Wednesday, 3 February 2021

COVID-19 mythbusters

Once the number of COVID-19 deaths in Britain reached over 1,000 a day, I naively thought that the people promoting the "casedemic" theory would shut up and go away. Instead, they have resorted to increasingly desperate and tenuous arguments to maintain their belief that there is nothing particularly unusual or threatening about this virus. 

Because the same dumb objections are raised repeatedly on Twitter, I started writing up prepared responses. Some people have told me they find them useful, so I am posting the whole lot below, with links to the sources. 

A few of them are, fortunately, getting past their sell by date because the denialists have moved on to new turf. At this stage in the second wave, pretty much all they have left is arguing against germ theory. 


1. Flu has not disappeared, but there are fewer flu deaths than average. In England in December, the age standardised mortality rate for influenza and pneumonia (which are grouped together because influenza usually kills through pneumonia) was 25.5 per 100,000, significantly lower than the five year average of 50.9 per 100,000 (Office for National Statistics). It was still the ninth biggest killer in that month, however. Flu has not gone away.

There are fewer flu deaths than average because hand-washing, social distancing, face masks, lockdowns, reduced air travel, etc. cut its transmission. New Zealand also saw a large drop in flu deaths, but has only had 25 COVID-19 deaths.

The idea that COVID-19 deaths are actually misdiagnosed flu deaths is absurd. Firstly, it would require Britain to have a massive flu season with an extraordinarily lethal strain of influenza at a time when the population has been adopting measures  since March that cut the rate of transmission.

Secondly, it requires most people who die of ‘flu’ to have tested positive for COVID-19 recently. Since ‘only’ 2% of the population has COVID-19 at the moment, that would be quite the coincidence.

Thirdly, it requires a conspiracy among doctors to put COVID-19 on tens of thousands of death certificates of people who actually died of ‘flu’. This conspiracy extends to all medics who treat COVID-19 patients and can see that the symptoms and fatality rates are different (and worse).

The more obvious explanation is that the virus which killed lots of people in the spring is killing lots of people in the winter.

2. Where are the excess deaths? They can be seen every week in the Office for National Statistics publication ‘Deaths registered weekly in England and Wales’ and in the monthly publication ‘Monthly mortality analysis, England and Wales’.

In December, the ONS recorded 10,594 excess deaths and 10,973 deaths from COVID-19. In the first three weeks of January alone, there were 17,342 excess deaths.

Excess deaths are deaths above the five-year average. That is the definition. Admitting that there are excess deaths while insisting that there are no more excess deaths than usual is a meaningless oxymoron.

Cumulatively, there will be more COVID-19 deaths in the second wave. There were 41,000 deaths with COVID-19 on the death certificate between March and July, and 60,000 since the start of November (see section 4 for the 'with Covid or of Covid?' question). The number of excess deaths, which has been affected by a lower-than-usual number of non-Covid deaths in the second wave, is also likely to be larger. As of 22 January 2021, there had been over 41,000 excess deaths since November.


3. False positives are not an issue. The false positivity rate of the PCR test is very low (well below 0.1%). We know this because when countries with hardly any COVID-19 cases do mass testing, they find hardly any positives. For example, in New South Wales, 135,000 people were recently tested, producing 93 positives. Even if they were all false positives (they weren’t), that would be a false positive rate of 0.069%.

The ONS does a random prevalence survey using the PCR test every week. It says the specificity of the test is at least 99.9% (ie. the false positive rate is no more than 0.1%). Between 31 July and 10 September, 159 of the 208,730 total samples tested positive. As the ONS says: “Even if all these positives were false, specificity would still be 99.92%” (ie. the FPR would no higher than 0.08%). 

The increase in positive tests cannot be attributed to the increase in testing. The number of tests has increased two or threefold since August while the number of positives has increased fifty-fold.

The idea that the PCR test has a 90% or 93% or 95% false positive rate came about because Michael Yeadon wrote an article in September arguing that if we tested people at random with a test that had a 1% FPR and there is very little SARS-Cov-2 in the community then most of the positives will be false (he said 90%). This is the base rate fallacy and it is not wrong. It is just irrelevant. We do not have low levels of the virus in the community, we are not testing at random and the FPR is much lower than 1%.     

