runny nose that’s definitely not covid. Now I learned that viruses are mutually exclusive for some reason, so that if you have a runny nose you most probably won’t have covid-19.
        Covid-19 has the loss of taste as a unique symptom.

        Reply ↓
        1. KJE

          Years ago, I used to lose sense of smell when hayfever was really bad – don’t seem to these days, so even that isn’t unique – although combined with persistent coughing and high temperature, it might be.

          Reply ↓
        2. Terry Sanders

          Viruses aren’t mutually exclusive. You can be infected with many different virus at the same.
          For example, more than 60% of under 50 year old’s are infected with Herpes simplex virus 1 (causes cold sores). This doesn’t bestow immunity from HSV-2 (genital herpes), HIV, COVID or any other virus.
          If virus infections were mutually exclusive then the solution would be simple, infect everybody with a benign virus.

          Reply ↓
          1. KJE

            However, different coronaviruses might be “exclusive” or at least confer some degree of immunity to others.

        3. Ali

          I always lose my sense of smell & taste for a few days when I have a cold. But a cold is just another form of Coronavirus apparently, so maybe that is just a symptom of having a Coronavirus generally…..

          Reply ↓
      4. sam

        the whole point is to keep the fear going until the vaccine is ready and everyone is forced to get it and get ‘health passports’ (covipass and ID2020) as per Tobias Ellwood
        The virus has gone, no-one is ill!

        Reply ↓
    5. Frango Asado

      Even if they were testing only people “with” symptoms, the question arises: what symptoms?

      Cough, sneeze, runny nose, headache, feeling off colour – all are also symptoms of flu (or just a cold).

      We have the unique situation of a disease with virtually no distinctive symptoms and no reliable test. A ghost in the machine! A disease so virulent that you don’t know you’ve got it until your test shows up positive.

      Reply ↓
      1. JohnC

        Apparently there is one unique symptom and that is loss of the sense of taste. This is the problem with what is happening in schools and universities. In schools children are getting a cough for example and they’re told to get tested. They have to stay at home until the result is known. Other children in their class and even their siblings continue to attend school. If the result is positive then their “bubble” is sent home.
        However, children get coughs and colds at this time of year every year. Likewise, students get infections at this time of year. With the false positives or positives from viral remnants the number of cases rises. If someone has measles for example, we don’t test for the measles virus as a random test, it’s based on signs and symptoms ( the rash is probably the only unique feature of measles). Likewise herpes zoster, it’s the blister like itchy rash starting on the trunk that diagnoses chicken pox.
        I cannot think of any viral infection that we test for rather than symptom based other than SARS-COV-2.

        Reply ↓
        1. JohnC

          Dr. K. May find this amusing, I got this from Pulse (a news feed primarily for GPs), after it was reported that schools were asking for proof that a pupil had visited their GP. One GP said that this happened to him, but stopped after a phone call to the school asking for his standard fee of £50 for the letter.

          Reply ↓
        2. AlanG

          I’ve lost my sense of taste (and smell) many times while suffering from the common cold and (especially) from hay fever. It is absolutely not a symptom unique to “Covid-19”.

          Reply ↓
    6. IMoz

      Hang on, I thought NPA and PPA were stated for applications of tests as indicated (i. e. when presentation justifies testing (symptomatic)), not if the the test were applied to a stochastically modelled experiment?

      Reply ↓
      2. Binra (@onemindinmany)

        The PCR primer is supposed to be a unique identification of something proven to be a cause of disease. I understand that different primers can be selected and used as well as other differences in testing parameters.

        If you can convince people that ‘cholesterol levels’ are disease markers you can make a mint as well as ‘medicalising’ or controlling the health under fear of threat. Regulars here understand this example although it is nothing to do with ‘infections’. However the principle remains that if you can set the narrative, define the ‘disease’ and control the testing and diagnosis parameters, you can literally play god over the lives and deaths of others. BUT – they have to willingly believe you by reacting WITHIN the narrative rather than first establishing is basis or otherwise in fact.

        The novel narrative is of latent infections requiring life-denying state controls into perpetuity.
        Clinical diseases that in themselves are almost harmless and doubtless serve a true function in the body, are being used as a basis for triggering and manipulating deep fear contagion that is being directed into fear of Life – and a willing compliance in mass suicide.

        The PCR test is NOT a worthy or reliable diagnostic tool – even IF primed to a correctly identified identity, definitively proven to be be the cause of clinical disease.

        Reply ↓
  3. Peter Swettenham

    So what I don’t understand is why Chris Witty and Patrick Vallance don’t understand this stuff (they are supposed to be scientists!) and continue to act as though all positive results indicate true COVID-19 cases and then draw silly graphs based on some fag-paper assumption that cases will double every week!

    Reply ↓
    1. johnandclare1970

      They know exactly what they’re doing! They’re going to drag this out, using false positives, until they roll out their vaccine, then they’ll claim victory! And never mind the hundreds of thousands they killed with their lockdown!

      Reply ↓
    3. johnplatinumgoss

      Chris Whitty is a virologist who wants to believe vaccines are the best treatment for any virus on the planet. I’ve read some of his papers. One he even co-wrote with Neil Ferguson and others a piece about tough decisions in reducing Ebola transmission in West Africa. But more recently (2019) he put his name to a paper on pumping even more money into vaccine manufacturers’ research. This is a useful tool for finding out who is getting the cash from Bill Gates Foundation, Wellcome Trust and Cepi.

      https://www.sciencedirect.com/science/article/pii/S0264410X19311971?via%3Dihub

      Reply ↓
  4. John

    Perhaps you could comment on what then should be considered ‘a case’. I came across this during my time down a rabbit hole and wonder if you think it it makes a fair point?

    “China has a very stringent definition of what constitutes a Covid case – requiring presence of clinical signs and symptoms AND in most cases, epidemiological exposure, AND a positive PCR test (at 35 amplification cycles or less compared to 45 for UK)) before counting a Covid case according to this source

    https://doi.org/10.1101/2020.04.26.20080911

    By contrast, World Health Organization, the US Centers for Disease Control and Prevention, the European Centre for Disease Prevention and Control, Public Health England, the Public Health Agency of Canada, and the Pan American Health Organization require ONLY a single positive PCR test in order to count a Covid case – clinical symptoms and epidemiological exposure are not required. This would mean that America (and UK) is significantly over-counting its cases (and, potentially, its deaths) compared to China.

    It is odd that WHO has apparently not noticed this anomaly and acted to ensure uniform international standards and definitions so that statistics are internationally comparable. It is also rather odd that both CDC and WHO had guidelines in place to minimise over counting of cases for previous viral outbreaks including SARS, MERS, Ebola and Zika, yet these safeguards have been removed in respect of SARS-CoV2 “

    Reply ↓
    1. IMoz

      I don’t think even China’s definition goes far enough: a proper infectious disease diagnosis requires a culture of the infectious agent, otherwise you risk mis-diagnosis. Think about it: someone presents with flu-like symptoms, you do you swab, get back SARS-CoV-2 positive, what if your patient actually has a metapneumovirus infection and SARS-CoV-2 was just debris?

      Reply ↓
    2. Dan Anderson

      Giving a link is helpful, but a link with some preface of content and conclusion would be more helpful. Just a sentence or two describing the link is what I’m suggesting.

      Reply ↓
    3. Dan Anderson

      April 17 – Wuhan changes criteria for diagnosis of Covid, raises death toll by 1290 that day. I recall it was based on x-ray/scans of lungs without testing. What has happened since then, I don’t know.

      Reply ↓
  6. Mr Chris

    Very clear.
    Would we be better going back to using actual cases, that is people with symptoms and hospital admissions?
    There is a lot of parroting of others countries having done better than the U.K. through testing and then isolating. Have they done this on the basis of reliable testing, if such a thing exists?
    I have the impression that the current policy is lockdown plus. Is the current resurgence real or imaginary based on unreliable testing. What is the anecdotal impression from doctors who contribute on here?

    Reply ↓
  7. Yvonne van Eck-Remmers

    Thank you Malcolm for this interesting information. In The Netherlands, the number of people ´infected´ is rising steadily (caused by increasing amounts of people tested), which gives the government a chance to impose all kinds of harsh measures. What I really do not understand is this: experts have stated that PCR-test are made for diagnostic purposes only, and should not be used the way they are used now. Is there really no other test that could be used? And what could one do to rule out the many false positives? A lot of people cannot go to work, visit friends and family while all they have is a simple cold, or not even that.

    Reply ↓
    1. Jerome Savage

      Yvonne – “experts have stated that PCR-test are made for diagnostic purposes only” , that point comes up again and again but is never dealt with that I can see

      Reply ↓
      1. Joe Dopelle

        Actually, PCR is completely unsuitable for diagnostic tests.

        As clearly stated by its inventor, Kary Mullis, who won the Nobel Prize for his invention of PCR. He explains that, because it can be used to amplify the tiniest trace of DNA billions of times, it will show up any DNA that exists in a given sample.

        Reply ↓
        1. Jerome Savage

          Thanks Joe,
          Assuming that is Kary Mullis speaking, that is very valuable footage as the “fact checkers” have been busy with numerous pages claiming that he never said that, Reuters being one of them. That certainly applies to google but that odd named search engine “duckduckgo” is relatively free of such claims. Its remarkable what is happening.

          Reply ↓
          1. Patrick Healy

            Jerome,
            Dear oh dear, those lying censoring fake news outfits g##gle and Rioters, there is nothing odd about Duckduck or Yippy if you want unbiased information.
            I admit their unusual names may be unfamiliar to many but well worth investigating.

  8. Jean Humphreys

    I want to weep. Is there anyone else out there with more than two brain cells to rub together?
    It is about time them in power were told to learn what the science really is, and stop playing their lovely board game with real people as pieces.

    Reply ↓
    1. Phil JW

      Some fantastic comments on here and I have learnt so much but this is one of the most meaningful ones, hence 39 thumbs up. I also feel like weeping, especially when when I look at my 7 year old son and wonder what awful future is being designed for him. I also feel like weeping as I have it, on good authority, that very high up medics, Trust Medical Directors etc and others even high up in NHS England also are getting really frustrated that a positive narrative and correct modelling is being actively suppressed. Actively suppressed.

      Reply ↓
  9. JDPatten

    Very cogently described!
    Are the sensitivity & specificity of tests used elsewhere in the world known? Australia, Sweden, Iceland? China??

    Are there ongoing efforts to develop that definitive “test test”?

    So, you don’t think it worthwhile to quarantine the false positives along with the positive positives?
    Who is it that’s not worth that? Two weeks.
    Oh, and ain’t the two weeks kind of arbitrary? Sort of like two meters or six feet. Nice and even.

    It’s been a bit more than six months. The virus certainly seems to have outpaced any sensible response on our part. General Human Nature? General corruption of medical science? Of politicians? All the above!? In what proportions? Has it always been like this and we just didn’t fathom it?
    OK. Good science takes a while. Meantime . . . what??

    Trust is gone.
    Perhaps it was always misplaced.

    Reply ↓
    1. Craig E

      @JDPatten I’m trying to find out for Australia what the cut-off cycle threshold is for RT PCR negative/positive via my local Pathology. Will ask them as well which PCR tests they use and what the stated sensitivity/specificity of each is. I heard the C(t) cut-off Australia uses is 30 but am trying to confirm.

      The study below is intriguing. They took 90 positive PCR samples and tried to culture the virus. For positives above 24 cycles none were able to be cultured implying that beyond 24 cycles only dead RNA/virus fragments detected. Only 26 of 90 exhibited culture growth

      https://pubmed.ncbi.nlm.nih.gov/32442256/

      Reply ↓
    2. Jerome Savage

      Sarah – facts, figures & a hypothesis based not on emotions or notions but reality – nothing to be concerned about. Where, might I ask, does the holocaust come in ?

      Reply ↓
  12. Trish C

    Malcolm – the basis of the calculations for determining how many of the positives are false positives seems to rest on an estimation of prevalence. From where does this estimation of prevalence come from? I.e how do they know this with any certainty?

    Reply ↓
    2. Mike C

      Well they could take the number of positive ‘cases’ they’ve found and divide that by the number of tests done. That would give a value of prevalence among the tested population. By my reckoning that would be 2.16%. If they’re sensible they could divide that by an arbitrary ‘2’ as Pillar 1 tests are done on patients and high-exposure people. If in doubt, just go with 2.16% it must be lower than that.

      Specificity is 95.6% so false positives are 4.4%… double the worst case prevalence.

      Reply ↓
      1. Dr Tim

        Are we simply using the positivity rate on a given day of testing as the prevalence? What about 14 day incidence? What about national cumulative incidence? Dr Kendrick, can you help me understand the current cases of 67,000 in the UK? Thanks.

        Reply ↓
        1. Mike C

          I based my 2.16% on cumulative count of all cases divided by all tests – numbers from the UK gov dashboard (ie since March). Crude, I know – but I can’t see how it could be any higher in the general population than in the tested population.

          Dr Tim mentions 14 day incidence. Is 14 day incidence a particular thing in assessing prevalence of a bug in a population? Or has it been calculated from some figures I haven’t found (or looked for hard enough)?

          Certainly I would expect infections to precede detected cases to precede consequent deaths – but by how many days on average? Looking at the peaks in UK data I came up with 23 days from infection to death – for those unfortunates who died in late March/early April in the UK.

          Reply ↓
          3. DaveL

            John Ionnadis just took another stab* at estimating overall IFR (infection fatality rate), which I liberally interpreted to mean that the IFR is about 9.7 times the total mortality rate (curve fit off his data, the very last graph). Total cumulative mortality rate for UK on 21 Oct is 0.070%; 9.7 times that is 0.68% IFR. CFR (Case fatality rate) is 5.60% on 21 Oct 2020. The ratio of CFR/IFR would be 8.2 in the UK case, based on that. That is the same thing as the ratio of actual infections to confirmed cases, of course; the fatalities cancel out. Prevalence of daily confirmed new cases on 21 October ’20 is 0.042%; therefore estimated actual daily prevalence would be 0.35%. Duration of infection not taken into account, since who knows what it is. Dr Kendrick says 0.1% actual prevalence, which given the accuracy of this calculation, doesn’t seem to conflict with it. Same order of magnitude, which is more than one could hope for. Of course, both could be wrong. The CFR/IFR ratio varied a standard deviation from 6.2 to 12.3, around the 8.2, so it’s pretty much a shotgun approach.

            *Publication: Bulletin of the World Health Organization; Type: Research Article ID: BLT.20.265892

        2. Her outdoors

          Let me help you. They’re making it up as they go along. At least that’s my conclusion which I reckon based on my fag packet analysis is just as good as theirs! 😂

          Reply ↓
        1. Mike C

          Many thanks for this and the other links Dr Tim. Yes, the maths looks right. It’s perhaps amusing that my throwaway suggestion of dividing the UK numbers by two to make some allowance for ‘Pillar 1’ (NHS/Public Health testing) compared with ‘Pillar 2’ (broadly speaking tests carried out on a self-referral basis) tests works nicely with the 3.4% compared with 1.7% figures you’ve provided for Ireland. I’ve no doubt that the 1.7% figure has more science behind it than my crude figure.

          Looking for a 1.7% prevalence bug and using a mass screening test with a false positive rate of 4.4% can’t go well.

          Reply ↓
    4. Martin Back

      In South Africa we have been in lockdown for 186 days. There are still people getting infected so we haven’t reached herd immunity yet. To be optimistic, let’s say 50% of the population had been infected. That works out to 50 / 186 = approx 0.25% new infections every day.

      According to this reference, the virus stays inside us for plus-minus 26 days. So at any one point in time we should have 0.25 x 26 = 6.5% infected. Sound reasonable?
      View at Medium.com

      Reply ↓
  13. Nigella P

    Thank you Dr K. My disappointment in the lack competence of our current Government and its so called scientific advisors has no limit, which is just as well as they remain constantly and relentlessly disappointing.

    I am so horrified that it seems to be ok to imprison students in their halls of residence on the basis of these deeply flawed tests. How has it come to this?

    Reply ↓
    1. Dr. Malcolm Kendrick Post author

      It has become a bit like the Salem witch trials. Once accused, you were doomed. It seems that all the weird and dysfunctional human psychology that we claim has been banished to earlier, less civilised times, lies very close to the surface. Once you get people sufficiently frightened, you can do anything you like – pretty much.

      Reply ↓
      1. Astrogeezer

        “Once you get people sufficiently frightened, you can do anything you like-pretty much”.Well said. Until you realise how easy it is for your mind to be manipulated, you remain the puppet of someone else’s end game.

