Monday, 31 January 2022

The Virus Stats that Cost Everyone a Lot

By Martin Cohen
This is a post about statistics. Not that I’m actually a great mathematician, let alone a statistician, but I do at least appreciate that the power of numbers to influence debates. And a debate in particular where statistics have been thrown around since the beginning of the Covid one. 

So, travel with me back nearly two years to the origins, and take a second look at some key metrics that have been tossed about ever since. One problematic measure has been the “Case Fatality Rate”. This was officially put by the United Nations at just under 1%, making Covid a very deadly virus.

The bit we can agree on is the definition. The CFR is the number of deaths from Covid divided by the number of “confirmed cases”.

The problem is that deciding who actually died “from Covid” is very murky. A typical report is that 95% of people dying from Covid have other co-morbidities. This means that they may actually have died from these rather than from Covid. The issue is exacerbated when you see that the typical age of someone dying “from Covid” is pretty much the age at which anyone dies.

So the numerator part of this crucial figure is HIGHLY debatable - and the denominator part, the number of cases is too. For starters, one problem is that what you, me and Joe Public understand as “a confirmed case” is someone who has symptoms and goes to hospital and is tested and found to have the virus. That would all make sense. But in fact, a case is simply someone who has the virus. And again, it is agreed that the great majority of people who encounter the virus never have any symptoms. These people are often not counted. This is why the number of cases a country has depends essentially on how much testing the government chooses to do.

To make matters worse, it depends on the criteria used for the test. The benchmark test, the so called PCR (polymerase chain reaction ) test, considered “the gold standard” for detecting Covid. The test amplifies genetic matter from the virus in cycles; the more cycles used, the greater the amount of virus, or viral load, found in the sample. Crucial to the test, then, is how many cycles are used -and that, perhaps surprisingly, is not a medical decision but a political one. In Europe, for example, The European Centre for Disease Prevention and Control does not recommend a specific maximum amplification cycle threshold for PCR tests. However, it does recommend that if the values are high, e.g. > 35, “repeated testing should be considered”. In other words, it recognises the results are unsafe.

Yet that decision on the number of cycles is not even communicated when a “positive test” is returned. As Angela Rasmussen, a virologist at Columbia University in New York told the New York Times, “It’s just kind of mind-blowing to me that people are not recording the C.T. values from all these tests — that they’re just returning a positive or a negative.”

Ultimately, there is no standard cycle threshold value that is agreed upon internationally. The U.S. Food and Drug Administration currently gives laboratory manufacturers autonomy in determining how many cycles are needed to determine whether a sample is positive or negative.

How accurate the test is matters, because everyone admitted to hospital, for whatever reason - maybe they had a heart attack - is routinely tested for Covid. If they are considered positive, and later on die, they will be counted as a Covid fatality, as “dying within 21 days of a positive test”.

So that’s three rather big question-marks lurking in the Covid data. But the next one, I think, is worse. This statistic purports to show vaccines save people from the worst effects of the disease.

It’s the statistic that led the CDC in the US to say:
“COVID-19 vaccines are highly effective at preventing severe COVID-19 and death.”
And you can read that in all the papers, in all the fact-checkers, and so you “might” think it must be true. However, statisticians at Queen Mary College in London, looked at the UK data (which is representative of other countries too) and concluded:
“Official mortality data for England suggest systematic miscategorisation of vaccine status and uncertain effectiveness of C19 vaccination”
They noticed that the official statistics showed that, following vaccination, there was a sudden surge in the numbers of UNVACCINATED people dying. The so-called ‘healthy vaccine’ effect. A less cheery explanation was that vaccines might actually be killing people - but if the deaths occurred within 21 days, as most side-effects do - being classified as deaths of “unvaccinated”.

This is a possibility, and adverse effects databases like the European EudraVigilance database and US VAERs ones currently report alarmingly high numbers, in apparently compelling detail – however the miscategorisation does not need to mean that vaccines are killing a lot of elderly people. Rather, fragile people are prioritised for vaccination, and thus skew the figures. However, by grouping vulnerable people together statistically to be vaxed and then … calling this group the unvaccinated, the authorities have very conveniently created an apparently miraculous positive effect for vaccines. That it is not really there is indicated that the positive effect – “vaccines save lives” - is not only for Covid but for ALL CAUSE mortality!

This is known. Yet far from accepting the statistics mislead, governments and drug companies surmise that the treatments may have unexpected general positive effects.

In reality, the statistical anomaly is large because in countries like the UK, the NHS Guidelines explicitly state that the most critically ill people are the ones who must be prioritised for vaccination in each age group.

Let me try to sum it all up in three sentences! Vaccine data shows most of the advantage from the jab in the first few months. Because Covid vaccination programs prioritise very ill people, a significant number of whom die in the following 21 days - not from the vax necessarily, just because they were, well, vulnerable. Whatever the reason, again under the official guidelines, these deaths are classed as ‘unvaccinated’, creating the ‘bad news’ for the unvaccinated and the amazing, parallel, health boost for those who are.

So there you have it. Some examples of how duff statistics alone, not anything more secretive let alone worrying, could have created a Ten Trillion Dollar “pandemic” that maybe never was. Worse still, they could have led to policies that really have been killing people.

3 comments:

Thomas O. Scarborough said...

There is no doubt that the policies have been killing people. I think, too, to borrow from John Ruskin, that we do not know ‘other losses and gains, far away among the dark streets’.

Yet I think that what is needed is differentiation of excess deaths in particular. Until we have that, it will be easy to cast doubt on all kinds of statistics, including the above.

Thomas O. Scarborough said...

An interesting meta-analysis of published last month by Johns Hopkins. Epictetus wrote, 'Happiness and freedom begin with a clear understanding of one principle: some things are within your control, and some things are not.' https://sites.krieger.jhu.edu/iae/files/2022/01/A-Literature-Review-and-Meta-Analysis-of-the-Effects-of-Lockdowns-on-COVID-19-Mortality.pdf

Keith said...

Where, Martin, I have witnessed considerable whipsawing over the last two years is in officialdom trying anxiously to align public policy and healthcare with the data and science. Shifting quicksands. That being said, I have cut plenty of slack for everyone in the unenviable position of shouldering the shaping of policy decisions. Key players, on the front lines of this exhausting public-health marathon, have had plenty of opprobrium heaped on them for miscues. Perhaps unfairly, given the learn-as-we-go-and-then-quickly-pivot circumstances that the scientists, policymakers, and organizational spokespeople have had to ride in the face of a *novel* coronavirus that has accounted for considerable illness.

Post a Comment

Recent Comments