Public Health Scotland are ditching their Corona statistics, “over concerns it’s misrepresented by anti-vaxxers.”

In an isolated demand for rigour, a bizarrely anonymous PHS official explains that these are just super simple numbers that don’t mean anything:

The case rates, hospitalisation rates, the death rates are very simple statistics, whereas for the vaccine effectiveness studies we use modelling, we compare people who have tested negative to those who have tested positive and match them on their underlining co-morbidities. …

What is happening is people are looking at those simple data and trying to make inferences about the vaccination, whether the vaccines work, inappropriately and sometimes wilfully.

There are so many caveats and they just pull certain figures out that should not be used.

What we are going to do is do a lot more on the vaccine effectiveness side and try and make people understand how effective the vaccine is.

For example we know it is 50 per cent effective against getting infected

1

, but that it is much higher effectiveness against hospitalisations and deaths which is the key thing really as that’s what we want to prevent.”

Of course, nobody has time for all these caveats, when it comes to complaining about how many unvaccinated people are in hospital at the moment. Millions of people across Europe have been excluded from public life for months on the strength of “very simple statistics” like those published by PHS. The only problem is that the numbers aren’t going their way anymore.

Yesterday’s PHS report has a whole wall of text defending their decision:

COVID-19 rates do not account for potential differences between populations such as behaviour towards social distancing, underlying health issues, or approaches to testing and how much they contribute to COVID-19 rates. For example, a study found that people with two doses of vaccine were more likely to test themselves for COVID-19 compared to those unvaccinated or with one dose of COVID-19 vaccine. This means that unvaccinated people may be less likely to test and report as a case resulting in lower infection rates among the unvaccinated. Comparison of case rates is therefore now inappropriate.

Emphasis theirs. The vaccinators, whenever their statistics go bad, are always quick to declare that there are systematic, population-wide differences among the vaccinated and the unvaccinated populations. There surely are, and these differences must colour our entire picture of how well the vaccines work. Whenever they’re not having statistical problems, though, the vaccinators develop a profound indifference to the problem of statistical confounding.

Individuals who have not completed their vaccine schedule may be more susceptible to a severe outcome and could result in higher COVID-19 case, hospitalisation and death rates in the first and second dose vaccine groups. For example, some of the older individuals who have exceeded the recommended time will have not received their next vaccine dose because of frailty or ill health. They may, therefore, be more likely to be hospitalised or die if they get COVID-19. Comparison of hospitalisation and death rates is therefore inappropriate.

So, it seems, “frailty or ill health” might be good reasons to avoid vaccination in Scotland. I wonder if that is the case elsewhere and what effect that might have on our overall impression of how well the vaccines prevent severe outcomes.

There is a lot uncertainty about the number of people in the unvaccinated population because we do not know if individuals are still resident in Scotland. This is unlike the vaccinated population who have had recent contact with the NHS due to vaccination. As case rates require accurate population estimates, the uncertainty in the unvaccinated population will result in an underestimate of case rates. Comparison of case rates is therefore inappropriate.

Excellent, so we can trust that the high age-adjusted case ….

…. hospitalisation …

… and death rates for the double-vaccinated are accurate, since these people “have had recent contact with the NHS due to vaccination.”

The report authors also plead the law of small numbers and vaccine fade:

Vaccinations have not been given to the entire population at the same time … . A number of people have not completed or exceeded their recommended dose schedule. This leads to reduced protection from the vaccine over time. …

Yes. This is why the boosters seem to work so well; the boosted population, by definition, has been vaccinated fairly recently. The question is how often they will need to be vaccinated in the future, to avoid ending up worse off than the naturally immune.

See also Igor Chudov’s remarks on the the PHS decision.