        Reply ↓
      2. AhNotepad

        If you are the right group, you can do anything you like, criminality not withstanding. Some of the organisations that can use criminal acts proposed in the latest security bill include HMRC, the health services, the police, the armed services, THE JUSTICE DEPARTMENT. There are others.

        Reply ↓
        1. Her outdoors

          What? Having seen other of your posts I know you are well researched. That is beyond scary. Can you share a link?
          On the road home the other night there was some graffiti ‘The world is flat’. Always laughed at that but now it wouldn’t surprise me if it is. I already know the aliens are here and running the show. ☹️

          Reply ↓
      3. Tish

        Well let us just be thankful that they haven’t started plonking black crosses on people’s doors. It’s probably pencilled in somewhere.

        Reply ↓
    2. The Dim Appear

      Nigella, I was reading an article in the Guardian the other day written by a head teacher of a primary school where she told of all the problems she is encountering because of COVID.

      Out of the whole school, the students, the teachers and the parents there were only two people that I had the slightest interest in hearing about. I think both were the same age, seven or so, one boy and one girl. Neither were related and the two incidents happened independently, but each child was caught trying to escape by climbing over the high fences that surrounded the whole school, much like all schools are now surrounded, trapped behind them to keep them safe. Neither was successful in escaping sadly. Brought a tear to my eye.

      That little girl is someone’s daughter, someone’s grandchild, someone’s sister, someone’s neice and someone’s friend. That little boy is someone’s son, someone’s grandchild, someone’s brother, someone’s nephew and someone’s friend.

      I’m sorry for being all emotional, but the mind altering device hanging on the wall keeps showing images of all these people being emotional about COVID, it’s highly infectious and I think I’ve caught it, so I can’t think straight any more, I’m welling up again and need to find some tissues.

      Did you know in the last year that 1.1M to 1.2M people in the US alone have died of heart disease or cancer. I think you’ll agree that’s a mind numbing figure but the mind altering device never mentions it.

      Back to your point about students being held in their halls of residence to keep them safe, I think you’ll find they’ll cope just fine, they went to primary school after all.

      Reply ↓
      1. The Dim Appear

        Just to clarify… when I said ‘it’s highly infectious and I think I’ve caught it’ I’m referring to the highly contagious ’emotional state’ regarding COVID rather than the virus COVID itself. I feel as fit as a butcher’s dog and have no symptoms of COVID whatsoever.

        Reply ↓
  14. elizabethhart

    In Australia 7,563,322 tests have been conducted, with 0.4% classified as positive.*

    Someone on another blog today suggested tests cost $200. I haven’t verified this, and haven’t found an Australian cost on the internet so far.

    But if we go with it, this would be: 7,563,322 x $200 = $1,512,664,400 😳

    I’m no expert, but it seems to me all this physically intrusive testing is a waste of time, resources and money, which would be better spent elsewhere.

    * https://www.health.gov.au/news/health-alerts/novel-coronavirus-2019-ncov-health-alert/coronavirus-covid-19-current-situation-and-case-numbers

    Reply ↓
    1. KJE

      Might be better spent on good treatment for those that are ill (with anything), and perhaps vit D tests, but it’ll be Summer down there soon so perhaps not needed

      Reply ↓
    2. Craig E

      @elizabethhart I wonder if that’s just the cost of the test and doesn’t take into account the cost of having people at Covid-19 testing stations, the cost of the lab equipment and lab staff time etc etc. Maybe it’s a lot higher than $200 per test. Eye-watering cost either way!

      Reply ↓
    3. Trish C

      I have heard that here in NZ doctors get paid $250 per test, so that’s maybe where your $200 comes from. I would imagine that the cost to the government (and therefore to taxpayers) is way more than that.

      Reply ↓
    4. Harry de Boer

      Realize that a treatment with the cure (hydroxychloroblahblah+) is 10x cheaper, and needs only spending at the onset of symptoms. So if those tests were right, only $6,000,000 would have been spent on those 7,563,322 subjects.

      Reply ↓
  15. The Dim Appear

    I might have to read through that a couple of times for it to sink in fully, but thank you for it.

    In other news I wonder if this MPs “rebellion” against Boris where MPs want to get to vote on further restrictions on life is just a move by Boris?

    Given that your average MP is probably as ill informed as much of the population are, if (surely I mean when) they are successful in getting to vote on future restrictions won’t they just fall in to line given that Boris has a majority anyway? I think all this does is provide him with a cover to do whatever he wants to next and have it appear more democratic.

    I suppose it will at least give MPs a chance to discuss things but I don’t feel any of them know what’s going on as I haven’t heard any of them talk about the things discussed on this blog.

    It concerns me greatly if I’m honest.

    Reply ↓
    1. Her outdoors

      Boris looks utterly terrified, look at his eyes. What of? And has a deathly palor. And Rishi Sunak has developed the same look recently. Again, what of?

      Reply ↓
      1. Zephyrbear

        They’re terrified of the runaway train that is Brexit hurtling towards them and unstoppable… COVID-19 is the perfect distraction until it hits, at which stage both will disappear leaving someone else to clear up the mess!

        Reply ↓
      2. Nigella P

        Total economic collapse I should think. Between Brexit & Covid I suspect the UK will become a husk of a country, barely first world. I do not feel optimistic about the future.

        Reply ↓
  16. johnplatinumgoss

    The PCR test was never designed to check whether a person has a specific virus. It was used to check for AIDS with a similar lack of success. My understanding is that anything over 30 amplifications is unlikely to give any indication of the disease. In the US they are using between 40 and 60 amplifications. And of course they have more people allegedly having COVID-19.

    From the start I have based my own fear of catching this virus on how many people I know who have been diagnosed with it and what harm it is likely to do to someone otherwise in reasonable health. The answer is still 1 person. I met her recently when out walking my dogs. She said it was bad at the time. Since then she has come round to my way of thinking, was happy to give me a hug, and asking when the next rally was against all these lockdown measures.

    I also see what is happening in countries which did not take the bribe from the IMF and did not lockdown: Belarus and Sweden. I only recently discovered Nicaragua.

    Nicaragua – further proof that lockdowns don’t work

    Reply ↓
    1. Patrick Healy

      John
      Do you know anyone, who knows anyone, who knew anyone who died FROM this Wuhan Flu?
      I certainly don’t and neither does anyone I know know.

      Reply ↓
      1. Dodi

        Yes. I work for a charity and one of our volunteers died on April 8. She was in her 80s but still driving and doing voluntary work. Her son (in his 60s) was hospitalised on the same day as his mother and was ventilated for weeks. He is now recovered as far as I know. No one else in the family was ill and at that time they wouldn’t have had any form of testing. She was a highly sociable and active woman and was in close contact with many people in the few days before she was hospitalised- including me. No one else had symptoms.

        Reply ↓
  17. Keith Hulme

    I understand the arguments for a large number of false positives, but with so many tests giving a negative result, how many false negatives could we expect. Sounds like this type of testing is a waste of time – and worse!

    Reply ↓
  18. sjglaze3

    Brilliant article. I’m sharing with anyone who is open-minded and scientifically literate enough to understand the implications of your article. Sadly, it’s a very small group of people.

    Reply ↓
  19. Astrogeezer

    Is it not the case that anti-body testing (using blood) rather than unreliable antigen swab testing should be the norm? It seems, from general discussions with acquaintances, that most of them would not be at all surprised to discover that they’ve already been infected, possibly as early as November/December 2019. Perhaps I’m missing something here. Why not have a blood test to settle it once and for all? Have we ‘had the virus’, have we ‘got the virus’, or do we have to wait for a ‘cure-all’ vaccine to protect us?

    Reply ↓
    1. Steve-R

      Except that unless you also test for IgA and T Cells you will still miss huge numbers of infections.
      Most antibody test are just for IgG and IgM which only appear after Covid-19 was severe enough to make you know you had had it!

      Reply ↓
    2. KJE

      I think that not everyone gets anti-bodies (or not the ones they test for) since many have immunity through T-cells (can be tested but isn’t AFAIK)

      Reply ↓
  20. Darko Mulej

    Just a small correction: when defining specifity

    SP = TN / (TN + FP)
    Where SP = specificity, TN = number of true negatives, FP = number of false positives. TN + FP = the total number of tests carried out..

    It’s not true, that TN + FP = the total number of tests carried out, but TN + FP = the number of real negative cases in the data.
    See wikipedia for nice drawing: https://en.wikipedia.org/wiki/Sensitivity_and_specificity

    Reply ↓
    2. Covideo Nasty

      Thank you.

      I’d read the piece, this evening, and thought my brain was failing. Very glad to find reassurance from your comment.

      Why hasn’t Dr Kendrick fixed the article? I suspect bewilderment, caused by the error, will put a lot of people off reading to the end, and, probably more importantly, deter them from sharing the piece.

      Reply ↓
  21. Simon Tilley

    Dear Malcolm
    As a follower of your emails I regularly pass them onto friends and family.
    In the past week I sent my brother your articles on Sweden and Belgian doctors re masks. He posted these on Facebook.
    I think, that Big Brother intervened, as both articles were pulled by their controllers as infactual.
    I thought you should know.
    Best wishes
    Simon Tilley

    Reply ↓
    1. Bev

      I shared the Belgian doctor’s letter on Facebook, too. While it wasn’t pulled, it got a warning of non-factual material attached to it. Strangely, I’ve just been back to my timeline to look at the share and the warning is no longer there. ??

      Reply ↓
  22. Peter FordO

    Based on this excellent post even the numbers of hospitalisations and deaths will probably be unreliable, and especially now we have the new app all numbers can only go up. It’s like a self-fulfilling prophecy and there appears to be no way out. All the more so as we may have a bad flu season because people have been lowering their immunity by social distancing etc and flu will be mistaken for covid.

    Reply ↓
    1. KJE

      Don’t get the app. Never fill in anything for track and trace – avoid anywhere that makes you do so. It’s the only way to save your livelihood. Why are they giving 500 quid to people on benefits to self-isolate but nothing to those on low incomes who don’t get benefits? How many people can afford two weeks on a max of 95 quid statutory sick pay or a minimum of nothing?

      Reply ↓
  24. jeanirvin

    I would share this but I just keep meeting people who say but ‘it’s still dangerous’ and ‘there are loads more people ill’. I give up!

    Reply ↓
    1. AhNotepad

      It happens to most people. I was talking to someone today who said they knew someone who died of covid, so they’re following all the rules. I asked if they were going to give up driving a car when they next heard of someone dying in a collision. I bet they don’t follow all the rules when driving.

      Reply ↓
    2. Jennifer.

      Jean. I understand why you feel there are more people out there expressing a fear of apparent escalating numbers, when compared with the prominent brains who aim to convince us otherwise. After all, we are being bombarded with facts and figures from all directions. Surely, those of us unable to decider higher mathematical equations, will err on the side of caution?
      Even our esteemed PM can’t work things out today, regarding the North East of England regulations. Not that that supports my argument, ha, ha.

      Reply ↓
  26. Marion Eaton

    On Mon, 28 Sep 2020 at 15:02, Dr. Malcolm Kendrick wrote:

    > > Hear.! Hear! Thank you for your clarity > > > > > Dr. Malcolm Kendrick posted: “28th September 2020 > > > > There has been a lot of noise about false positive COVID19 tests in the > news. So, I thought I would try to explain what it all means. Or do my best > anyway. > > > > There are two measures in most medical screening tests which are usually ” > > > >

    Reply ↓
  27. Jerome Savage

    And this. Although the line “according to one test manufacturer” immediately prompts the comment, “well they would say that wouldn’t they” https://www.health.harvard.edu/blog/which-test-is-best-for-covid-19-2020081020734
    “some experts recommend repeated antigen testing as a reasonable strategy. According to one test manufacturer, the false positive rate of antigen testing is near zero. So, the recent experience of Ohio Governor Mike DeWine, who apparently had a false-positive result from an antigen test, is rare”

    Reply ↓
  28. Kevin McCarthy

    Thank you for this superb explanation Malcolm.
    The only thing I am not getting is the increasing % of test results that are positive.
    Also question about increasing hospital admissions. based on the acknowledged statistic that 92% of all deaths recorded as COVID deaths have at least one underlying condition, would it be fair to assume that 92% of “Covid” hospital admissions also have underlying health conditions and might there be a possibility that they are being admitted to hospital for the underlying health condition and when tested on admission are recorded as COVID admissions if they test positive?? @Malcolm – any of your contacts able to give an opinion on this?

    Reply ↓
  29. LA_Bob

    “Recently, it has come to my attention that one of the primers – an 18-base primer for a region of the RdRP gene – has exact sequence homology with a region on human chromosome 8.”

    Someone, please help me out here. I clicked on ref 3. This takes me to an April 2020 blogpost where this issue is discussed. In the comments Bernd Paysan says,

    “People, the virus is an RNA virus, and chromosome 8 is DNA. It’s not going to be replicated by the RT-PCR, because this process starts with the Reverse Transcriptase (RT!), and therefore does not copy DNA.”

    Is this correct? Are we talking about a problem which is not a problem after all?

    Seems like there are enough diagnostic issues that there is no need to shout, “Squirrel!” But, if I’m wrong, please let me know. Thanks.

    Reply ↓
    1. KJE

      I thought that point of the PCR test was to amplify DNA. Since it can’t do anything to RNA, the RNA has to be changed to DNA first – that’s the reverse transcripterase bit, so it seems that if one of the primers in a test looking for DNA is some commonly existing DNA, it’s going to find it. But I am probably totally wrong. Why don’t all the tests use the same primers?

      Reply ↓
    2. JustSaying

      Reverse transcriptase is the first step, yes. It generates a DNA sequence from the RNA sequence, which is required because the core PCR process only works on DNA, not RNA. The DNA sequence on chromosome 8, if present, will be copied; one of the two primers will match it and the polymerase will then make a complementary copy of the sequence. However, because the sequence on chromosome 8 only matches one of the two primers, the complementary copy won’t match the other primer in the next thermal cycle, so there will be no chain reaction. Put more concisely, the “P” part of the PCR will work for the chromosome 8 sequence but the “CR” part won’t. The growth will be linear rather than exponential; if you start with a single copy of the chromosome 8 sequence, then after 30 thermal cycles you will end up with just 30 (complementary) copies of it, instead of 2^30=1 billion copies for the sequence that matches both primers. So provided you set your detection thresholds accordingly, there should be no confusion between the target sequence and the chromosome 8 sequence.

      Reply ↓
      1. LA_Bob

        JustSaying, thanks for this explanation. Very helpful.

        Would you please post a link about this? Most of us here are lay folks (I know I am). I would like to understand more about the PCR process. I’ve read the Wikipedia article about, but I need more.

        I’m concerned the “senior scientist” Dr Kendrick quoted is not familiar with what you discussed and may be wrong about the false positive incidence with the 18-base primer. The thermal-cycle threshold issue is nasty enough without more confusion over false positive COVID tests.

        Thanks for your help.

        Reply ↓
        1. JustSaying

          Hello, I couldn’t find much online discussion about what happens in PCR with just one primer (the linear amplification as described is inherent in the PCR process) but there’s a passing mention of what happens in this Wikipedia article https://en.wikipedia.org/wiki/Variants_of_PCR under “asymmetric PCR”. As a deliberate technique, it’s usual to use an excess of one primer rather than one primer alone, because with just one primer, as mentioned, you don’t get a chain reaction so it’s not really PCR. There’s some discussion here: https://www.researchgate.net/post/PCR_with_single_primer .

          Something else to note is that when using a single primer there’s nothing to say where the other end of the copied sequence should be apart from the duration of the thermal cycle, so the copied sequences won’t be of uniform length, as they would be with standard PCR using two primers.

          If you were after more general tutorial material on PCR (Wikipedia articles can be rather dense) then Khan Academy is often a good place to try for technical education. Here’s their video on PCR https://www.khanacademy.org/science/ap-biology/gene-expression-and-regulation/biotechnology/v/the-polymerase-chain-reaction-pcr and you will also find an article in the sidebar on that page.

          Reply ↓
          1. LA_Bob

            Thanks. The Khan Academy video and one other article were good places to start. I still need to check out the other links, including the ones from Martin Back.

      2. Martin Back

        They separate the RNA from any DNA, protein etc in the sample before applying the RT-PCR test (see my post elsewhere this page), so if that step is properly done there shouldn’t be any DNA from any source to contaminate the results.