Public Health England wrongly claimed that only 7% of COVID-19 cases would be identified through testing at airports. This led some people who don’t understand science or statistics to assume that the FPR is 93%. PHE was actually talking about the number of people with the virus who would become symptomatic during the course of a flight. Even if the PHE claim related to testing accuracy, it would be a 93% false negative rate, but it wasn't. 

False negatives are far more common, partly because the PCR is so invasive that people don’t do it properly, but the ‘sceptics’ are strangely uninterested in those.

4. With Covid or of Covid? In the UK, COVID-19 deaths are recorded in two ways. One method is based on COVID-19 being mentioned on the death certificate. The other counts death as Covid-related if they took place within 28 days of a positive Covid test.

This means that it is possible for someone to be admitted to hospital, test positive and die from another disease. It is also possible for someone to be admitted with another disease and catch Covid while in hospital. Indeed, this happens too much; the Telegraph reported in December that 16% of hospital patients with Covid caught the virus in hospital.

Those who seek to downplay the epidemic focus on these patients. Of course, Covid is still a problem if you catch it in hospital, and the 28 day cut-off for deaths means that people who die of Covid after 28 days are not included in the figures, but the ‘sceptics’ aren’t worried about under-reporting.

What we can say for sure is that the number of excess deaths in 2020 closely matches the number of Covid deaths. If you want to claim that a large number of ‘with Covid’ deaths were not from Covid, what were they from? What other disease has surged in the last twelve months? Why do cases rise a week or so before hospitalisations rise if the patients are testing positive in hospital?

As of January 23rd, less than two per cent of the population have COVID-19 and yet around 40 per cent of people dying have COVID-19. We know that COVID-19 is a lethal disease. Occam’s Razor says that COVID-19 is killing the vast majority of the people who die with COVID-19.

This is confirmed by the death certificate figures which the ‘sceptics’ rarely mention. These figures are very similar to the figures based on the 28 day cut-off. As of February 2nd, 108,013 deaths have been recorded based on the 28 day rule and 112,660 have been recorded based on death certificates.

Contrary to popular belief, it takes more than having COVID-19 mentioned somewhere on the death certificate for a death to be classed as a COVID-19 death. The ONS distinguishes between deaths involving Covid and deaths in which Covid was the underlying cause: “In most cases (90.2% in England and 88.2% in Wales) where the coronavirus (COVID-19) was mentioned on the death certificate, it was found to be the underlying cause of death.”

5. No, they are not ‘lockdown deaths’ 

Now that the Covid ‘sceptics’ have mostly given up pretending that there are no excess deaths, they have pivoted to claiming that the excess deaths are not caused by the viral pandemic that has killed millions worldwide, but by the effects of lockdown itself.

It is not clear what diseases are supposed to be causing these ‘lockdown deaths’. One claim is that the NHS has become the ‘Covid Health Service’ and people with other ailments are not being treated. It is certainly true that hospitals overrun with Covid patients have had to cancel elective surgery and some routine screenings and operations. This is likely to cause deaths in the long term, if not immediately, but this is the result of COVID-19, not the lockdowns. Allowing more people to be infected and hospitalised with the disease is unlikely to improve matters.

In nine out of the ten last weeks, there have been fewer non-Covid deaths than the five-year average. COVID-19 deaths comfortably account for all the excess deaths recorded this winter. This should not be too surprising. New Zealand and Australia both saw fewer deaths than average in 2020 despite extensive and draconian lockdowns. Lockdowns reduce the ability of other infectious diseases to spread and reduce some other causes of mortality, eg. road accidents.

There have been excess deaths in hospitals, care homes and private homes in recent weeks. An excess death is a death that would not be expected to occur, all other things being equal. You cannot infer that an excess of deaths in the home would not have occurred in any case. People are more reluctant than usual to go to hospital this year, quite understandably, but that doesn’t mean the people who die at home died from lockdown. There were a similar number of excess home deaths in the summer when there was no lockdown. Deaths that would have occurred anyway are simply being displaced to the home. The only way you could convince yourself otherwise is by believing that all the people who perished from a ‘lockdown death’ also happened to test positive for COVID and/or had a doctor who committed the criminal act of putting the wrong cause of death on the death certificate.

If the excess deaths were lockdown deaths, they would take place at a similar rate across the whole country. Instead, there are more excess deaths in regions with high rates of COVID-19. That is because they are not caused by lockdown, but by COVID-19.

