alberta gets caught palming cards on covid vaccine efficacy

and we get real quantification on the size of the definitional datacrime. it shocked even me.

ok, i’m just gonna say it: this is great work.

joel smalley (a gatopal™ and another “booted by bluebird” refugee) nailed this. it’s a thing i’ve been wondering about for some time but i never found a dataset from which i could quantify it.

but joel did and this is great stuff. he literally had to scrape pixels to get this. it’s A+ sleuthing. go read it.

(and you know it’s the good stuff because alberta already took it down the minute they realized what they had blabbed. but the internet, well, the internet remembers. let’s help it route around censorship.)

Alberta just inadvertently confessed to fiddling the COVID vaccination stats.

in short, the issue goes like this:

there is a 2 week window of immune suppression post vaccination. it roughly doubled the base rate chance of healthy people getting covid back before even delta. (it’s likely far worse now given omicron and OAS issues) this has been demonstrated in trials.

this gets not only covered up, but willfully misattributed. i wrote THIS on it some time back. it’s pretty simple.

this is the essence of bayes. it’s how you measure and aggregate real relative risk and outcome. but the definition of “vaxxed” as “dose 2 +14 days” which breaks this utterly.

it attributes the risk of running across the field to “staying in the foxhole.”

if you get sick or hospitalized or die in that period, you get called “unvaxxed.” the increased risk you face due to immunosuppression should be associated with vaccination. instead, it gets attributed to lack of vaccination.

this is frightfully dishonest.

it’s also frighteningly effective if you seek to lie about efficacy.

bayesian datacrime: defining vaccine efficacy into existence

i laid this out as just a basic walk though on the math. the numbers were thought experiments. but they clearly show that you can make a vaccine with zero efficacy and actual net harm from early immunosuppression look highly effective with just this one definitional game.

i could not quantify it. joel did. the numbers are eye popping and the 2 week period of greatly accelerated vulnerability is plain for anyone to see.

it’s outright glaring. it might well be a 5X accelerator for hospitalization and 3-4X for deaths over that 2 week period vs prior base rate. it might well be 10X in cases.

these are stunning numbers. the fact that they are not being talked about is outlandish. the fact that they are being hidden and misattributed by people who ought to know better is insidious.

here’s hospitalization for covid:

see how base risk rate soars in the first week and then sits at a high plateau for the second? yeah, that all gets called “unvaxxed.”

that’s getting shot running across the field and getting called a foxhole death.

47.6% of hospitalizations post vaccination were in the first 14 days. and the rate falls off rapidly right after. (left is data by day, right is aggregate)

i mean, you can see the inflection clear as day. this was not chosen at random. this was a laser guided precision strike to shift the blame. it’s literally reducing the overall vaxxed hospital count by half and adding them to the unvaxxed.

this is the same garbage that went on in the clinical trials. they did not count this period when calculating efficacy.

deaths are even more stark with 55.6% in the first 14 days post jab.

and it all got shifted the same way. this one dodge alone means that ~56% of all post jab covid deaths get called “unvaxxed.”

this data extends all the way to the present (100% of outcomes). i confirmed this with joel.

there is no way to recover a real, comparative data signal from this.

given the huge front loading on cases, hospitalization, and deaths, this one definitional stunt could literarily be the entire source of apparent vaccine efficacy and that may well include the drug trials.

it also means that (barring other bias selection) the vaxxed are more likely to also be “recovered” than the unvaxxed.

consider:

for the sake of example, let’s you have a base rate of 5% infection per month. (this number is arbitrary, but we need to use something.)

once you are infected, you do not get infected again.

we presume vaccines have zero effect on infection rates after the initial immunosuppression period. (this gets MUCH worse if they have negative VE’s, but as you’ll see, it’s not needed to make the case)

but immunosuppression means the vaxxed get 63% of total infections in the first month. (this is about right per joel. again, note the conspicuously beneficial placement of 2 week cutoff…)

so, infection rate looks like this in vaxxed vs unvaxxed. (this is a little contrived but the basic relationship it lays out is sound)

now presume that once you get covid and recover, you have a 99% lower chance of being hospitalized for covid in the future (this is roughly correct).

so take a baseline 10% case hospitalization rate in a naïve population (much too high but makes the math easy).

and we get this: (this is case hospital rate as reported in a case randomly distributed in the population, so the downward slant is the building of herd immunity to severe disease)

~25% apparent vaccine efficacy from literally doing nothing except getting people sick faster. (ironically, the strategy so many have railed against so vehemently)

and this is BEFORE we pile the hospitalizations from the first 14 days into “unvaxxed” and salt their number while shaving the vaxxed. do that on the order of 50% of reported hospitalizations (as above, 47.6% in two week misallocation window) and the vaccines could well have negative VE on hospitalization and death and you would not be able to see it.

it would show up in cases though, and what do you know, it does. bigly.

that’s some magic trick.

it’s also a nasty possible explanation on why all cause deaths are up in so many places that vaccinated. that figure will catch this issue even if the drug trials and public health reporting is all set up to misattribute it. seeing it rise year on year despite milder variants raises some very pointy questions.

why are northern european (except for sweden) all cause deaths up in 2021 vs 2020?

what we really need are all cause deaths data, cohorted by age, comorbidity, and vaccination status starting right from the day of your first jab.

it’s 100% out there. lots of the single payor system countries will have it.

and if it were good, i suspect we’d have seen it by now.

the fact that we have not speaks volumes.

and so will using these same bogus definitions of “dose 2 +14 days” if and when they do release such ACD data.

that’s not how you analyze an outcome and they know it. this is week one of trial design 101 stuff.

this was not an accident. this was a choice.

the drug companies that ran these trials know more about how study design and definitions affect reported outcomes than anyone on earth, bar none.

they did not screw this up.

they did not pull it out of a hat.

they stacked the deck.

how have you liked the cards you’ve been getting?

the FDA tried to release the actual trial data over 75 years. (yes, years) boy does THAT make you wonder…

it’s time for deep, serious transparency, and it looks like we may get our shot:

and i know that a lot of us are looking forward to it.