        Reply ↓
  30. Simon

    The test package insert link [1] is no longer working — I believe this is the current equivalent:

    Click to access Xpert%20Xpress%20SARS-CoV-2%20Assay%20ENGLISH%20Package%20Insert%20302-3750-Xpress%20Rev.%20E.pdf

    That said, in looking for that I also found this sibling:

    Click to access Xpert%20Xpress%20SARS-CoV-2%20Assay%20CE-IVD%20ENGLISH%20Package%20Insert%20302-3787%20Rev.%20A.pdf

    Which appears to claim 100% specificity for a similar or related test? Perhaps you could glance and clarify? (I am unsure how to guess/know which tests are actually being used in practice.)

    Reply ↓
    2. Dr Tim

      The PDF you refer to states the following:
      “Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) were determined by comparing the results of the Xpert Xpress SARS-CoV-2 test relative to the expected results. Results of these 90 archived clinical NP swab specimens are shown in Table 3. The PPA was 97.8% (95% CI: 88.4% – 99.6%) and the NPA was 95.6% (95% CI: 85.2% – 98.8%).”
      The NPA is the same as specificity, thus 95.6% is accurate as Dr Kendrick states.
      Go to https://www.medcalc.org/calc/diagnostic_test.php and use a prevalence of 0.1; I am using a prevalence of 0.7 in Ireland. Create a 90% sensitivity (which is on the high end of clinical sensitivity) and a 95.6% specificity scenario with a 0.1% prevalence and you have a 2.01% positive predictive value (PPV). In this same scenario using a 0.7% prevalence you would have a 12.60% PPV. However, I have not been able to find a definite specificity for Ireland, as it is simply not clear what PCR test we using. However, Professor Philip Nolan has hinted is 99.7%. With a 0.7% prevalence and a 90% sensitivity you have a 67.90% PPV.
      BTW, analytical sensitivity and specificity does not equal clinical sensitivity and specificity.
      https://asm.org/Articles/2020/April/False-Negatives-and-Reinfections-the-Challenges-of

      Reply ↓
      1. Simon

        You are quoting the first pdf, which I already said was the same as the one in the OP (just with updated url since the one in the OP no longer works, at least for me).

        The second link says “The overall performance of Xpert Xpress SARS-CoV-2 for all 30 samples combined shows a positive percent agreement (PPA) of 100% (95% CI: 88.7% – 100%) and a negative percent agreement (NPA) of 100% (95% CI: 88.7% – 100%).”

        Which is what I am asking about.

        Reply ↓
    3. JustSaying

      Hello Simon, the results reported in the two versions of the PDF are, as you indicate, from different tests. For the results shown in section 19.1 of the second (older) PDF from May 2020 they took a set of nasopharyngial specimens that were known to be COVID-free and spiked some of them with various concentrations of COVID viruses, in a randomised and blinded way. So these were semi-artificial specimens. The test results in all cases accurately reflected whether or not the specimen had been spiked, which is where those 100% numbers come from.

      The 95.6% NPA in section 22.1 of the other (September 2020, “EUA only”) PDF is from a test in which they used a set of specimens previously tested by an RT-PCR system that had already received an Emergency Use Authorisation. The Cepheid results were then compared with the expected results. They don’t say which other system they were comparing their results with. They do mention (Table 3 footnote b) that the two false positive specimens were collected during the pandemic, leaving open the possibility that there could have been something dodgy about the specimens (welcome to real life). The number of specimens was a bit low if we are after definitive PPA and NPA figures.

      Cepheid also recently published a paper (Loeffelholz et al) on some other tests that they ran on their COVID-19 PCR cartridges https://jcm.asm.org/content/58/8/e00926-20 Table 2 of that paper shows the results of a test against several other PCR systems. Relative to those other systems, overall PPA was 99.5% and NPA was 95.8% which is consistent with the September 2020 PDF.

      All of the test results have quite wide 95% confidence intervals, so they are more like an approximate guide than a precise reflection of reality.

      Reply ↓
  31. Trish C

    Who is making money out of producing all these PCR test kits? It seems the world would be a far better place if we just stopped testing asymptomatic people.

    Reply ↓
  32. Neil Charles

    Good evening Dr Kendrick. I have posted a link to your recent blog. I received a reply stating “he makes a major error in his calculation though. Where he states SP=TN/(TN+FP) he then goes on to incorrectly state that TN+FP = Total Tests.

    Would it be possible to clarify this ?

    Reply ↓
  33. Neil Charles

    Good evening Dr Kendrick

    I posted a link to your recent blog and received a reply stating “he makes a major error in his calculation. Where he states SP=TN/(TN+FP) he then goes on to incorrectly state that TN+FP=Total Tests”

    I note from another reply to your blog that TN+FP=The total number of real negative tests.

    Can you clarify this ?

    Reply ↓
    2. Martin Back

      If you look at the table, there were 45 known positives and 45 known negatives.
      The test identified 44 positives and 43 negatives correctly.
      Therefore the table shows:
      Positive Percent Agreement (PPA) = sensitivity = 44/45 = 97.8%
      Negative Percent Agreement (NPA) = specificity (SP) = 43/45 = 95.6%

      The formula SP = TN / (TN + FP) is wrong.
      It should be SP = (TN – FP) / TN
      Where SP = specificity, TN = number of true negatives, FP = number of false positives.
      TN – FP = the number of negatives correctly identified

      If 43 of the 45 negatives were correctly identified, then 2 were incorrectly identified as positive, i.e. were false positives i.e. TN = 45, FP = 2
      Thus SP = (45 – 2) / 45 = 95.6% as before

      Reply ↓
      1. Ben Palmer

        Your formula seems to be off.
        Specificity = True Negative / (False Positive + True Negative) x 100
        https://www.easycalculation.com/statistics/specificity.php#:~:text=Formula%3A%20Specificity%20%3D%20True%20Negative%20%2F,%28False%20Positive%20%2B%20True%20Negative%29%20x%20100.Your formula seems to be off.
        Specificity = True Negative / (False Positive + True Negative) x 100
        https://www.easycalculation.com/statistics/specificity.php#:~:text=Formula%3A%20Specificity%20%3D%20True%20Negative%20%2F,%28False%20Positive%20%2B%20True%20Negative%29%20x%20100.

        Reply ↓
        1. Martin Back

          The two formulas give you almost the same result. Using the numbers in the table
          Specificity = NPA = TN / (FP + TN) = 45 / (2 + 45) = 0.957 = 95.7% (your formula)
          Specificity = NPA = (TN – FP) / TN = (45 – 2) / 45 = 0.956 = 95.6% (my formula)

          A larger percentage is better, therefore you would expect Xpert Xpress to quote the larger percentage if they were allowed to. But they quoted 95.6%, so I assumed mine was the accepted formula.

          In words, my formula = “What percentage of the actual negatives were correctly identified?”
          Your formula = “What percentage of the total negative tests were actual negatives?”

          Reply ↓
    3. JohnC

      Sensitivity is the number of negatives in a sample that should all test positive (false negatives)
      Specificity is the number of positives in a sample that should all test negative (false positives)
      I had to do my infection control and protection mandatory training the other day, one of the slide was about viral infections and it stated that most viral infections are without symptoms and people don’t realise they are infected.

      Reply ↓
      1. dcfl51

        Are you sure about this. I thought the definitions were :
        Sensitivity : the number of positives in a sample that should all test positive
        Specificity : the number of negatives in a sample that should all test negative
        Hence, in an ideal world, sensitivity and specificity should be as close as possible to 100%.
        False negatives = 100% – sensitivity
        False positives = 100% – specificity

        Reply ↓
  34. Rodrigo

    “They are then using these figures to terrorise the population, and to justify decisions to impose local lockdowns, and increase nonsensical general restrictions which are having a massive impact on people’s lives and their health, and also on the economy, particularly hitting small businesses hard”
    Thanks Malcom, Hopefully there is a growing momentum around this view but it’s a hard slog trying to encourage others to recognise the devastating and even lethal impacts and consequences of lock down.
    One huge resource to those of us making the case against such growing dystopian measures would be far more medics speaking out about their concerns. I think it is the one factor that would encourage members of the public to have the confidence to begin to question the mainstream narrative and start a process of investigation and open debate. Most of the media won’t like it but some may pick up the gauntlet.
    What does it take to get medical professioinals into the reality and ethics of all this?

    Reply ↓
    1. Aileen

      Rodrigo, you’re absolutely right. We do need more medics to speak out. *But* Dr Heiko Schoening from Germany was arrested at Saturday’s London protest, held for 22 hours without charge and had his laptop, phone and a copy of a book called “Coronavirus: False Alarm” seized. He didn’t even get to deliver his address.

      Dr Adil has been suspended for a year (I think) for asking the wrong questions.

      Not many people are in a position to risk losing their livelihood. Those who do speak out are extremely brave. It is no coincidence that that many of those professionals (not necessarily medics) who do speak out are retired: John Lee, Lord Sumption, Michael Yeadon.

      Reply ↓
  35. The Wizard

    Thank you for this post Dr K. The Casedemic being used to justify ever more ridiculous measures is simply preying on the ignorance of the population, most of whom are scientifically illiterate. Seemingly intelligent individuals are just parroting the rubbish pushed by the BBC and the MSM.
    I was recently in a conversation with an acquaintance and tried to explain that when there is a low overall disease prevalence (as is widely acknowledged, ONS puts it at 0.1%) the false positive rate (FPR) of the rt-PCR will result in virtually all positives being false. He wasn’t interested; I am deluded apparently.

    HandyCock has reluctantly stated that the FPR is 1% , but independent experts have expressed that it is much higher. This coupled with the ridiculously high cycle threshold of 45 and dubious primer(s) means the results are about as valuable as half a rotting dead fish in an empty crisp packet !

    How many of these “cases” are actually unwell? Do they have ANY symptoms at all? Why were they tested in the first place? Why are people being paid to take tests? This is all a theatrical performance better suited to the West End or Broadway. Whitless and HalfBalanced pushing 50,000 cases per day EVERY day is more hardcore porn than even a seasoned adult performer could handle.

    I was in attendance at the political rally in Trafalgar Square on Saturday. The organisers had a fully approved risk assessment in place and permission from Westminster Council too. The entire gathering was peaceful, many families with children and we listened to a number of speakers without hindrance for a few hours. Then when a prominent German doctor took to the podium, the riot police charged. He was about to speak medical facts and it is very telling that this is when the police moved in. I was yards from this highly articulate, well dressed and peaceful gentleman who was simply trying to spread the message of truth and hope. Clearly, the authorities were never going to let the truth get out. I truly despair.

    Reply ↓
    1. David Lilley

      What does Matt Hancock mean by the False Positive Rate. It is quite possible that 1% of ALL tests yield a false positive result. But the percentage of POSITIVE tests which are false is much higher, possibly a majority of them. I doubt whether Matt Hancock would understand the difference. It is quite usual for politicians to spout percentages without knowing what they are percentages of.

      Reply ↓
    2. Her outdoors

      I heard today what Vallance said in his graph presentation.
      ‘This is not a prediction but simply a way of looking at things.’
      Really? Well matey I look at things too and I don’t see what you see. I’ve seen better knitting patterns. I only know 1 family that has had Covid, husband first with man flu followed by wife and 4 teenage and older kids, in March. Ill for 3 weeks, poorly for 6 in total. Medics didn’t test but said 99% sure. Despite living isolated lives with all contacts known, they know no-one else who has had it, even now 6 months later. Figure that out?

      Reply ↓
      1. KJE

        Yeah. I had a cold. I live alone, I don’t visit anyone. The only people I see are outdoors walking dogs or masked in shops. Where did the cold come from?

        Reply ↓
    2. kzb

      This is what I was thinking ! However, does not a positive test result class you as a “case” rather than simply “infected”? In which case it is the CFR (not IFR) that is higher due to FPs giving an erroneously high number of cases.

      Reply ↓
  37. Stuart Cairns

    Dear Dr Kendrick

    Apologies for asking a silly question but have you challenged the Westminster and Hollyrood governments about this? Also, are you ok for me to challenge the Scottish Government with this?

    Thanks and best regards

    Stuart Cairns

    Reply ↓
  38. brainunwashed

    Dr Fauci Says Asymptomatic People Have NEVER Driven Outbreaks!

    So why are they testing asymptomatic people
    (and this from Bill Gates “business partner”)

    Reply ↓
  39. Chris Emblen

    You are absolutely right on the statistics but miss the point that the PCR tests 3 genes so the rate is closer to 99.99% assuming no positive unless all 3 are.
    @bealelab on Twitter explains it very well.

    Reply ↓
  41. Trust me I'm Not a Doctor.

    After reading this article I was in a confused state and my brain hurt. I was going to get an IQ test but they’re not reliable either.

    Reply ↓
  43. Maggie Filkin

    Bull’s eye, Malcolm. Allowing the population of otherwise healthy people to become infected with Covid 19 will rid us of the disease. It seems to have happened in Sweden as Sebastian has reported from his hospital in Stockholm. If our experts don’t wise up, we shall live with a roller coaster situation of whenever we take the brakes off, infection rates will rise and we will need to reapply the brakes again. We are just prolonging the situation.
    We must shield the vulnerable, but let’s get on with it.
    It’s counter intuitive but it is the correct course of action.

    Reply ↓
    1. AhNotepad

      Maggie, Allowing people to become infected, I think will happen regardless of interference by politicians pretending they are doing anything for people’s good. Unfortunately some, including the UK health minister (what an oxymoron) have used the phrase “let it rip”. This conveys an uncaring attitude, even if it is very similar to “allowing people to become infected.

      Reply ↓
  44. robertL

    Quoting: Politicians and Health Officials are basing their numbers of cases entirely on the results of these tests, which are not fit for this purpose.
    They are then using these figures to terrorise the population, and to justify decisions to impose local lockdowns, and increase nonsensical general restrictions which are having a massive impact on people’s lives and their health, and also on the economy, particularly hitting small businesses hard. end quote.

    Boris was elected to deliver Brexit. After that he is rapidly dispensable or disposable.
    Why? because he does not have a clue about everything concerning this virus. Because he is very very badly advised. The BUCK STOPS with the PM!

    .

    Reply ↓
  45. dearieme

    “One source of false positives is the persistence of fragments of viral RNA long after a patient may have recovered and is no longer infective.” I suspect that’s wrong, doc. The test by its nature finds fragments of viral RNA. A test-of-tests would establish how accurately it does so but it wouldn’t distinguish infective virus from noninfective viral fragments.

    The fact that such fragments may be noninfective is surely just another defect in the testing procedures, additional to the problem of false positives.

    In other words I suggest that:

    bogus positive results = false positives plus irrelevant positives.

    Note the “plus”.

    I suspected that the original demand for near-universal test-and-trace was, if you’ll pardon the expression, utter bollocks, and I still suspect it.

    Reply ↓
    1. Joe Dopelle

      Perhaps a simpler way of putting it is that any discussions about false positives rather miss the point.

      The real point is that the PCR is not, never was, and was not designed to diagnose specific diseases. It is a way of multiplying DNA, that’s all.

      Scientists started using it to try to diagnose because they have nothing better. It’s as if they were trying to dig a hole in a garden, but had not spades or shovels. So they have resorted to using knives and forks.

      Reply ↓
  47. elizabethhart

    Speaking of testing, what do you think about this?

    The World Health Organisation wants to roll out 120 million rapid-diagnostic coronavirus tests to help lower-income countries make up ground in a testing gap with richer countries — even if it is not fully funded yet.

    At $US5 ($7) apiece, the program initially requires $US600 million in funding and will start as early as next month.

    The rapid tests look for antigens, or proteins, found on the surface of the virus. They are generally considered less accurate — though much faster — than higher-grade genetic tests, known as PCR tests, which are used in many wealthier nation.

    My emphasis.

    Rapid coronavirus test which diagnoses COVID-19 in minutes to be rolled out by World Health Organisation – ABC News
    https://www.abc.net.au/news/2020-09-29/who-unveils-rapid-diagnostic-testing-in-lower-income-countries/12712462

    Reply ↓
  48. annedempsey1968

    Thanks so much for your elucidation of the data on PCR testing for covid 19 testing, explaining the significance of false positive and false negative results. It really helps to resolve the data to down to earth interpretation for us non statistical utterly savvy people. I studied statistics many times, in preparation for undertaking my thesis for my bachelor of science in physiotherapy, my advanced masters (as they term it in America), my clinical doctorate in physiotherapy in physiotherapy, the later two in Boston, the first in Dublin, Ireland. I was good at maths and physics and chemistry in school and college, and while I had a fairly reasonable grasp on statistics and interpretation of scientific research findings during these endeavours, I regrettably have to admit I have become very very rusty, probably mostly due to the fact that I am really a clinician and not really born to be a researcher. I totally love reading research papers if they appear to offer anything of apparent clinical relevance, I ashamedly have to admit that I do what one should never do, I read the abstract, results and conclusions, and attempt not tm be too easily influenced, as have been made aware frequently, that anyone can prove their bias with conveniently using manipulative statistics, that those who are not rusty readers can decipher with better awareness due to better understanding of the stats and research reporting. You do a great job of giving a clearer message to people who can understand the information when expressed in more more understandable normal language, for people who don’t regularly converse in statistical scientific language, but most people are able to understand the message when explained in common language. Akin to there isn’t a stupid student, the one trying to inform is very blessed if can find the appropriate way to communicate the message in the way the listener can understand. Thanks

    Reply ↓
  49. David Bailey

    Thanks Malcolm for this very revealing piece – and indeed thanks for the anonymous scientist that wrote most of it.

