23 Comments

Can you explain “analysis based on person-years” please?

Expand full comment

Comments under Catastrophic Health Collapse (part 7) cover this and part 6 is also worth another look, but here's the nitty-gritty pasted again:

"Person years is classically used to account for differing exposure to a treatment such that 100 people getting jabbed in January will notch up 600 months (50 person years) of post treatment period between them come June, compared to only 300 months (25 person years) for folk getting jabbed in March. We can also use 'person years' as a form of weighting such that a cohort of 100 60-year-olds will possess a combined age of 6,000 person years, compared to 3,000 person years for a sample of 30-year olds. When using person years in this manner - as I have done for part 6 - you get age-weighted results that are similar to those if we just count heads and ignore age at admissions. The difference between the two highlights a shift in the age profile, which can be very useful."

Expand full comment

I may be dumb but I can’t see why or how a time dimension has a place in comparing admissions and treatments. How have you weighted a 5-month vaxxed person versus a 2-month vaxxed person versus unvaxxed? You say at the start that vax status is binary.

Expand full comment

It was an idea I wanted to try out to avoid logistic modelling but it might just suck. Ideally I'd spend time with a simulation study, but I may re-jig the analysis so that we're back in head counting territory and accept the many issues that go with that. I suspect the whole article is pointless but at least subscribers have got something to chew over.

Expand full comment

I've revised the article based just on head counts alone. This is less confusing (though we need to acknowledge confounding factors). The differences are rather interesting indeed!

Expand full comment

Yes - seems to raise a lot of questions. Shame you don’t have more data - I guess you can overanalyse stuff when your source data is restricted. However ONS presumably have the full data.

Expand full comment

Making a meal from a few scraps :-) Individual Trusts will have the full data providing they've got staff who can do the wizardry to link databases, but that's a big issue right there. After speaking to folk I have my doubts on data reliability.

Expand full comment

So do I yet it is THE most essential thing. Who audits this stuff and have they been asleep at the wheel? Are they any good at their jobs? Why is it left to Prof Fenton, HART and your good self to check the numbers? A stratum of assurance is missing from our health data.

Expand full comment

Re your last paragraph, worth noting that in the Pfizer trial, outside of the predetermined first 180 positive test results, overall the difference was a mere 12% between vaccinated and Placebo !

Expand full comment

Too many confounders to hoist this particular flag IMO. And of course one could also invoke "it would have been worse". Maybe some control tests could be done. Did we already do a "covid severity index" (e.g. ratio of respiratory procedures to positive covid tests)? I can't remember. Or maybe check a control infection that has no vaccine (e.g. ratio of respiratory procedures to all-admissions, in those with a positive RSV test, versus those with a negative test)? Maybe that was also done already but I forget.

Maybe I'm dyslexic, but it seems to me age-weighting exacerbates age-confounding, as opposed to the opposite, likely making vaccines look worse. But as you've stated, it seems to do little in any direction.

Maybe the ideal code to analyze would be something that patients in essence request/prescribe themselves that indicate their symptom level, as opposed to doctors presuming unvaccinated people need more procedures. I can't think of anything.

Expand full comment

Yes, it's a rather dubious attempt. I did have 'crude' in the title but maybe that should be 'rude'! Yes, we have CSI and a few other bits and pieces and that can be dialled in to a logistic model. I'm not convinced about the weird age weighting method but fancied trying it out. Back in HART report 20 I was toying with a construct called 'base risk of hospitalisation' and I attach a link to the latest draft. The issues with this sort of work are abundant such that I might pull the article and stomp around a bit - but at least subscribers will have a copy.

https://drive.google.com/file/d/1-UYepKXGuzkqITnbK5tWlFGJ4j3dNvuI/view?usp=share_link

Expand full comment

Is Covid really a respiratory disease? I mean I've had it, twice, and it didn't really affect my breathing at all. My upper respiratory tract was pretty well clear throughout.

b) what kind of other conditions/ side effects/ complications might someone vaccinated report to hospital for? I mean, heart issues seems likely.

Expand full comment

Well, that's how it was designated and sold to the profession as well as the public, so that's the mast that I nail it to for 'bottom-line' numbers.

A different strand of research (and one I may pursue) is what COVID actually did to folk. Back in autumn 2020, when I followed case studies, I listed out all the symptoms that were being reported, which quickly became a bewildering matrix. I came to the conclusion that either the medical profession was deluded or we were looking at a list indicative of radiation poisoning. Most curious indeed!

As for vaccinated folk with troubles we could use the same data to look for trends in their symptoms, and I may do this whilst I'm at it.

Expand full comment

others appear to be tracking down blood disorders and cancer.. all way above my paygrade anyway, but I'm eager to learn more, as in Scotland we are still as of April 2023 experiencing higher death rates than the five year average... (10% higher in numbers, maybe 4% age adjusted). And still people are coming down with Covid - my multi vaxxed young friend attended a concert two weeks ago, fell ill, and is still testing positive.

Expand full comment

Yes indeed. A colleague of my wife's who had been clear of cancer for several years suddenly regressed and died within 4 weeks after his third booster. Her best friend managed to survive clotting in her cerebellum, ditto. I lost both uncles through clotting, ditto. All her multiple-boosted fellow teachers have been plagued by various illnesses and test positive at the drop of a hat. Bizarrely nobody questions the experimental mRNA therapy - I suspect they'd rather not know at this late stage. It would be great to track all this but the data dump I possesses only runs to Sep 2021 so is of limited use.

Expand full comment

I have a different take on this post. Excess deaths, iatrogenesis and the catastrophic health collapse, the transmission of influenza, vaccine efficacy. These are important, but highly complex and contentious issues. For these and the other issues covered, this blog is in the very top echelon! I learn something new with each post, but - more importantly - each of them challenges me to sharpen my critical thinking and analytical skills. The analysis that goes into all of these posts is original and thought provoking. Please keep up the great work. Now, about that slice of Mrs. Dee's scrumptious Somerset Apple Cake...

Expand full comment

I wear the obligatory cilice of the statistician 😄

Expand full comment

Thank you for this. Highly interesting.

I think the vaccines turned people into superspreaders.

For an illustration of this based on US data: https://q3deathwave.pervaers.com

Expand full comment

Another excellent bit of baking.

I think an ingredient error may have occurred here though:

“I am forced to conclude that world governments ought to get their money back because the *unvaccinated* population have been faring less well than their *vaccinated* counterparts.” A Faucian slip?

Expand full comment

Enjoying the optimistic 1M target in the Build Back Better Biscuits GoFundMe. Maybe easier to reach with a suitably titled money-triggering research grant application:

“Rising Temperatures, Crumbling Dough: An Investigation into the Impact of Climate Change on Biscuit Texture and Quality"

Credit to ChatPGTips

Expand full comment

LMAO - what a project! I can't wait for the press to print the headline: 'a new study shows the world's biscuit supply may be in danger'

Expand full comment

Whoops! Corrected. Should have gone to Specsavers.

Expand full comment