About 1,500 Days and Counting . . . (Part III)

Welcome to part III of this series about my assessment of COCK (Crazily Overexaggerated Covid Kookiness). It’s my celebration of the 4-year anniversary of the start of these wonderful times. In part II the issue of testing was the focus, but I didn’t quite cover everything I wanted to say before Substack informed me that the email electrons would go on strike if I tried to post an article of that length.

Driving the insane testing program was this notion that there were a whole bunch of otherwise perfectly healthy people who were, nevertheless, infected with The Cough of Calamity™ and spewing out deadly danger with every breath they took.

The Science™ agreed with them according some media articles

I wasn’t able to properly test this theory because, unfortunately, the phrase “what a load of fucking bollocks” is 6 words and not 4.

This is representative of the “quality” of The Science™ we were lavishly treated to.

These kind of ridiculous articles, like the most unwelcome of sharts, kept surprising us all in both the media and The Science™ journals.

These scrapings from the accidental underpants of academic publishing were rushed through. More sensible and measured works that went against the narrative grain were treated to a kind of academic Loperamide and publication was stymied.

This rather unpleasant deluge of demented drivel was The Science™ we were supposed to “follow”.

In part II of this series I looked at the mathematics that underpins medical testing. It’s well-known, basic, probability stuff that every Doctor knows, unless they’ve been to one of the ‘top’ colleges where they could replace all that difficult technical stuff with courses on the social justice of interpretive Lego, or similar.

What we saw was that medical testing is underpinned by ideas from probability. The definitions used for the accuracy and reliability of medical tests are probability questions.

We saw that we needed the idea of conditional probability to properly understand these definitions.

The relevant one today is how we approach the issue of a false positive. This became something of a hot-button issue during the COCK era. But what is it and how did we cock it up so badly?

---

Your False Positive is Positively False

The false positive rate is defined by the specificity of a medical test. If SP is the specificity¹ then the Test False Positive rate is

TFP = 1 - SP

This is, as we shall see, itself a probability.

This is a false positive at a ‘test’ level. What we saw in the previous article is that the false positive rate at a population level, PFP, can be very different, especially during times of low disease prevalence (and when you, stupidly, test everyone, even the asymptomatic folk - otherwise known as healthy people).

It is far from impossible to have, for example, TFP = 0.01 (which represents a specificity of 99%) but find that the PFP is, say, 1/4.

What this means is that 75% of the positive results obtained by testing the population are arising from false positives at the test level (which only occur 1% of the time). This is the significant effect that low prevalence can have.

But there’s a problem here. To badly misquote Spock from Star Trek, “It’s a false positive, Jim, but not as we know it”.

Let’s remind ourselves of what the specificity means.

You take a population and (at least conceptually) split them into two ‘rooms’. In one room go all the people who definitely have the disease. You ignore this room from now on.

In the other room go the rest. These are people who definitely do not have the disease.

We test the people in the non-disease room. The specificity is defined to be

The probability the test result is negative given the person tested does not have the disease.

In math terms we might write this as P(T=0 | D=0). This is just the squiggle version of the words in the definition.

A perfect test (in terms of specificity) would not yield any positive result in this non-disease population.

The problem is that there’s a ruddy great big elephant in the room about to drop a steaming great fresh pile of trouble to brighten up our day.

What we’re actually doing here is mixing up our probability questions.

The test is not distinguishing between disease vs no disease. It is testing for the presence of specific gene sequences.

A positive test result (and not a false positive) does not mean one has an active infection. More pertinently, it does not tell us about whether there is a transmissible infection, which is the salient point.

We can better think of the much-touted “gold standard” PCR test as a test for exposure to the virus in some form.

We’re exposed to viruses all the time. In every breath we take. Most of these will not be capable of causing a subsequent infection. Even those that do have this capability are usually dealt with very quickly by the body’s defences.

If we ran a PCR test at high enough thresholds we could probably detect in our bodies almost any respiratory pathogen out there. We’re not “infected” in any reasonable definition of that term.

In terms of the goals of the testing program what we really want to find out is whether someone can transmit the virus, or not. This is the ‘room’ split we’re really interested in. But the test is not doing that ‘split’ for us. It’s not capable of properly answering that question.

The important general point is that a false positive is defined by the probability question you’re trying to answer.

It’s a false positive with respect to a particular probability question.

The test was designed to detect viral fragments and not to test whether someone ‘has’ the disease or whether someone is infectious. The test is answering an entirely different probability question to the one that it was claimed to.

We’ve all become somewhat accustomed to the modern fetish for dicking about with definitions to serve a political agenda. In Woke Woo Speak™ we have, for example

But it also happened during COCK, too. We had re-definitions of things like herd immunity, vaccination, and pandemic, to serve a particular narrative.

In medical terms a covid ‘case’ was re-defined to be someone with a positive PCR test, rather than someone presenting with covid symptoms and who tested positive.

This was, in my view, a mass scientific fraud perpetrated on the general population.

It cost billions and served no useful purpose².

Even if we forget all the technical stuff and the statistical illiteracy in implementing mass testing, even if we thought the tests were all just tickety-boo, they achieved nothing to mitigate the spread of the infection.

It was all so fucking pointless - and unnecessary.

When the virus has already taken hold with thousands infected you’re just wasting your time implementing any kind of Test and Trace program, irrespective of the ‘quality’ of the test.

Right at the very, very, initial stages of an outbreak where only a handful of individuals have been exposed is the only time it might, possibly, make sense - and then you’d only be testing potential contacts.

Just to remind ourselves of just how stratospherically stupid it all became

There weren’t enough particles of Snarkon created during the Big Bang to fully describe this level of idiocy.

1. In the last article I used the symbol F for the specificity because I wanted to distinguish it from the sensitivity for which I used the symbol S.

2. In health terms. Obviously it made a small number of people much richer - so it served a ‘useful’ purpose for them.