    “Recently, it has come to my attention that one of the primers – an 18-base primer for a region of the RdRP gene – has exact sequence homology with a region on human chromosome 8”

    Does “exact sequence homology” mean “the same sequence”?

    If I read that right, if a certain part of human chromasome 8 is in use at the time of the test – i.e. it is being transcribed into RNA – then one particular variant of COVID test will detect it as if it were COVID-19. That is utterly absurd – how can that possibly be used as a test?
    Furthermore, I would have assumed that every primer would be checked against the entire human genome to ensure there was no match.

    Reply ↓
  50. Dr Tim

    I thought I would try redoing the math. Here it goes:

    SP = TN / (TN + FP)

    Where SP = specificity, TN = number of true negatives, FP = number of false positives. TN + FP = the total number of people not infected

    Now the latest Government figures from Monday 7th September state that 350,100 tests were carried out and 2,948 people tested positive 2. So, if we apply the above equation to our PCR test and the Government’s figures, we get:

    TN + FP (# of people not infected) = 350,100 – 2,948 = 347,152

    SP = TN / (TN + FP)
    0.956 = TN / 347,152

    Therefore, the number of true negatives is:

    TN = 347,152 * 0.956 = 331,877

    Therefore, the number of false positives, FP we would expect from 350,100 tests is:

    # of people not infected = 347,152

    347,152 = TN + FP
    FP = 347,152 – TN
    FP = 347,152 – 331,877 = 15,275

    The math was out only by a bit (original article quoted 15,404).

    As the article states above, this is still more than five times the number of positive tests reported, which means we cannot have any confidence that any one of those positive tests represents a genuine case.

    This hinges upon the 95.6% PCR specificity in the UK. In Ireland, we may be using a 99.7% (not confirmed as of yet).

    Reply ↓
    2. Martin Back

      Or to put it another way, you could get the same number of positives if the test was 99.16% accurate.
      350,100 X (1 – 0.9916) = 2948 approx
      Conclusion: The Xpert Xpress test is not good enough at only 97.8% accuracy.

      Patrick Charles, a Melbourne doctor, tweets that their tests are more accurate:
      “Specificity for the assays we are using (performed at MDU and VIDRL) appears to be in the range of 99.9%. Several times (when case numbers were low) we have done many thousands of tests without a single positive.”

      Reply ↓
        1. Bart Merwede

          offcourse the test is 100% fals positive, this is because the genetica sequence SARS-CoV-2 they compare with is nog from a virus. How can you otherwise explain that already > 114,000 sequences of “the coronavirus” have been uploaded to gisaid.org? are we supposed to believe that each sequence is frim a pathogenic virus. Think about this: When they do the pcr test they find a different sequence EACH AND EVERY TIME. How should we believe than that there is “the coronavirus” when these pseudo-scientists have already found more than one-hundred-thousand sequences and are finding something new with each test?

          Reply ↓
  51. Craig

    Thank you for interesting discussion as always.

    As others have mentioned, the fact that at least some percentage of the ‘ testees’ are symptomatic ie self selected positives, will change those simple numbers. More importantly, if someone tests positive then has another test some suitable time later and gets another positive result, the specificity will be somewhere between that 0.956, and 0.998 for the two tests combined.

    This is very informative from a back-of-the-envelope point of view but I’m not so sure that when you are dealing with large populations with rates of effects over time intervals, on the various subsets of that population, that it comes down to simple algebra. In fact I’m sure it isn’t.

    Friends in the US tell me that the test rate is declining, it is hard to get a test, but the incidence rate is rising?

    Reply ↓
  52. David Bailey

    ” More importantly, if someone tests positive then has another test some suitable time later and gets another positive result, the specificity will be somewhere between that 0.956, and 0.998 for the two tests combined”

    Not necessarily, because if some of the sequences match human sequences( as discussed in the original article) the two tests may well both show a false positive.

    Reply ↓
    1. KJE

      Or the test might be detecting dead particles from a previous infection. I bet we’d be seeing all sorts of “epidemics” if they PCR-tested at 30-40 cycles for other common viruses

      Reply ↓
    2. Craig

      Granted. But that needs to be taken into consideration differently I think. It will never not return a false positive, no?

      Reply ↓
      1. David Bailey

        If the fragment of RNA that is used as a primer for the test, matches the DNA in a human chromosome, then yes there will be false positives.

        These primers are really short – 20 bases, so if the material under test is contaminated with other bacteria, viruses, or fungi, then these might also contain that sequence and give a false positive.

        Reply ↓
  53. Studs Ramrod

    Thanks again for this. Like many, I am seen as a witch who must be burned if I dare mention any of this stuff.
    BUT apart from the self inflicted harm on society of using dodgy maths and science, there is no possibility of any 2nd wave (i’m not even sure there was a first) other than the regular flu season and patients being classified as Covid “cases”. So dying “with COVID”,not “of COVID”.

    Reply ↓
  54. Eric

    Hmm, the quoted senior scientist must be using a single target test. I posted this about 6 weeks ago that most labs in Germany quote a specificity of > 99% and posted test data from France and Germany that showed a total positive rate of about half a percent during the summer.

    In fact, that rate has been creeping up in Germany in spite of a doubled number of tests (> 1 million / week since mid-August):
    https://www.spiegel.de/wissenschaft/medizin/coronavirus-infizierte-genesene-tote-alle-live-daten-a-242d71d5-554b-47b6-969a-cd920e8821f1

    It’s the fourth graph labelled “Wie viel wird getestet”. The first tab shows total negative and positive results, the second shows the positive rate.

    More importantly, here’s an article (in German, unfortunately) that takes a close look at the claim that false positives are behind the rise:
    https://correctiv.org/faktencheck/hintergrund/2020/09/09/pcr-test-auf-sars-cov-2-warum-in-der-praxis-falsch-positive-ergebnisse-selten-sind

    First, they quote a round robin from Mai:

    Click to access 340%20DE%20SARS-CoV-2%20Genom%20April%202020%2020200502j.pdf

    The round robin used heat inactivated (!!!) samples. For the three regular samples of diluation ratios of up to 1:10,000, the sensitivity of all labs was 99.7%. For the fourth sample with a dilution of 1:1,000,000 (!), it was still 93%.

    For the negative samples, the specificity was 97.8 to 98.6%. This seems comparable if slightly better than what the scientist says.

    HOWEVER, these number are for SINGLE TARGET tests. At least in Germany, most labs do a dual target test as a standard, and if they don’t, according to Dr. Drosten, they follow up positive results with a second test for a different target.

    So one must multiply the false positives. For a dual target test and a worst case false positive rate, this is 0.035*0.035 = 0.00123, so 99.88% specificity.

    One lab group that does about 8% of tests in Germany has been doing triple target tests since March, so they are confindent their specificity is 99.99% or better. Another lab spokesmen confirmed that they do a second test only if the first is positive. Yet another lab said they do a careful review of each positive test and decide case by case. The RKI seems satisfied that the false positive reports are negligible.

    Reply ↓
  55. Eric

    Re: the video we had here about CO2 behind a mask, this article (in German, unfortunately) is really interesting:
    https://correctiv.org/faktencheck/hintergrund/2020/09/25/was-passiert-wirklich-wenn-der-co2-gehalt-unter-einer-stoffmaske-gemessen-wird-ein-experiment

    The journal Correctiv visited the German Federal Office for Environmental Protection and did some experiments with two scientists there. Main takeaway: CO2 sensors are optimized to measure concentrations of below a few 1000 ppm. Just breathing without a mask onto a sensor that sits in free air will completely saturate it for half a minute or more, so the sensor will only measure the peaks of our breathing because it cannot follow the breathing frequency.

    Secondly, they say that if the CO2 couldn’t escape from the mask during exhalation, it would have to inflate like a balloon.

    Reply ↓
  56. comlabs2019

    While I broadly agree with the ‘high case-count scepticism’ articulated here, let me ask you (Malcolm) a relevant question:

    It strikes me that so many times in medical research, people focus on static, snapshot-type numbers instead of observing the the DYNAMICS.

    During the first three weeks of September:

    The growth in daily testing was under 133%
    The daily ‘case’ count rose by over 230%

    If someone were to muse that the daily cases might be wholly or largely false positives, then how would they explain this dynamic disparity?

    Surely there is SOME kind of real world trend here – no?

    Note: I do not assume that a true false positive is an ACTUAL ‘infection’ – that’s a different narrative to be sceptical about!

    Thanks,

    Nick

    Reply ↓
    1. Eric

      I think we can discount false positives as a reason for the high rates.

      Whether the true positives really pose a problem is entirely debatable. Death rates are going up everywhere but are still pretty low compared to this spring (but really reflect infections that happened at the beginning of August.

      From the table from the guardian, deaths in the last two weeks were about 0.5% of new infections in NL, UK, F, D but more like 1% in E and I. And this is not taking into account that deaths are delayed, i.e. reflect infections when the daily rates were maybe half the current rates.

      So we are currently seeing a CFR of maybe 1-2%, which is not so much lower than in spring, considering we are doing a lot more tests now.

      The big question is who is dying? Are the fatalities only people who were within weeks of dying anyway and happened to test positive for CV? Or do we still have folks with comorbities or even young people without dying in droves who still had years to live? How about lang haulers, do we have statistics about this?

      Reply ↓
        2. Eric

          I was trying to account for the two week average that the guardian has applied + the average time of 5 weeks from infection to death. I read that number recently but can’t find it right away. So I was assuming 6 weeks, but that may well be only 4 or 5 weeks. It does not matter really. Looking at Johns Hopkins data for France, daily infections for the second half of September were averaging about 10,000 and taking only 4 weeks, averaging about 3,500 in the second half of August. For the sake of eyeballing the CFR, this is good enough.

          Reply ↓
          2. Eric

            I think I have shown that multiplying the apparent CFRs by 2 to account for the dynamic of case numbers is reasonable.

          3. Dr. Malcolm Kendrick Post author

            Really. How can you possibly know? I think it is far more possible that, at present, people are getting a positive Covid19 test then, coincidentally, becoming ill with something else, then dying – of that something else. Covid19 has nothing to do with it. However, because they had a positive Covid19 test within 28 days, they become part of the Covid19 death count. Equally, someone can be in car crash, enter hospital seriously ill, have a swab done, be found to be positive – then die, as as result of the car crash injuries. They too, will be recorded as a Covid19 death. ‘Death from any cause, within 28 days of a positive Covid19 test’ is the terminology used. Currently, it is far more likely we over over-counting than under-counting.

            What is ‘the dynamic of case numbers’, by the way? I have not come across this phrase before. It does not appear to have any obvious meaning.

          4. Eric

            Are you trying hard not to understand me? 🙂

            Just trying to eyball the minimum time between infection and being counted as a fatality in the two week intervall given by the guardian, recorded cases in France were about 3500/day in the second half of August and the second half of September, so if we get an apparent CFR of 0.5% in September, this is really a CFR of 1% because infection happened in August.

            I know, we are certainly not finding all the infected, and we may be labelling people as infected due to a positive test who are not really ill. The point I am trying to make is that the current CFR is not orders of magnitude lower than it was in spring. I personally find this unsettling, especially as it looked like the fatalities would stay down as recently as August.

            And yes, the daily deaths are still low enough to be accident victims testing positively, but we really need to know who the serious cases and fatalities are today!

          6. Eric

            Glad we’re on the same page on this. Are we banging our respective heads collectively or do you know but won’t tell?

            On a more serious note, this must be perfect stuff to do a few papers on. I wonder why they haven’t surfaced yet.

          8. Eric

            Again, this is possible but you’d have to assume some malice for it to be true.

            France publishes infections, hospitalizations, ICU admissions, and deaths by region.

            Germany publishes infections, ICU admissions, and deaths by county.

            All these indicators besides infections were still giving cause for optimism well into August but not so much recently. The big question is, what is the makeup of the people experiencing serious illness or dying? Whatever the answer is, it does not seem to be newsworthy.

          9. Eric

            Read something about Paris yesterday, where ICUs are filling up again. One doctor was quoted that a third of those in ICU are under the age of 65.

          10. AhNotepad

            Eric, there now comes the difficulty of distinguishing between genuine virus symptoms and those caused by psychological pressures, and the non-pharma measures used to indoctrinate the population. Bad enough for adults, https://youtu.be/ektfBatKG3w far worse for children.

  57. Philip

    So I have seen that the policy in the UK now is that if someone tests positive, they are tested again and only get counted as a case with 2 positive tests…is this correct and would it significantly reduce the potential for false positive tests?

    Reply ↓
    1. Eric

      If the reasons for false positives were random, this would be the case. However, in case of remnants of other Corona viri, one would need a different test that tests for a different gene or other marker of the virus.

      My understanding is that people do not get another swab, but rather another test is run on the same sample. Some labs will do that in the first run, i.e. analyze for 2 or 3 different targets at the same time.

      Reply ↓
      1. KJE

        Off topic, but it really irritates me when people write”viri” which means “of a man” or “men” when they actually mean viruses. Of perhaps you really did mean Corona men – the guys who used to deliver that brand of fizzy pop from a wagon. Obliviously didn’t even do O level Latin.

        Reply ↓
        1. Eric

          I wasn’t entirely awake yet when I wrote this (different time zone), but thanks for the condescension. If you recall, we have discussed this before, and I had plenty of Latin. But maybe that is why you wrote obliviously?

          Reply ↓
        1. Eric

          Mapping a neologism back to classical Latin is not straightforward, nor is deciding how to form the plural in a modern language or using Latin to argue how it should be done, and different modern European languages have found a different solution, which constitutes potential for tripups.

          Let’s take fruit. This is an uncountable or mass noun in English when you talk about say a basket of cherries, but there are also different fruits, such as apples and oranges.

          Back to virus. Viruses is accepted mainstream, but there is some evidence of usage of viri from the 19th century or more recently when talking about computer viruses. Usage is usually the best indicator in linguistics, rather than theoretical constructs.

          When we had that conversation before, someone proposed vira, which KJE and I agreed was indefensible from the point of view of classical Latin.

          If you subscribe to the explanation that virus goes back to virus with a long i for poison, viri is also indefensible. My latin teacher used to say that the modern word virus was coined because people were impressed that is could fell strong men. He didn’t explain the nominative, though…

          So my question to native speakers: if you are talking about a container containing only one species of virus, would you use the singular or plural?

          Reply ↓
  58. Eric

    First of all, German new cases have “shot up” from ~300/day in the first week of July to about 1800/day now, so maybe the growth rate was not as high as elsewhere, but it is certainly a multiple of what it used to be.

    The round robin report said that there was a good number of international labs involved. However, the test design was target specific. I have not read all of the report, but while they give target specific rates, they don’t seem to talk about dual target testing as a requirement. So maybe people can find out locally how testing is being done? *

    By the way, the non-profit that did the round-robin has an English language page (likely with the report available in English), and there seem to be more round-robins coming up:
    https://www.instand-ev.de/en.html

    * From today’s guardian:
    https://www.theguardian.com/world/2020/sep/28/coronavirus-world-map-which-countries-have-the-most-covid-cases-and-deaths

    At the very bottom of the article, there is a list of reported cases and deaths in the last two weeks:
    Spain 145,099 1407
    France 140,471 711
    UK 64,785 337
    Netherlands 25,893 108
    Germany 23,617 101
    Italy 21,374 218
    Sweden 5,880 29

    With the caveats that deaths reflect the infections that happened about four weeks earlier, that age structures may be different, and dynamics have been different, I still get the impression that the UK, France, the Netherlands, and Germany have roughly the same ratio of new cases and deaths, whereas Spain and Italy are ~2x higher. My reading is that the case numbers for the UK and France are therefore not inflated by false positives, and that Spain may be undertesting.

    Reply ↓
    1. Yvonne van Eck-Remmers

      Don’t trust the numbers coming from NL, last week more then half of the reported deaths were actually deaths that occurred in.. April 🤬

      Reply ↓
    2. Jennifer

      Mike S. My paranoid thoughts exactly. And as to the text message I received yesterday urging me to download the NHS app……not on your nelly.