6. The WHO has not changed its advice to clinicians about testing for SARS-CoV-2, nor has it said that the PCR test produces a lot of false positives.

On 20 January 2021, a theory circulated on the internet that the World Health Organisation had begun advising medics to retest anyone who tested positive for SARS-CoV-2 if they don’t have symptoms. This was taken as a tacit admission that tests for the virus - and the PCR test in particular - produce a large number of false positives which result in an inflated case count and the NHS being under-staffed because workers have to self-isolate after falsely testing positive. One version of the theory claims that the WHO published the new advice on the day of Joe Biden’s inauguration in order to reduce the number of reported cases under his presidency.

The WHO webpage is titled ‘WHO Information Notice for IVD Users 2020/05’ and can be summarised as “Read the Manual”. Its stated purpose is to “clarify information previously provided by WHO”. It provides a very brief summary of the WHO’s advice published in Diagnostic Testing For SARS-CoV-2, published on 11 September. The webpage was first published on 13 January, not 20 January, and it only reiterates existing advice. It includes the lines: “as disease prevalence decreases, the risk of false positive increases. This means that the probability that a person who has a positive result (SARS-CoV-2 detected) is truly infected with SARS-CoV-2 decreases as prevalence decreases, irrespective of the claimed specificity” and “Where test results do not correspond with the clinical presentation, a new specimen should be taken and retested”

The first sentence is an almost verbatim copy of what the WHO said in the previous version of the advice on 14 December (now deleted but available on Wayback Machine). The second reflects what the WHO says in the Diagnostic Testing For SARS-CoV-2: “When test results turn out to be invalid or questionable, the patient should be resampled and retested.”

NHS staff are not told to self-isolate if they test positive once. On the contrary, they are told to “arrange a confirmatory PCR test via the established testing routes”. Staff who contract COVID-19 are told not to resume regular testing for 90 days in order to avoid false positives resulting from fragments of inactive virus. The same 90 day advice is given to the general public.  

Retesting people who test positive without presenting symptoms is perfectly sensible, but since the specificity of the PCR test is above 99.9%, there are a very few false positives to uncover and it makes little difference to the overall case count, let alone the number of people dying from COVID-19. As the advice is not new, it obviously has nothing to do with Joe Biden’s inauguration.

7. This is much worse than 2017/18
There were 50,000 excess winter deaths in 2017/18, leading some people to claim that the UK’s second wave of COVID-19 is no worse than a bad flu season. An article in the Telegraph by Sarah Knapton made this argument on 18 January 2021. She claimed that there had been 12,100 excess deaths in the second wave whereas “in a bad winter flu season, around 22,000 excess deaths would be expected.”

Unfortunately, the Telegraph article is based on a misunderstanding. 2017/18 was certainly a bad flu season. England and Wales saw more excess winter deaths (50,100) than any year since 1975/76. But even in an average year, there are around 30,000 excess winter deaths. There are always excess winter deaths because they are defined as the difference between the number of deaths that occur in winter (December-March) and the average number of deaths that occur in the preceding August to November and the following April to July.

That is totally different from the excess death figures published each week by the ONS, which are deaths above the five-year average for the time of year. For example, in the week ending 15 January 2021, there were 4,220 more deaths than have occurred on average in the last five years in the second week of the year, and thus 4,220 excess deaths. This was the 14th week in a row with excess deaths. This is not normal, by definition.

If you look at the total number of deaths in England and Wales, you can see that 2018 saw more deaths than usual thanks to the bad flu season, but it was nothing like 2020:

2016: 525,048
2017: 533,253
2018: 541,589
2019: 530,841
2020: 607,173

Since the start of November, there have been more than 40,000 excess deaths and this total will continue growing for at least several weeks. We won’t know how many ‘excess winter deaths’ there are until the figures are calculated in the summer, and the figures may be skewed by the COVID-19 deaths in the autumn, but they will very likely set a new record. In any case, you cannot compare the excess deaths being recorded now with the excess winter deaths of previous years. Although the names are similar, they are two totally different measures.

For more on the Covid charlatans, lunatics and grifters who throw misinformation at gullible people, see this Full Fact article, my Quillette article and the Anti-Virus FAQ.

No comments: