Mass testing, ICU occupancy and vaccine efficacy

On the current situation in Sweden

The Swedish media narrative around covid is now focusing on three things.

1. An alleged surge in cases (people who are PCR-positive irrespective of symptoms or anything else; or people who test positive with an antigen test and present with symptoms typical of covid)

2. That the majority of ICU patients are unvaccinated, and

3. That the vaccines therefore are very effective at protecting people from serious illness and death.

This is basically the entire rationale for the return and the slow intensifying of the restrictions and the implementation of the vaccine passports. Incidentally, all of this is in spite of 2021 actually giving us Sweden’s second-lowest number of deaths per capita in history. Yes. Read that again.

And believe it or not, we never had this few deaths per capita in the older cohorts (see SCB.se, the government’s agency for statistics).

The surge in cases is to a high degree an artifact of the PCR testing. We see panic headlines reporting record numbers of cases, but there’s never any reference to the relative number of tests performed. That said, the ratio of positive tests in relation to the total is now actually a bit higher, yet this is exactly what you’d expect with a virus that has gone endemic and spread throughout the population for at least two years.

(However, the official definition of a case also includes a positive diagnosis coupled with a positive antigen test, which possibly confounds the data to some extent, since these antigen tests are not included in the tally of performed tests.)

At the same time, there are basically no deaths, the ICU occupancy is below average, and there are almost no covid-positive ICU admissions or deaths per case:

(source: folkhalsomyndigheten.se via softwaredevelopmentperestroika.wordpress.com)

For this reason, it’s slowly being acknowledged that the vaccines are useless with regard to stopping transmission. You’re not going to have a record number of cases coincidental with a vaccine uptake around 80% if they make a significant difference in this sense.

Focus is therefore now almost exclusively on how the vaccines significantly protect us from “serious illness and death”.

While this conclusion is also kind of a stretch in light of the above data since the absolute numbers are reduced to such a level that the margin of error starts making inferences very unreliable, we nonetheless get daily news reports emphasizing how the vaccines must be working since the unvaccinated positively swamp our ICUs. Among the ≈40 patients.

Looking at this data, to begin with, total numbers in the tally include the data from before the vaccine rollout. The numbers from week 52 state that 3 804 patients out of 3 995 were unvaccinated, i.e. the give you the total for the entire year. But to use this data point in anything akin to reporting on the relative numbers of vaccinated vs. unvaxxed ICU patients is entirely nonsensical, since that would implicitly ascribe efficacy to the vaccines before they even hit the market, and way before we had any meaningful uptake. They’re probably not that good.

Any such argument will also chalk up to the vaccines all the other factors that reduced overall ICU numbers before the advent of vaccination, such as e.g. seasonality or natural immunity. Fortunately, I don’t think this line of reasoning is very common, but to even address the total numbers in this way and in this context is problematic.

Moving on, we’re dealing with a definition of “vaccinated” that decidedly favours the impression of vaccine efficacy. In the published reports on ICU occupancy, Sweden’s public health agency counts patients as unvaccinated until two weeks after their second dose. This practice has been discussed and criticized elsewhere, not least since we know most AEs take place during the first two weeks after treatment. This counter-intuitively pushes most sufferers of severe adverse vaccine events into the “unvaccinated” category.

This also holds true for the consequences of the immediate immunosuppression that follows from the vaccines (which is nothing unusual or peculiar to the covid vaccines as such).

On top of this, it must be emphasized that we’re dealing with very small sample sizes:

There are remarkably few patients with covid in the ICU from the summer and onwards. The small sample sizes renders confounding factors potentially enormous, especially in light of the endemic spread of the virus.

In other words, since everyone and their mom has covid and people in general thus are likely to test positive, anything that just moderately correlates with being unvaccinated while also negatively impacting your health, such as poverty or immigrant status, may have a huge impact on the relative ICU occupancy of the two categories in question. Tiny sample sizes = large margin of error, especially with lots of confounding variables.

And of course, the de facto sample size shrinks further if you account for the fact that only a fraction of the ICU patients are actually there due to covid in the first place. I.e. I don’t care if you see a correlation between being unvaccinated and tending to slip and fall in the shower - we have no business bringing vaccine efficacy into the discussion of any set of cases unless there’s a clear causal link between covid infection and ICU occupancy.

Finally, the official Swedish narrative on how the unvaccinated are swamping the ICUs, implying the wondrous efficacy of the vaccines, is contradicted by reams of data on fading or absent vaccine efficacy from other countries, such as Israel, the UK or Denmark.

In summary, the authorities in question really need to immediately publish detailed data on everything from precise vaccination status to gender, age, co-morbidities, socio-economic factors &c. Without this, it’s entirely impossible to ascertain anything meaningful from this jumble of pre-packaged information and miniscule samples, and the authorities’ and the media’s claims of an obvious vaccine efficacy remain unwarranted.

And once again, there are still basically no deaths, the ICU occupancy is below average, and there are almost no covid-positive ICU admissions or deaths per case.

Edit: an article just dropped in one of our major newspapers. The lead:

“The data on covid-19 patients may be misleading. Several regions alert the FHM that the numbers may in fact be lower.

One example is patients being treated for fractures but who are counted as ‘covid-19 patients’ due to testing positive during admission.”