      Reply ↓
  60. Steve

    The real bombshell in this article is the revelation that one of the primers used in the testing has the exact sequence homology with a region on human chromosome 8.
    I would like to know if all humans or just a part of the population have this chromosome, if the answer is all or most then surely the politicians and their advisers have some serious questions to answer.

    Reply ↓
    1. AhNotepad

      Steve, from what I see, most politicians do not answer most serious questions. They might answer a question that they were not asked, and if it had been, it would have been trivial.

      Reply ↓
  61. IMoz

    Back in the real world, according to the ONS, deaths for week 38 (ending 18 Sep) of 2020:-

    Total: 9523 (down from 9811 prev. week, up from 8945 5Y average)
    ‘flu/pneumonia: 1197 (up from 1125 prev. week)
    COVID-19: 139 (up from 99 prev. week)

    ‘flu/pneumonia:COVID-19 ratio is just over 8.6:1 (11.4:1 prev. week), COVID-19 mentions as a proportion of all-cause deaths: 1.46% (prev. week 1%), ‘flu/pneumonia: 12.57% (prev. week 11.47%). Looks like a seasonal respiratory infection rise to me. Why are those nosophobic of COVID-19 not getting referred to psych. yet???

    source: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsregisteredweeklyinenglandandwalesprovisional/weekending18september2020

    Reply ↓
    1. Trish C

      The real world figures don’t make for good headlines. Let’s stick with propping up the worldwide nosophobia. And thanks for introducing me to a new word 🙂

      Reply ↓
  62. Ken Garoo

    “Indeed, currently, there is no current agreement as to what ‘infected’ means with COVID19. Does it mean finding viral particles in the nose, sputum, or throat – or all three? Does it mean finding viral particles in these places, and also isolating it in the bloodstream, or lungs? Does it mean finding evidence of antibodies specific to COVID19 two to three weeks following ‘infection?’ Or what? It would be nice to know.”

    In early 2020, the Chinese published their protocols for confirming Covid-19 diagnosis. That involved two PCR tests taken two weeks apart and a chest x-ray. This was presumably in symptomatic cases. They stated that the PCR test on its own only worked in 60-70% of cases – again symptomatic cases.

    One question – are the diagnostic tests for a symptomatic ‘drive-by’ individuals carried out in the same laboratories as the tests for symptomatic, possibly hospitalised individuals? If so, there is a significant potential for cross-contamination symptomatic to asymptomatic.

    Reply ↓
  63. Ken Garoo

    Another question: Are the same number of cycles used for symptomatic and asymptomatic cases? If the number of cycles is considered the gain of an amplifier, and the primers, a narrow band-pass filter, then (pseudo-)random (asymptomatic individuals) ‘noise’ as input will produce a an audible output if the amplifier gain is high enough.

    Reply ↓
    1. JohnC

      As far as I’m aware there’s no difference in cycles as I doubt that the symptom status is included on the patient information.

      Reply ↓
      1. Terry Wright

        absolutely: no difference

        you just keep flogging the test; flogging the number of cycles;

        till you eventually get what you want: something you can call a “positive”

        …. just keep going; that is the mantra

        Reply ↓
  64. Ken Garoo

    Regarding a ‘gold standard’ test, someone at the 53rd SAGE meeting on Covid-19 held by Zoom on 27 August 2020, claimed that PCR is the ‘gold standard’.

    “Tests used for mass population testing particularly in low prevalence settings and
    populations could result in higher false positives than symptomatic testing using lab-based PCR tests, which could reduce public confidence in testing. Double testing may be required to reduce false positives (with PCR as the gold standard).”

    The minutes for the series of SAGE Covid-19 meetings are available at:

    https://laworfiction.com/2020/06/sage-minutes-all-in-one-document-searchable-and-indexed/

    Reply ↓
    2. AhNotepad

      Ken, thanks for that link, it leads to a document for those who would like to know the legal position of mask wearing requirements.

      Reply ↓
  65. Mike Smith

    Sooooo, if we fast forward past the part that suggests we now have a (fake) casedemic on our hands. Eventually some of those false positives will die, especially if poorly people are being tested. Is that essentially what we are now seeing in the daily death figures ? People who happened to have died and also thrown a false positive ?
    The country has become like a death circle of ants, mindlessly going around and around with no particular plan in the vain hope that we will have a couple of months of zero Covid deaths, which of course can not ever happen.

    For all we know, we could be completely over Covid now but due to the noise in the testing data we will never know!

    Utter and complete madness.

    Reply ↓
    1. Patrick Healy

      Mike if you think things are crazy in Britain, take a look a the real lunatic Mayor or New York New York.
      He says he will keep New Yorkers in house arrest until there are zero cases of this Wuhan Flu.
      I think what this Marxist meant “or until President Trump is defeated”

      Reply ↓
    1. Mike S

      Or are even false positives this time around ? Or more likely that the PCR test is picking up viral fragments from the first infection ? Its all bonkers

      Reply ↓
  67. Malcolm Ramsay

    Malcolm, you say “… some studies quoting figures as high as 50%. That is, half of those told they are not infected with COVID19, are probably infected”

    That’s a bit misleading,. To make any sense, the false negative rate has to be related to the number of (true) positive results, not to the number of tests done – it would be absurd to relate it to tests done.

    According to Wikipedia, “the false negative rate is the proportion of positives which yield negative test outcomes with the test, i.e., the conditional probability of a negative test result given that the condition being looked for is present“. So (with 100% specificity) a 50% false positive rate would mean that, for every positive result there is another that is missed; a 20% false positive would be one in five missed; 2% would be one in fifty missed.

    Reply ↓
  68. Martin Back

    RNA is bound up tightly within a virus or cell. It has to be released and converted to DNA to be tested. This is quite a process. The following is summarised from https://www.assaygenie.com/rna-extraction-for-covid-19-testing

    Most RNA extraction kits contain sterile buffers, solutions that lyse cells in order to access the RNA within the cells
    – First, the patient’s sample is mixed with the solution that lyses the cells in order to release the genetic material.
    – Then, cells are then treated with a buffer to inactivate RNases that may interfere with the RNA extraction step. (Cells and the environment secrete high concentrations of enzymes that destroy nucleic acids.)
    – Then RNA is extracted, usually by organic extraction which includes separating RNA from DNA by coating the sample in acidic solutions containing guanidinium thiocyanate, sodium acetate, phenol and chloroform.
    – This is then followed by centrifugation, which physically separates the RNA from DNA and other proteins. The RNA remains in the upper aqueous phase, whilst the DNA and other proteins stay in the middle or settle to the bottom.
    – Total RNA can then be recovered by transferring the upper aqueous phase containing the RNA into a new tube and precipitating the RNA by using isopropanol, a less concentrated version of ethanol.
    – This sample is then centrifuged, and the RNA will form a pellet at the bottom of the tube and can be isolated.
    – RT-PCR is the process of converting RNA into cDNA. It uses RNA as a template instead of DNA, which is seen in the standard PCR method. First, the enzyme reverse transcriptase uses the RNA template to produce a complementary single-stranded DNA strand called cDNA. This process is known as reverse transcription.
    – After, the enzyme DNA polymerase is used to convert the single-stranded DNA molecule into double-stranded DNA which can be used as templates for a PCR reaction. This allows the double-stranded DNA to be copied and amplified exponentially with the use of primers and thermocycling .

    Reply ↓
    1. David Mullen

      I would be a little careful of any thing that describes isopropanol as “a less concentrated version of ethanol”.

      Reply ↓
  69. Jerome Savage

    Long transcript here Malcolm if you will allow. Its Irish politician V top medical bureaucrat today and is telling.
    “HSE CHIEF CLINICAL OFFICER ADMITS PCR TEST FOR COVID CAN GIVE FALSE RESULTS AND SAYS HE DOESN’T KNOW HOW MANY CYCLES ARE BEING USED TO COME UP WITH TEST RESULTS. THE HIGHER THE NUMBER OF CYCLES THE LESS ACCURATE THE TEST AS IT CAN PICK UP REMNANTS OF HISTORIC VIRAL INFECTIONS
    This morning the Chair of Dáil Covid Committee Michael McNamara T.D. asked Dr. Colm Henry, Chief Clinical Officer with the HSE about the risk of false positives and false negatives in PCR testing.
    He admitted “One of the problems with this test is that it can pick up a residual RNA weeks after active infection and after a person is no longer infectious.”
    Scientific studies show that this is more likely to happen after 30 cycles or more.
    Dr Henry was repeatedly asked about the number of cycles the HSE use and then admitted that he didn’t know”. Video link is available.
    “Transcript
    Chairman
    I will ask a brief follow-up question about testing. There has been a lot of discussion, particularly on social media, about PCR testing. Some international doctors are claiming that the number of cycles used in a PCR test means that a person can be shown to be positive if he or she had the virus some time ago but are no longer symptomatic. Equally, I have spoken to Irish doctors who have said that they are more concerned with false negatives than false positives. Would Dr. Henry comment on that?
    Dr. Colm Henry
    That is a good question. On false negatives, the PCR test is the most sensitive one we have but it does not provide absolute sensitivity. The sensitivity relates to and correlates closely with how symptomatic an individual is.
    Clearly, if somebody is actively shedding the virus and is quite sick then one is much more likely to get a positive test when that person actually has the virus. When somebody is asymptomatic it has lower sensitivity. Such is the nature of all tests that they have a stronger sensitivity in symptomatic and asymptomatic individuals. I ask the Chairman to repeat his second part of his question.
    Chairman
    What number of cycles of testing are likely to show that people have had the virus a very long time ago, are not symptomatic and, maybe, not even be shedding anymore? There is also the issue of a margin of error that is part of all tests.
    Dr. Colm Henry
    Yes. The manufacturer defines the number of cycles that go with the test. We do know because the PCR test picks up on RNA it is possible to find residual RNA after infection has resolved and the person is no longer infectious. We have seen that, and sometimes in some cases, for many weeks after a person ceases to be sick and infectious. One of the perks or problems with the test is that it can pick up on residual RNA weeks after active infection and after the person is no longer infectious.
    Chairman
    How many cycles are utilised in Irish testing?
    Dr. Colm Henry
    I cannot answer that but I will come back to the Chairman with an answer.”
    Video https://www.facebook.com/177185598977651/posts/3840988352597339/?vh=e&extid=LMI8ah7Y3sv7doZ9
    Transcript
    https://www.oireachtas.ie/en/debates/debate/special_committee_on_covid_19_response/2020-09-29/3/
    A contributor to the page suggested it was 45 cycles.
    As a lay person, I have difficulty understanding the logic in his comment; “Such is the nature of all tests that they have a stronger sensitivity in symptomatic and asymptomatic individuals”

    Reply ↓
    1. goransjoberg2015

      In Sweden People now seems more scared. Entering a huge elevator today the only guy inside raised his hand and said I could not enter claiming Covid.

      I am fed up!

      Reply ↓
      2. Dan Anderson

        If one person wants to ride an elevator alone, and you don’t mind riding with two aboard, then you get on despite the “halt” order and say “you wait for the empty car you want, this one is fine for me.” Individuals acting in their own best interests seems reasonable, but YMMV.

        Still, looking at the couple of hundred deaths next door in Norway, and the few hundred next door in Finland, and compared to the 5900 deaths in Sweden, with 500 new cases a week, it might pay for a vulnerable person in Sweden to be extra cautious when possible. Contagion is a matter of numbers, of exposures — of habits of hygiene, physical spacing, masks when no natural or vaccine immunity is present and no treatment available.

        Reply ↓
  70. Colin MacDonald

    Looking at data from Scotland the percentage of positve cases has increased from 0.5 to 5 since July. If the false positive rate is around 5% how do we account for the much lower covid-19 rate in July? The simplest explanation is that there has been a genuine increase in positive tests. I wonder though if positive tests were immediately re run back in July, a realistic thing to do when we were only reporting 10 positives a day, if our anonymous PCR scientist is correct then surely that’s exactly what would happen in a well run testing station. However as the case load increases there simply isn’t the capacity to do this. So perhaps some, perhaps most of the increase is down to an inability to double check the results.

    Reply ↓
  71. Steve

    Perhaps it’s time to concentrate more on the psychiatric aspects of this medical event which we are living through rather than the epidemiology.
    Today I met some friends who I haven’t seen since January and I was shocked. They both used to be pragmatic and relaxed about risk but have now become paranoid and risk averse and became very agitated when I dared to suggest that there was an alternative narrative to the official one. They talked about a forty year old nephew who will not visit his thirty seven year old sister for fear of infection and will only go to work and return home to shut himself away with his equally terrified wife.
    Their basic attitude was that if we relax for one minute and stop following orders the virus will rage through us all and wipe us all out.
    Are there any psychiatric specialists on this blog who would care to comment on what I have described.

    Reply ↓
    1. Jerome Savage

      One Case of compulsive cleaning, an obsessive compulsive disorder had a particularly sad outcome – that I am aware of, Am just surprised there are not more.

      Reply ↓
    2. Janet Love

      Also a fulfillment of the commentary in Luke Ch. 21 – in this context, ‘men’s HEARTS “failing” them for …. FEAR.
      In short, a generated emotion is doing the damage, and the implication being this FEAR is /will be, more damaging than the observed or imagined catastrophe.
      As a weapon of conquest, ‘FEAR’ is brilliant. It’s relatively cheap, easily delivered to selected targets and best of all, not only self-sustaining – it self-replicates at the expense of it’s host.
      Like a virus.

      Reply ↓
  72. JB

    Click to access Dr.-Lees-paper-on-testing-for-SARS-CoV-2.pdf

    Testing for SARS-CoV-2 in cellular components by
    routine nested RT-PCR followed by DNA sequencing Sin Hang Lee 1*

    … Using this protocol to re-test 20 reference samples prepared by the Connecticut State Department of Public Health, the author found 2 positives among 10 samples classified as negative by RT-qPCR assays [2 false negatives]. … Of the 10 samples classified as positive by RT-qPCR assays, only 7 (7/10) were confirmed to contain SARS-CoV-2 by heminested PCR and DNA sequencing of a 398-bp amplicon of the N gene [3 false positives]. …

    One group of scientists in Australia tested a commercial RT-qPCR test kit and found its positive predictive value to be only 55.56%. The authors suggested that any positive results derived from one commercial test kit should be confirmed using another nucleic acid test or nucleotide sequencing [6]. …

    Reply ↓
  73. JB

    In the U.S., hospitals want their Medicare bonus for Covid patients WITHOUT having to do a PCR test, citing a possible huge false negative rate of 37%!
    ——
    https://www.foxbusiness.com/money/hospitals-covid-19-test-requirement-medicare-funding
    ——
    Meanwhile, according to the CDC, if you’ve got a headache and nasal congestion, you can be diagnosed with Covid. And if anyone can decipher the last sentence below for me, the sentence with the words “utility” and “validity” in it, I would be greatly appreciative!
    ——
    https://wwwn.cdc.gov/nndss/conditions/coronavirus-disease-2019-covid-19/case-definition/2020/08/05/

    Clinical Criteria

    In the absence of a more likely diagnosis:

    At least two of the following symptoms:
    fever (measured or subjective),
    chills,
    rigors,
    myalgia [muscle pain],
    headache,
    sore throat,
    nausea or vomiting,
    diarrhea,
    fatigue,
    congestion or runny nose
    OR

    Any one of the following symptoms:
    cough,
    shortness of breath,
    difficulty breathing,
    new olfactory disorder,
    new taste disorder
    OR

    Severe respiratory illness with at least one of the following:
    Clinical or radiographic evidence of pneumonia,
    Acute respiratory distress syndrome (ARDS).

    Laboratory Criteria

    Laboratory evidence using a method approved or authorized by the FDA4 or designated authority:

    Confirmatory* laboratory evidence:
    Detection of severe acute respiratory syndrome coronavirus 2 ribonucleic acid (SARS-CoV-2 RNA) in a clinical or autopsy specimen using a molecular amplification test

    *The terms confirmatory, presumptive, and supportive are categorical labels used here to standardize case classifications for public health surveillance. The terms should not be used to interpret the utility or validity of any laboratory test methodology.

    Reply ↓
    1. JohnC

      I think that this is what it means:
      It looks as though there are three classes of tests that they use. However, although a test has been given a classification of confirmatory, it cannot be inferred from the classification given that the result is necessarily meaningful. It just means that the test methodology meets the requirements of the classification.
      If you have a test that looks for parameters A B C then that is considered supportive
      If you have a test that looks for parameters A B C D E then that is considered presumptive
      If you have test that looks for parameters A B C D E F G then that is considered confirmatory
      What you cannot deduce from the classification is the QUALITY of the results obtained from a given version of the test.
      Consider the old roadside breath test with the crystals that changed colour, that produces a result that shows an illegal blood alcohol level, that’s all. This I would categorise as presumptive. The next step is a urine or blood sample, this produces a number and I would classify that as confirmatory.
      Even the roadside tests that show a number are still presumptive, they still need a second test.
      These tests have to be regularly audited and checked otherwise their results cannot be considered valid.
      What the CDC are saying here is that the tests are not necessarily audited for quality control purposes.

      Reply ↓
  74. JB

    I had a contractor at my house earlier this summer and he told me that he has high blood pressure and that he’d been on a new drug for it lately, and that ever since he got on the new drug, he’d had a dry cough that had been driving him crazy. I thought that was really interesting. Remember when “dry cough” was the first telltale Covid symptom that the media was going on about?
    —-
    https://themedcircle.com/6-medications-causing-unexplained-cough/

    14 Medications Known to Cause Persistent Dry Cough

    Reply ↓
    1. James DownUnder

      For me it was candesartan. – smugly identified by my GP. as ‘Candesartan Cough’ – Proof of compliance in taking all my pills… :/

      Reply ↓
      1. Steve-R

        Classic side effect of some ACE inhibitors.

        Just started to get if after about 3 years and a change to a cheaper version, so I stopped. Still here 10 years later so I doubt they were as good a losing weight taking exercise and avoiding polluted traffic fumes n=1.

        Reply ↓
  76. Martin Back

    Viruses swap genetic material fairly indiscriminately. Is it possible that the increase in positive tests is a result of a related coronavirus acquiring the RNA sequence the PCR test amplifies, perhaps from someone who has a cold or flu and Covid-19 at the same time?

    ” Recombination involves the exchange of genetic material between two related viruses during coinfection of a host cell… Recombination between two retroviruses gives rise to novel viral progeny with reassorted genes. “ — Chapter 43 Viral Genetics
    https://www.ncbi.nlm.nih.gov/books/NBK8439/

    Reply ↓
  78. anndevlin

    How about this for a false positive? I have some friends who had an appointment to go for a test. They turned up at the test centre but the wait was too long so they decided to go home. A few days later they received their test results in the post – CV POSITIVE. I’d love to know how they worked that one out.

    Reply ↓
  79. IMoz

    So this morning I realised we (collectively people who can do maths) explain this false-positives in a wrong way…

    Here’s a simpler explanation that most can relate to:-

    Everyone knows home pregrancy kits. The requirement for using them (it expressly tells you on the insert) is that you need to have missed your period. Having done a cursory search, the false positive rate for those tests is between 1 and 4 per cent (which is actually in line with what SAGE guesstimated qRT-PCR tests were); bear in mind that that is 1–4% if you test as indicated: actually missed your period! So for 1000 women who missed their period and had unprotected intercourse upto 40 will be “tested” as pregnant when they are actually not, which is why the insert to those tests actually tells you to get a doctor’s appointment to do a proper investigation to determine if you are pregnant.

    The (un)common sense suggests that if you were using the pregnancy kit and still haven’t missed your period, the likelihood is that the false positive rate would be higher. Now, let’s take it further, and let’s say that even if you hadn’t missed your period and carry out the test, the false positive rate remains at 1–4% but let’s now get men to pee on the stick! For every 1000 men, you will end up with upto 40 “pregnant;” that, ladies and gentlemen is your qRT-PCR testing of “asymptomatics”—40 pregnant men in 1000, now scale that up by how ever many tests are done on daily basis! For some “strange” reason even the MPs and those making arbitrary rules would find the pregnancy in men ridiculous…

    Reply ↓
      1. IMoz

        This is the danger with analogies 😉 The problem with the false positives is that axiomatically it shows positive when there is no beta-HcG

        Reply ↓
      1. Tish Farrell

        You do not surprise me! But where on earth is all this going? I’ve just been listening to Dr. Sucharit Bhakdi, one of Germany’s most eminent medical scientists, who is highly concerned about any vaccine for this particular virus – whether conventional or RNA. He believes that between 80-90% of Germans, and pretty much everyone else, must have degrees of existent immunity; that our immune systems remember infections by other corona viruses which of course are extremely common. He and his wife, Dr Karina Reiss, have the English translation of their book coming out tomorrow, ‘Corona: False Alarm’. On amazon unfortunately, but at least it’s available (so far).

        Reply ↓
        1. AhNotepad.

          No need to pay the South American river traders, that ebook is available for free, and there is a link somewhere on this blog. Search for “corona false alarm”.

          Reply ↓
  82. David Bailey

    I don’t know if it is proper to copy this letter, which I found in the Daily Telegraph, so I leave it to Malcolm to decide:

    SIR – The vast majority of students who graduate from medical school pursue careers that use the tenets of clinical medicine. A correct diagnosis is typically dependent on a clinical examination using all of the doctor’s senses. In other words, clinical 
doctors are wholly immersed in a life of risk.

    A tiny minority of medical graduates pursue a career in public-health medicine. That these scientists are intelligent is without doubt, but they inhabit a world of epidemiology and theoretical modelling, which keeps risk at a distance. No wonder, then, that when the Cabinet seeks their advice about minimising death, the message is to lock down and shield until the problem goes away.

    As a general surgeon, I have not encountered a single Covid-positive patient for more than four months. In this same time, my clinical activity is running at 20 per cent of what it normally would be and the expected numbers of referrals for patients with cancer has fallen dramatically. Yet I am now working in a region that has gone back into lockdown.

    The Government is following the advice of a population of doctors who are risk-averse. It is possible to protect both the economy and the vulnerable, and to resume treating more lethal diseases. However, this requires the Government to seek and follow advice beyond the world of public-health medicine. It should listen to the Royal College of Surgeons and to oncologists.

    David Scott-Coombes FRCS
    Llysworney, Glamorgan

    Reply ↓
  83. unicodetantos

    Dr. Kendrick:
    I’m from Spain and I usually read your blog. It’s really refreshing to find people who are actually thinking and not in panic.
    I agree with your conclussion in this post, but I have a question.
    You have wrote this:
    “Using this one in a thousand figure. This means, if you randomly tested ten thousand people, you would expect to find ten COVID19 cases [forgetting the false negatives for now].
    On the other side of the coin. If the false positive rate is one per cent, you would have an additional one hundred false positives cases”.
    I thought that the false positive rate is the number of false positive tests among the tests that have given a positive result. But if I understand your explanation, you are saying that the false positive rate is not exactly that…
    Probably, the problem is my english… Could you clarify this question?
    Thank you so much.

    Reply ↓
    1. Malcolm Ramsay

      “I thought that the false positive rate is the number of false positive tests among the tests that have given a positive result.”

      I don’t know how Malcolm Kendrick finds time to engage with all the comments so I’ll answer this for him.

      The number of false positives is not related at all to the number of people who truly have the virus, it’s only related to the number of tests done on people who don’t have it; it’s the percentage of the people tested who are in fact clear of the virus but show as having it.

      So, if there’s a false positive rate of 1% and you test 1000 people who don’t have it, you’ll get 10 positives (all of which are false). If, instead, 100 of the them had it and 900 were clear, then (ignoring false negatives) there would be a total of 109 positives – 100 of them true, and 9 false. So false positives aren’t very significant when the overall prevalence is high but are very important when it’s low.

      If the false positive rate was related to the number of positive results (as you understood it to be) the number of false positives would go up in line with the number of true positives – which would be very strange!

      I hope that’s clearer.

      Reply ↓
  84. The Dim Appear

    Does it follow that if someone tests positive for COVID using PCR and they really do have COVID that a subsequent antibody test would confirm this? Is there an antibody test?

    If blood was taken at the same time as the swab and then a couple of weeks later could this prove how accurate PCR is depending what the antibody results show?

    I’m guessing if it could they’ll never do it.

    Reply ↓
      1. Steve-R

        They probably don’t make IgG and IgM if their T Cells and possibly IgA antibodies clear the infection. If you only test for IgG and IgM and don’t find them it is wrong to conclude that antibodies weren’t made.

        Reply ↓
    1. James DownUnder

      Research ONLY ? – Surely using it to “diagnose” a Dread Disease is ‘Research’ – Research into terrifying a nation into mindless OBEDIENCE and economic self-destruction.!!!

      Reply ↓
  86. JohnC

    From https://www.bbc.co.uk/news/health-54353888
    “The researchers found that older adults admitted to hospital who were classified as frail were more likely to have had delirium as one of their symptoms, compared with people of the same age who weren’t frail.”
    “But it’s not yet clear why this extreme confusion or delirium happens”
    Why are they surprised?
    it is known that older people have increased confusion from bacterial infections such as UTIs.
    It’s also a red flag for sepsis for all ages. It also happens in dehydration.
    Its one of the criteria in assessment of community acquired pneumonia CRB65.
    IIRC confusion is one of the criteria for deciding if a person is frail in the community

    Reply ↓
    1. Martingale

      JohnC My wife works in the French retirement home system of EHPADs. In March, from observations, people were warned to look out for diarrhea and confusion, including people taking a fall who didn’t normally fall. When these 2 criteria were observed, they had about a week before “normal” Covid symptoms appeared, and had time to isolate them. (Currently there’s nobody with symptoms, and it’s been quiet for months).

      Reply ↓
  87. Ivan Lowe

    The Base Rate Fallacy, which is at the core of this discussion, is counter-intuitive and writers are struggling to explain it. May I draw your attention to successful writers?

    Fenton N and Neil, M 2010 Comparing risks of alternative medical diagnosis using Bayesian arguments
    Journal of Biomedical Informatics vol 43 p485-495 https://pubmed.ncbi.nlm.nih.gov/20152931/

    The authors use various ways to visually explain false positives and false negatives and the Base Rate Fallacy. They also use a flow diagram which is simple to use and understand.

    I tried out several online calculators. This one http://araw.mede.uic.edu/cgi-bin/testcalc.pl
    also allows you to punch in the data in different ways and presents results in several ways including a very helpful flow diagram.

    When teaching the base rate fallacy I used to encourage my students to play with the figures to see what happens when the numbers change. In particular, whenever the incidence of a disease (= prevalence) is 1% or lower , then unless that test is super-accurate (which in the real world is unlikely), then most (verging to all) of the positives will be false. This is the way numbers work.

    I have found it easy to say to people that a true case will have symptoms meriting treatment and the government is not using language fairly when they label those positive PCR tests as positive cases.

    Reply ↓
    1. Martingale

      Ivan Lowe The Fenton and Neil paper is great! you only need to look at figures 1 to 3 to “get it”, the rest is good to read, but that first part just stands out in its simplicity. I’ll try this out on people around me.

      Reply ↓
  88. Jerome Savage

    Interview by Celia Farber in April with David Crowe– Canadian researcher, with a degree in biology and mathematics, host of The Infectious Myth podcast, and President of the think-tank Rethinking AIDS
    It debunks the conspiracy element to an extent but does point the finger at usual suspects to another extent. Celia does open with some colour but if we can get over that, the remainder of the article is a treat.
    Crowe – “They don’t know how to end this. This is what I think happened: They have built a pandemic machine over many years and, as you know, there was a pandemic exercise not long before this whole thing started.” ” So, this beautiful pandemic machine is a lot like…let’s use an example of an aircraft simulator. So, pilots are tested on an aircraft simulator. if you’re flying and there’s a loud bang and you see smoke coming from an engine on the right hand side, this is probably the first time a pilot has ever been in an airplane that had an engine failure. But he’s tested this scenario 25 times on an aircraft simulator. So, he knows exactly what to do without being told. He goes through the procedure. He doesn’t have to think” “So, a pandemic simulator is just like that. But there’s, there’s never been like an actual real pandemic since they built this machine. So, there’s this huge machine, it’s got a red button on it and it’s like if you ever detect a pandemic starting, you press the red button. We don’t know exactly what happened, but I think the Chinese government was embarrassed cause they were being accused of covering up a pandemic. They said, okay, you know, we want Western approval for our medical system so we’re going to press the goddamn red button. Or they did. And then everything followed from that. The problem is that the simulation was never based on reality.”
    He is quoted as saying “So, we’ve essentially been taken over by the medical Taliban, if you like”
    Oooh – strong stuff says I.

    Was the COVID-19 Test Meant to Detect a Virus?

    Reply ↓
    1. AhNotepad

      A paragraph from the link:

      “The people who have taken all your freedoms away in recent weeks, they’re social engineers, politicians, globalist thought leaders, bankers, WHO fanatics, and the like. Their army is composed of “mainstream media,” which is now literally a round-the-clock perfect propaganda machine for the Gates-led Pandemic Reich.”

      This is a “must read”.

      Reply ↓
    1. AhNotepad

      I don’t have confidence in the sciencedaily article when I read this “the research shows that in both the United States and in Europe, sustained transmission networks became established only after separate introductions of the virus that went undetected.”

      Where does this come from? It looks a bit speculative to me.

      Reply ↓
  90. elizabethhart

    Malcolm, re ‘COVID19 tests’…shouldn’t these be called SARS-CoV-2 tests?

    I’m confused by the terminology. In this regard see below my rapid response published in The BMJ on 21 September 2020:

    Covid-19 vaccines…or SARS-CoV-2 vaccines? Clarification needed
    https://www.bmj.com/content/370/bmj.m3258/rapid-responses

    Is it correct to use the term Covid-19 vaccines?

    Shouldn’t these vaccines instead be described as SARS-CoV-2 vaccines?

    It grates with me to use the term Covid-19 vaccine, it doesn’t seem right.

    Terminology was similarly incorrectly used with the initially titled ‘cervical cancer vaccines’, which are now more correctly described as HPV vaccines.

    Surely the vaccines should be described as SARS-CoV-2 and not Covid-19?

    This is an important distinction and it should be correct – can we have clarification on this?

    Reply ↓
      1. Steve-R

        I, for my sins, end up listening to a lot of phone-ins and if the correspondents are typical of the ‘everyone’ you mention then it is not the case that they know what is meant.
        Perhaps an understanding of human diseases needs to be included in the junior school curriculum to try, in time, to dispel the rule of General Ignorance in public discourse on all matters of health.

        Reply ↓
        1. AhNotepad

          Steve-R, With the current lies that are spun by pharma and the subsequent bias inflicted on the medical profession, the chance of giving junior school children something approaching facts would be small.

          Reply ↓
      2. elizabethhart

        I disagree. As you have said previously Malcolm, “…terminology really, really matters”.

        Despite all the ongoing fear-mongering, it’s my understanding the SARS-CoV-2 virus isn’t a problem for most people, particularly the young, i.e. they’re unlikely to be troubled by the disease associated with SARS-CoV-2, i.e. ‘COVID-19’.

        But calling the vaccine ‘COVID-19’ insinuates it’s preventing this disease, when in many cases the virus doesn’t proceed to serious disease anyway.

        I’m not claiming medical expertise in this area…but who can actually? It’s pretty shocking how clueless so many so-called ‘medical experts’ are about this virus. Here we are, 2020, and it’s surprising how ill-equipped the medical industry is to deal with a virus outbreak.

        Of course, from the get go, it’s all been about ‘the vaccine’… This relatively benign virus is just the mechanism for the real game, which is Bill Gates’ desire to vaccinate the entire world, along with his camp followers. A humungous vaccine market is riding on the back of this virus, and who knows what else…

        I stick with my argument, it is incorrect terminology to call this vaccine ‘COVID-19’, it should be called SARS-CoV-2 to ensure people are properly informed.

        Reply ↓
    2. JohnC

      People talk about the flu vaccine (pick and mix), shingles vaccine (Herpes Zoster), chicken pox (Herpes Zoster) vaccine when they are talking about the illness not the virus. I think that referring to the illness associated with the virus when discussing the vaccine is not too bad. However, I get really narked when they refer to Covid-19 cases, tests or infections when they really mean SARS-COV-2, as you cannot have an asymptomatic covid-19 case; the test is for SARS-Cov-2, and the infection is from SARS-COV-2.

      Reply ↓
    1. KJE

      I had an exam candidate with a big bushy beard yesterday. We supply those cheap blue masks that you can get in supermarkets and his sat so far out on top of the beard that there’s no way it could even have stopped spitting, let alone a tiny particle of something, but we complied with the needs of our COVID-safe risk assessment

      Reply ↓
      1. IMoz

        How anyone in the government has the “brain” to come up with the idea that a piece of cloth that gets saturated with water, bacteria, fungi, and viruses is a good way to *reduce* respiratory infection is mind-boggling, to call it merely idiotic would be to grossly insult the idiots; and then have the audacity to make laws to mandate it!

        I asked a few people yesterday to show me the face-side of their masks yesterday: they were light orange/brown (and I have no idea with what) instead of bright white as they were at the start! This got me thinking: just think what level of contamination the dark-coloured face masks that are so prevalent on the market have, and people breathe through that!

        But what is actually worse is that so few doctors/ID specialists are actually speaking out against it… Luckily, it looks like some surgeons are beginning to publicly call out the BS now too.

        Reply ↓
        1. KJE

          Well, a) he’d have the same problem everywhere – forced shaving is not required b) all candidates are behind perspex screens and everyone knows that viruses can’t go over or round those c) there are a load of other methods of disinfection that we have to do – letter of the law and all that. And we all (except the government apparently) know that masks can’t stop viruses anyway.

          Reply ↓
  93. Terry Sanders

    For false negatives you say:
    “With COVID19, there are a lot of false negatives, with some studies quoting figures as high as 50%. That is, half of those told they’re not infected with COVID19, are probably infected”
    Shouldn’t that be:
    “Half of those infected with COVID19 are told they are not”

    Reply ↓
    1. Malcolm Ramsay

      I’m glad someone else noticed that. I commented on it a couple of days ago (less succintly than you) and Malcolm said he’d have to think about it. It doesn’t really affect the main thrust of what he was saying but I think it does undermine his credibility somewhat.

      Reply ↓
      1. KJE

        If you never get any symptoms, how would you now that it was a false negative or a true one – and if if you later did get symptoms, how would know you hadn’t caught it since the test? Basically the test just seem so inaccurate and the symptoms of the disease so diverse (now, not originally) that testing seems pointless and can be made to show anything. Lots of expense for no benefit. It is in the interests of the scaremongers to shout about false negatives and in the interests of those who want to go back to earning a living to shout about false positives – apart from probability theory, both sides seem impossible to prove. Why don’t we just treat the symptoms and cure the sick?

        Reply ↓
  94. Tish

    Can I ask those of us who believe the government has a hidden agenda, what they think the agenda might be? The vaccine is one big issue but what about the finance and a possible new order of jobs and society? And anything else?

    Reply ↓
      1. Tish

        AhNotepad,
        How ugly. Kings of the whole world – never been done before. What feathers in caps and their own grandchildren get their liberty safeguarded by wealth and power.
        I hope they remember Shelley’s Ozymandias.

        Reply ↓
        2. KJE

          Well, all they need is enough slaves to keep their services running and they can swan around hunting, having parties, surveying all they own and whatever. I think they haven’t factored in how bored they’ll get once they’ve got it. B&M G have three kids but you never hear about them, perhaps they are a disappointment

          Reply ↓
    2. Jerome Savage

      There is one other factor that is never considered. Its increasingly applying itself to every and any area of human activity and that is the avoidance of risk and any possible litigation. If a standard is applied, no matter how ridiculous, it’s strict adherance will guarantee lminimum liability. The US culture of litigation is insidious has & is spreading.

      Reply ↓
    1. Jean Humphreys

      Many years ago there was a state occasion where the Duke of Edinburgh was said to have toothache. Everyone believed it was because they could not trust him to be diplomatic in the way he did so well. There was a lot of joking about having diplomatic toothache.
      Can we suppose that POTUS is having a political virus – is his positive test true? After all a SNP MP has got herself inot deep trouble over a positive test. She went to Westminster, before getting the result. Nowhere can I find any mention of her being ill – you wouldn’t want to travel from Scotland to London on the train if you were feeling ill and feverish.

      Reply ↓
      1. KJE

        Sadly, MPs seem to be largely brainwashed and poorly-read. Mine actually believes that it is possible to completely suppress a virus and have a Zero-Covid strategy.
        “I am concerned at what many perceive to be an overall strategic approach taken by the Government in response to the pandemic that is about “flattening the curve” by suppressing the pandemic instead of pursuing a “zero covid” strategy to entirely eliminate transmission of the virus nationwide. There is evidence that many of the countries which are now seeing comparably low levels of community transmission of the virus, and consequently have been able to limit the economic damage of the pandemic, pursued this strategy of elimination and rapidly responded to localised outbreaks through an effective Test, Track & Trace system.
        Moreover, I am worried by the sharp rise in the number of cases in the UK since the end of August and at the Government’s strategy of telling people to return to work and compelling the reopening of schools and universities without first getting community transmission of the virus down to a manageable level.”

        I assume that’s the opposition line

        I really think they need to be educated as to what the rt_PCT test actually does and therefore what a positive test means, and also about the history of viruses and the fact that none (apart from possibly smallpox) have ever been eradicated and that viruses were probably here before humans and will after we have all gone. Labour obviously cares little for the livelihoods of working people.

        Reply ↓
    2. David Bailey

      It is remarkable that so many people in politics seem to have got the virus – and this must be very inconvenient for President Trump, and Boris’ bout came at a time when he was resisting pressure to start locking down. I wonder how many PCR cycles they used – 50?

      I think Trump is far more savvy than some people think, and hopefully he has the sense not to let himself go to hospital unless absolutely necessary.

      By contrast, most people are baffled because they have not encountered the virus.

      Personally I would try to avoid being tested.

      Reply ↓
        1. Steve B

          The Trumpists (Trumpers?) who comment on this site and don’t believe that that Covid-19 even exists must be confused.
          It must be fake news or a devilish plot by Democrats or a cunning plan by the Republicans to get the sympathy vote
          Please let us interested observers know.

          Reply ↓
          2. KJE

            Does anyone not in the USA give a stuff about what’s happening in America? I certainly don’t. Not that I really know about it as I don’t watch the news.

          3. Steve B

            The the USA still has great power and influence, so what happens there probably affects all our futures.

            If the UK has to go cap in hand to the U.S for a trade deal, the you’ll see what economic power the U.S has

  97. Dr No

    Dr No has written about these important problems a number of times, most recently in the post “The Sensibilities of Sensitivity and Specificity” (https://dr-no.co.uk/2020/09/25/the-sensibilities-of-sensitivity-and-specificity/). It is a huge problem.

    That the current numbers just don’t make any sense at all is also a huge problem, covered in various Dr No tweets – see https://twitter.com/_dr_no_ in particular this one (make sure you use twitter with images visible, so not on a mobile) https://twitter.com/_dr_no_/status/1309563714325098497 about how the numbers from the 1st Aug (twitter Pimenta/Hartley-Brewer ‘debate’ about #covid false +ves). This is another way of looking at the nonsense results your anonymous contributor alludes to.

    It is also very important to bear in mind there are two sorts of false positives. If we really wanted to get in a mess we could call them true false positives and false false positives… but let’s avoid that semantic nightmare, and see why they matter.

    The first sort of false positive is the one that appears when there is no viral RNA present. None, zilch. This is the one that features as the flip side of specificity, ie if the test is 99% specific, then the false positive rate is 1% (false +ves/(false +ves + true negatives)). These are the false positives that multiply when the prevalence is low (even a low percentage of a big number can be a big number).

    The second type of false positive is the test which is positive and correct (the viral fragment is there, so it gets counted as a true positive in the 2 x 2 table), but it is false in that the person testing positive does not have active covid-19 (the hair in the room from someone who left weeks ago idea). At present, how many such tests happen is an unknown…

    The other important thing not to lose sight of is the gold standard test for covid-19 RT-PCR tests has by and large been another PCR test, meaning in effect a sort of echo chamber validation. What the positive test is really telling you is that here is a person who will probably test positive to another PCR test. Not exactly diagnosis as we used to know it…

    Reply ↓
    1. Jerome Savage

      Last paragraph is humorous I think. Sounds like, if I can use a very simple but terrible analogy –
      Having a plain ordinary dipstick to test that oil levels are low. And to confirm it, using another nice shiny embellished dipstick that’s guaranteed to give the same answer – no ?

      Reply ↓
    2. Martin Back

      To me, a confirmatory test has to be a different sample, drawn by a different person,sent to a different laboratory, and tested for the presence of a different sequence of RNA. Otherwise there are too many possible procedural, contamination, or equipment errors that can be repeated.

      Reply ↓
  98. Roger Terry

    Dear Dr Kendrick

    You suggest that, if the false positive rate may be over 4%, as stated by the test vendor you quote, then the results stated by ONS and others are meaningless, and policy decisions based on these are wrong. This is true, even accepting that the folk presenting for tests are more likely to have the disease than the general population.

    However, is the false positive rate truly as high as that? The data suggests otherwise. Prof Heneghan’s article in August “COVID Cases in England aren’t rising: here’s why” (https://www.cebm.net/covid-19/covid-cases-in-england-arent-rising-heres-why/ ) states that the number of pillar 1 cases per 100,000 tests in July in England was around 200 – ie 0.2%.

    The latest COVID infection survey pilot found “In the latest six-week period, there were 291,732 swab tests, and a total of 419 positive tests”, ie 0.14%.

    Surely these results imply:
    – the proportion of false positives is less than 0.2% of tests
    – of the 6914 positive tests reported on 1st October (from c. 250K tests) at most 500 were false positives

    This suggests that the current apparent rise in cases is real and not an artifact.

    Roger Terry

    Reply ↓
    1. Roger Terry

      But I can see that this argument is wrong, or at least incomplete, as it neglects the degree of correlation between a positive test, an infection, infectiousness, and illness, which is Dr Kendrick’s underlying point. Whatever condition the covid test detects is truly increasing, but this may or may not indicate that citizens are at risk.

      Reply ↓
      1. KJE

        Or it always there and they are running more cycles of the test, so finding more instances of it? if there is no transparency about tests and number of cycles, primers used and so on, it’s impossible to know what is going on.

        Reply ↓
      2. David Bailey

        If the number of PCR cycles was increased surreptitiously, there would still be a degree of correlation because if the test would have been positive after 35 cycles (say), it would still be positive after 45 cycles, but many more false positives would be coming up as well.

        At present the situation is absurd – we do not even know whether the tests are running at a fixed sensitivity, or if this is being continually ramped up.

        In addition, I would not disregard the powerful nocebo effect (inverse placebo effect) on some people who receive what they may consider to be a death sentence. If someone with a positive COVID test dies shortly after, nobody is going to challenge the diagnosis.

        Reply ↓
    2. Martin Back

      pillar 1: swab testing in Public Health England (PHE) labs and NHS hospitals for those with a clinical need, and health and care workers
      pillar 2: swab testing for the wider population, as set out in government guidance
      https://www.gov.uk/government/publications/coronavirus-covid-19-testing-data-methodology/covid-19-testing-data-methodology-note

      You can expect people in hospitals to have a much higher rate of infection than the general population, so the Pillar 1 test will have a far lower proportion of false positives.

      Reply ↓
      1. Malcolm Ramsay

        “You can expect people in hospitals to have a much higher rate of infection than the general population, so the Pillar 1 test will have a far lower proportion of false positives.”

        Doesn’t that depend on why the false positive results are happening? If a significant proportion are a result of the test picking up other viruses, then the false positive rate might be higher in a symptomatic population than it is in the overall population.

        Reply ↓
    1. AhNotepad

      Snce they are stabbing people now with zero informed consent, the next step, (oops what am. thinking, they are already doing it) will be vaccinating on day of birth “to protect them”. This will be except for the Rothchild, Gates, Rockefeller, Tedros etc progeny who will be unselfish and forego their vaccine so they remain “at risk” to demonstrate their sacrifice for humanity.

      Reply ↓
    3. KJE

      So death vaccine for the over 65s (oh, dear, but inevitably some people have side effects) as there’s no point giving them contraception and anti-fertility for theh younger age groups. I still say “avoid”.

      Reply ↓
  101. Eric

    For those who’d rather discuss grammar:
    On Friday at 5:18 p.m, press secretary Kayleigh McEnany announced that the president was being moved to Walter Reed hospital. Here is her full statement:

    President Trump remains in good spirts, has mild symptoms, and has been working throughout the day. Out of an abundance of caution, and at the recommendation of his physician and medical experts, the President will be working from the presidential offices at Walter Reed for the next few days. President Trump appreciates the outpouring of support for both he and the First Lady.

    Reply ↓
    4. The Dim Appear

      President Trump did look at little pale in the pictures I saw of him after his positive covid diagnosis (assuming the test is accurate), not his normal color at all. I wish him a speedy recovery from whatever it is that ails him.

      The whole thing does look slightly stage managed by his team though, I mean is it really necessary for four hundred doctors to line up outside his hospital to give a press conference?

      Do you think that after this illness he will decide covid is really a threat and arrange global lockdown 2.0 with other world leaders decimating what remains of the economy? What if he’s been watching David Atenburgh’s mind numbing programming while in hospital and decides that global warming / cooling / climate change is man made and decides to act on that?

      Reply ↓
          1. Steve B

            What an extraordinary thing to say, unless it’s sarcasm If it’s not, then i don’t know whether to laugh or cry.
            I think it’s time for me to stop reading these comments until the subject matter gets back to cvd etc ,and sanity prevails

          2. AhNotepad

            Steve B, like trump or not (and I don’t), he has been the least war mongering president for a looooooong time. So David’s statement is somewhat reasonable, and not lacking sanity. You should look at the alternative possibility.

          3. KJE

            Actually, yes, that’s one of the few things I have noticed about the USA recentl: they haven’t declared war on anyone or interfered in the way others want to run their countries – got to be a plus.

      2. AhNotepad

        The Dm Appear, if trump has mild symptoms and recovers quickly, why not look at it as not a threat, and relax the oppressive restrictions. Keeping people in fear will damage their health, and many will succumb to trivial ailments, and even die if the fear is gret enough.

        I went to collect some screws and drills this morning, no mask. There was someone in front of me and when he turned round, he had no mask, nor did one of the shop staff (there were two staff). Meanwhile Doris was pronouncing earlier we could be bumping along to Christmas or beyond. All based on grossly unsound science with no consideration for the consequences.

        Meanwhile https://youtu.be/ektfBatKG3w a video about some of the serious damage being done by the muzzled sheep proponents.

        Reply ↓
  102. Rob

    I agree with the moral of the story, however this statement is incorrect:
    TN + FP = the total number of tests carried out.
    I think:
    TN+FP is the total number of well(free of disease) individuals only, so u can’t use 350,100 as total.
    350,100 = TN+FP+TP(true positive)
    And you don’t know the TP number
    Cheers, tx for fantastic info

    Reply ↓
    1. Martin Back

      You are correct in that TN + FP is the total actual negatives. The total tested however would be TN + FP + TP + FN. One equation with four unknowns. You need another three equations and you haven’t got them, so any calculation is a bit of a thumb-suck.

      In fact, looking at the table, calculating the PPA and NPA percentage to two decimal places based on just 45 test subjects is nonsensical. One false result and you are out two whole percent. You would need to test 10,000 known true positives and 10,000 known true negatives to justify that level of accuracy.

      Reply ↓
      1. Martin Back

        Oops, they calculate the PPA and NPA percentages to one decimal place i.e. three significant figures, therefore they need 1,000 each of true positives and true negatives as test subjects.

        Reply ↓
  103. Tish

    I suspect that contraceptive vaccines could be an important factor because something with some ‘moral’ overtones would, I would have thought, been necessary to get EVERYONE on board at meetings held to discuss the current awful homogeneous strategies. There is so much agreement (or else forced silence). Bribery and group membership behaviour may well have played their parts but at least some of the deliberators might have needed to be convinced by such talk as that of a desperate world population crisis. Developing countries may be particularly targeted.

    Reply ↓
    1. AhNotepad

      I wonder what Ursula von der Leyen would think of that vaccine. She has rather a lot of children. Yet another case of one rule for them, another rule for us,

      Reply ↓
  104. shirley3349

    For more news on Dr Rainer Fuellmich’s initiative to launch a class legal action in the USA in which anyone anywhere damaged economically or physically by the anti-Covid measures may join, he has today launched a video in English, a translation of one he gave a couple of days ago in German, explaining who he is and what his position is. It is called Crimes against Humanity, and was still available on YouTube when I checked a few moments ago.

    Reply ↓
      1. Studs Ramrod

        I’ve seen this all the way through and he names individuals and companies and it hasn’t been pulled from YouTube.
        What he claims is so incredible that I’ll need to see further clarification before I believe it. However he quotes the reputable scientists so he has done his homework.
        Sadly though, it could be the only thing that makes sense.
        But this would probably prove Godwin’s Law wrong.

        Reply ↓
  105. JB

    The article entitled “ Coronavirus: surprisingly big problems caused by small errors in testing” by Christian Yates illustrates the false-positive/false-negative problem quite well.

    Reply ↓
  106. JB

    I am quite concerned about National Public Radio (NPR) here in the states, hiring Dr. Jha at Harvard over and over again, who concludes with every new study that he does for NPR that we must do more and more testing to put an end to this pandemic. The latest figure is that “ideally, the U.S. would need just over 14 million tests a day.” (https://www.npr.org/sections/health-shots/2020/10/01/915793729/can-the-u-s-use-its-growing-supply-of-rapid-tests-to-stop-the-virus)

    It smacks of profiteering to me. I wonder if NPR and Harvard have huge stock positions in a testing company and are, together, trying to push up the stock price.

    Reply ↓
  107. Tish

    The wearing of masks seems to be increasing people’s fear. More people are now wearing them outdoors where they are not expected to. Some people hang back from or rush past me in shops because I’m not wearing one. But this morning I encountered a lady in a Scottish supermarket who, like me, was not muzzled, and she gave me a beautiful smile and made my day. So shouldn’t we who are maskless take any opportunity to smile at and so show approval to anyone shopping who is not in a mask (although they are not often encountered)? A bit of solidarity is cheering.

    Reply ↓
    1. KJE

      I’ve always thought the local lockdowns were punishments for areas with lots of Labour voters – although the Labour line seems to be just as bad as the govt if my prat of an MP is anything to go by.

      Reply ↓
  110. AhNotepad

    Some of this might explain the total chaos we see now. (Not that I trust the guardian).

    ”I have known Johnson since the 1980s, when I edited the Daily Telegraph and he was our flamboyant Brussels correspondent. I have argued for a decade that, while he is a brilliant entertainer who made a popular maître d’ for London as its mayor, he is unfit for national office, because it seems he cares for no interest save his own fame and gratification.”

    https://www.theguardian.com/commentisfree/2019/jun/24/boris-johnson-prime-minister-tory-party-britain

    Reply ↓
  113. Bill M

    I’m sure it’s been covered in the 3 or 400 replies but surely you mean that PCR is best used for screening – might you have it – than diagnosis – do you have it, which would require a more precise test.

    It would also seem, given the actual results quoted, that the error rate can’t be 4%. If there’s an inherent false positive rate, then given any true infection rate in the population, the number of false positives clearly can’t exceed actual positives. Of course if we knew what the actual Infection rate was, then we could easily back solve the number of false positives. But since this question is clearly being avoided by government, we’ll never know.

    It should, however, be possible to get a sense as to whether false positives are driving the numbers by comparing the geographical distribution of positives vs wave 1. We’d expect – given how PCR “works” – false positives to be related to residual viral material In the subject. If so, there will be more residual viral material in areas that were more infected in wave 1, so more false positives in subsequent tests and ergo higher apparent infection. Precisely the opposite of what we would expect if were a fresh wave hitting the country.

    Reply ↓
    1. Martin Back

      The test manufacturer would say that if you get a positive result and there was Covid viral material in the sample, it’s not a false positive, even if that viral material was left over from an infection that has passed. But we would classify it as a false positive because the person was not actually infected at the time of the test.

      So really, you need two positive classifications:
      PIP = Post Infection Positive
      CIP = Current Infection Positive

      The problem is, there’s no single test that can distinguish between the two positives.

      Reply ↓
  114. Jerome Savage

    The irish parliamentary review group the “Dáil Covid Committee” that asked pertinent questions off the TMB’s (top medical bods) just last week and uncovered policy that would have a car crash victim labelled a CV19 casualty of testing positive, has been disbanded. It also uncovered data in relation to false test results but after many attempts, failed to get confirmation of the number of cycles involved in the testing process.

    Reply ↓
  115. shirley3349

    Rainer Fuellmich is not a medical doctor but a lawyer as Leifur stated. In Germany, someone of his professional status would usually have a doctorate and would use this title in his professional life, and anywhere his qualifications might become an issue. A medical doctor is ein Arzt, a lawyer, ein Jurist, which makes the distinction clear. The normal, formal mode of address is still, to my knowledge, Herr Doktor /Frau Doktorin in all professions.

    Reply ↓
    1. shirley3349

      Like many others, I’m afraid I have spelt Dr Fuellmich’s name wrong. It i s Reiner not Rainer. Both are pronounced the same, rhyming with climber.
      Rainer is the more usual, like Rilke, the poet, and Fassbinder, the film director. I have never come across a Reiner before, though I have met several Rainers, including the son of a old Viennese friend.
      She taught me much of my German, when we were sunbathing on the flood plain behind the north-eastern bank of the Danube in the summer of 1964, long before they dug a further canal to carry the flood waters.
      The river w a s blue on occasions. It also had an amazing current, even in August, well after the main Alpine thaw. To cool off, we used to swim in the river, never venturing more than about 10 feet from the bank. The water was very cold. After a few seconds we would swim down stream, letting the current carry us leftwards into the bank. We then had to walk several hundred yards back to our starting point.
      Those were the days!

      Reply ↓
  116. Jennifer

    So Trump has just tweeted that he feels better than he did 20 years ago, and will be leaving hospital in the next few hours.
    I hope his medics are taking into account the side effects of large doses of steroids may have on the human body and brain.
    Euphoria being the one that worries me most. And he has his finger on the button? Covid 19…….Bah! The least of our worries.

    Reply ↓
  117. Norman

    I am about to read this. It sounds a positive move
    https://unherd.com/2020/10/covid-experts-there-is-another-way/

    Anecdotally, I think that there could be some ‘risk compensation’ for mask wearing, i.e. people starting to take less care over distancing. I’ve noticed this in my last two visits to the supermarket.

    It would be hard to do a good scientific study of this. As Prof Carl Heneghan said, we don’t even seem to have a good study of the efficacy of masks in normal everyday life. But maybe the government should pay for one, as it thinks that we must be muzzled (unlike a) the Scandinavian countries, which have few masks in use or b) Japan, where masks are voluntary).

    By the way, I’m told that if a non-mask wearer is shouted at on entering a small shop, as I have been once, then this may be discrimination under the Equality Act 2010. There is it appears no legal requirement to give a shop owner or staff a reason for not wearing one. If one is pestered and asked for more information, this may breach one’s privacy and the right, e.g. to keep one’s medical data private.

    I’m only a ‘mere’ scientist. So if there are any lawyers on here, is it a potential breach of the Act? Breaking the Act led to substantial damages in a case involving disability and housing law
    https://thenegotiator.co.uk/midlands-estate-agency-faces-paying-damages-following-no-dss-court-case/

    Reply ↓
    1. AhNotepad

      Norman, the human race has lived for all its time with viruses. There is no way of avoiding them. It is the responsibility of people to look after their immune systems, no one will do it for them, except for young children where their mother comes in handy. Muzzles and distance do nothing. If you think they do, explain how you are “safe” with 2m when sand from the Sahara gets deposited on your car, in the UK. Look up the study done by Surgeon Neil Orr at Severalls hospital, Colchester, where patients fared better after operations where the medical staff DID NOT WEAR MASKS.

      Non-pharmaceutical methods do not reduce risks, they are there to engender fear and force compliance. This is a psychological operation, and the aim is command and control.

      Reply ↓
    2. KJE

      The government itself says in the section under Exemption cards
      “Those who have an age, health or disability reason for not wearing a face covering should not be routinely asked to give any written evidence of this, this includes exemption cards. No person needs to seek advice or request a letter from a medical professional about their reason for not wearing a face covering.” The NW Equalities organization, whose name I forget, says that insisting on masks for those who are exempt might be classed as a hate crime under the Equalities Act 2010. I’d certainly cite it if challenged.

      Reply ↓
      1. shirley3349

        I go out very little as my mobility is poor. If asked politely by the man with the trolleys at our local supermarket, why I am not wearing a mask, I say truthfully, “Heart condition”, whereupon the look on his face suggests he expects me to drop dead at once!
        Never had any problems either with officialdom or the general public. Maybe I look far worse than I feel?

        Reply ↓
        1. KJE

          No one has ever asked me, so I must look rough! I have emphysema and heart arrhythmia. I get out as much as I can, though.

          Reply ↓
  118. elizabethhart

    Here’s a link to a recent paper aiming to justify actions taken in Australia…

    Note this paper cites the infamous Ferguson Imperial College report (reference no. 18)

    Coronavirus Disease Model to Inform Transmission Reducing Measures and Health System Preparedness, Australia

    https://wwwnc.cdc.gov/eid/article/26/12/20-2530_article?ACSTrackingID=USCDC_333-DM39150&ACSTrackingLabel=Latest%20Expedited%20Articles%20-%20Emerging%20Infectious%20Diseases%20Journal%20-%20September%2028%2C%202020&deliveryName=USCDC_333-DM39150

    Reply ↓
    1. AhNotepad

      They mght oppose it if they get to hear of the cases of transverse myelitis (aka polio) from the trial vaccines.

      Reply ↓
    2. Mike S

      This jumped out at me:

      “Lessons from past vaccination campaigns suggest that a generous compensation program for people who have serious vaccine side effects should be a centerpiece of these efforts. ”

      Let the fraud begin…

      Reply ↓
  121. Ken Scott

    Dear Dr Kendrick

    Thank you for your updates.

    Is this a direct quote from your source?

    Politicians and Health Officials are basing their numbers of cases entirely on the results of these tests, which are not fit for this purpose.

    They are then using these figures to terrorise the population, and to justify decisions to impose local lockdowns, and increase nonsensical general restrictions which are having a massive impact on people’s lives and their health, and also on the economy, particularly hitting small businesses hard.

    It seems to me there is a grand case of medical larceny being inflicted on the people of UK (and other countries) by governments and their scientific advisors.

    Why are the likes of you, Mike Yeardon, Ivor Cummins, Sunetra Gupta and others not being heard by the powers that be?

    I am confused and very concerned because Nicola Sturgeon seems on the brink of inflcting more restrictions.

    Keep the great work going Dr Kendrick.

    Ken Scott

    Ken Scott (MPRCA), Founder, ScottAsia Communications – UK & Thailand http://www.scottasia.net 
 UK — Email: kens@scottasia.net Office tel: (+44) [0]141 632 8521 Mobile: (+44) [0]7949 077959 Skype: ken.scott.asia
 Thailand – Anchalee Sriwongsa (Ann) Mobile: (+66) 91 806 0484 E-mail: ann@scottasia.net Skype: anchalee.sriwongsa

    ScottAsia Communications – Delivering Great PR for Travel Industry Brands

    >

    Reply ↓
  123. elizabethhart

    A recent BMJ rapid response by Janet Menage, retired GP:
    https://www.bmj.com/content/371/bmj.m3925/rr-0

    Dear Editor

    “The best decisions are based on the best science”, the article quotes.

    However, the CDC states on page 39 of its 13th July 2020 document entitled,’ CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-PCR Diagnostic Panel For Emergency Use Only Instructions for Use’ (1) :

    “Since no quantified virus isolates of the 2019-nCoV are currently available, assays designed for detection of the 2019-nCoV RNA were tested with characterized stocks of in vitro transcribed full length RNA (N gene; GenBank accession: MN908947.2) of known titer (RNA copies/μL) spiked into a diluent consisting of a suspension of human A549 cells and viral transport medium (VTM) to mimic clinical specimen”.

    What does this mean? And is it the “best science”?

    Reply ↓
  124. elizabethhart

    Dr Mike Yeadon, former chief scientific advisor, Pfizer
    ‘Government are using a Covid-19 test with undeclared false positive rates’
    17 September 2020

    (Includes some adverts)

    Reply ↓
    1. elizabethhart

      Also see this article by Dr Clare Craig:
      COVID-19 and the false positive trap:
      https://spectator.us/covid-19-false-positive-trap-seasonal/#

      Imagine a world where COVID-19 has been eliminated. To be certain this is true, the government conducts regular tests at random. The number of positive results should be zero, right? Wrong. There will always be a proportion of cases tested that come back with a false positive test result. Thankfully, for COVID-19, the false positive rate is less than one percent of tests done. But it is not zero. It will be impossible for us to ever reach zero. Why? Because COVID-19 cannot be eliminated, even if it is likely to evolve to be more benign and become a seasonal problem like influenza.

      Reply ↓
  126. Eric

    Brand new and not yet peer reviewed (and from my alma mater, too):

    Click to access 2009.05732.pdf

    This is from a group of mostly theoretical physicists who usually model complex materials.

    Just by breaking down infection rates by age group and using age dependent infection fatality rates, they were able to model deaths in Germany. Their key findings:
    – actual fatalities match prediction very well
    – there is no indication that detected infections are harmless now
    – nor that a major percentage of infections is currently not being detected
    – elder cohorts were well protected into September which is when higher infection rates in younger cohorts started spilling over
    – if trend continues, will see 200-300 deaths / week in Germany by end of October, i.e. half of what it was beginning of May

    Reply ↓
  127. Harry Ludder

    Hi Malcolm

    Very interesting, as usual.
    I have a friend who works at one of the testing labs. He asserts their PCR test is flawlessly brilliant of course.
    My query is, that if there are so many false positives, why is the at this stage is there not a swathe of people with a second positive result.
    Whether you have Covid or not, at 1% false positive rate, the likelihood of 2 positive tests is 1 in 10,000. At a rough figure of 20 million tests carried out in the UK, we should have about 2,000 people flailing their arms in the air, hysterically screaming they have it again and they have no immunity!!!
    Where are all these false repeat victims?

    Reply ↓
    1. Dr. Malcolm Kendrick Post author

      Well, I have seen patients who tested positive, then negative, then positive again. In one case we got negative, positive, negative, negative, positive, negative, positive. He came through my unit twice, then ended up in a care home that I visit. One of the care home staff tested positive for eight weeks – with no symptoms. We have had many who tested negative when ill, ended up in hospital, then tested positive. We have had patients test negative in hospital, came in to us, then tested positive. We have had patients with clear symptoms, indicative of COVID19 who tested negative, then died. At present, we have many patients who have tested positive, who have no signs or symptoms at all.

      From time to time I have even seen patients get ill, test positive for COVID19, then go to hospital and die, of COVID19. These, I must say, have been a minority.

      Reply ↓
  128. Alisina Z

    Dear Dr. Kendrick,

    As I couldn’t figure out the answer myself in the last couple of months, there is no other choice than bother you with this question.

    To give you some back ground. I’m de Dutch family doctor/GP and have my own practice nearby Rotterdam. I have read all your books (twice) and the blogs (since last 2 years). Your book Doctoring Data I have given to open minded colleagues as a gift for several occasions.

    Now my question regarding to this part: First, you have to know the estimated prevalence of the disease in the community. That is, the total number currently infected. Last time I looked it was one in nine hundred. For the sake of this calculation I shall call it one in a thousand. [Or, to put it another way, sixty-seven thousand people in the UK (population 67 million) are currently infected with COVID19].

    If the PCR-test is not that accurate, how can the estimated prevalence be any correct? With other words, how can one measure the prevalence if this measurement goes through the same inaccurate PCR test?

    Thank you

    Alisina

    ________________________________ Van: Dr. Malcolm Kendrick Verzonden: maandag 28 september 2020 16:04 Aan: booalisina@hotmail.com Onderwerp: [New post] False positive tests

    Dr. Malcolm Kendrick posted: “28th September 2020 There has been a lot of noise about false positive COVID19 tests in the news. So, I thought I would try to explain what it all means. Or do my best anyway. There are two measures in most medical screening tests which are usually “

    Reply ↓
  129. Mike Roberts

    I don’t follow the calculation. You don’t know the TN figure but have used it in the divisor of your calculation, assuming that TN+FP is the total number of tests. In doing this, you are assuming that none of the positive results are true positives. You’ve then used these assumptions in trying to calculate the TN figure. From the information given, it isn’t possible to do that calculation, as far as I can tell. I’m a little suprised that you don’t see the calculation of 15,404 as the number of false positives that you would expect. Given that the number of positives was way less that that, where is your explanation for why you didn’t see at least 15,000 positive results (or some number in that vicinity)? Your estimate of a 1 in 1000 prevalence is also unlikely to be correct, or even close to correct, at every point in the case trajectory (e.g. for those countries that stamped out the disease for periods of time, one would expect that prevalence to be close to zero whilst other countries with overwhelmed health systems may have a prevalence of much more than 1 in 1000).

    Reply ↓
  130. N

    You lost me when u called the Cepheid xpert test an antigen test.

    The Cepheid Xpert Xpress SARS-CoV-2 RT-PCR test is the “Gold Standard” COVID-19 antigen test used in our laboratory.

    Reply ↓

Leave a comment