COVID: Sweden was right. Everyone else was wrong.
(twitter.com)
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (22)
sorted by:
If the vaccinated group are just healthier people, like they were in the VA study I saw, then you shouldn't expect it to converge afterwards. They should have lower death rate post-covid as well, so converging could indicate lasting damage from the vaccine.
If the people who didn't get the vaccine didn't get it because they were in hospice then the vaccine wouldn't have kept them alive, and generally the older you are the more likely you are to be in hospice and know you're dying soon. Here you might expect the lines to converge later, because the people in hospice now thought they were fine a year ago and got vaccinated.
There's lots of ways simple data like "number of people who died" can be very misleading. Another example is the vaccines delay when you'll get covid rather than preventing it, so many vaccinated people got different strains from unvaccinated. Vaccinated with omicron vs unvaccinated with delta is attributable at least somewhat to luck; had the virus gotten more deadly the unvaccinated who got it early could have been better off.
And that's why we use randomized trials instead of self-selected ones, so you know it's not any number of confounding factors, but for the vaccine the control groups were disbanded so we don't have the data that says the only difference is the vaccine and by how much.
While I, in general, agree with your concerns about backwards looking analysis, confounders, and misleading data, in this case the numbers are so extreme that occam's razor points strongly in the direction of causality (or at least a shared cause). And since we don't have randomized prospective trials, we have to look backwards and do the best we can.
Speaking of shared causes, if you have any data or anecdotes to suggest that most of the unvaxxed people died in hospice and weren't given the vaccine because it was pointless, please send it my way, because it would drastically change my opinion (of course!)
Regarding differing average health levels of the two groups and the convergence. Totally agreed that this convergence could indiciate long-term health complications from the vaccine. Right now, if I had to bet, I would say both that the vaccines saved a lot of elderly lives, but will also cost lives in the future due to side effects (and given the extreme age stratification of death rates from covid should never have been given to the non-elderly).
And yes, it is 100% possible that the vaccines 'got lucky' by saving lives by preventing people from getting delta until omicron came about, but in that case the vaccines still saved lives. And I most certainly am not advocating for -more- mRNA vaccinations post-omicron.
Finally, since I never mentioned it - looking at just excess deaths in general between the two groups is basically the best you can do, since it limits bullshit from covid classification, but it is still susceptible to bad vaccination classification. More discussion here: https://reason.com/volokh/2023/01/10/no-lockdown-sweden-seemingly-tied-for-lowest-all-causes-mortality-in-oecd-since-covid-arrived/
There's not any doubt that the vaccine helps some people or that it's way more deadly than any other vaccine since variolation was a thing. The question is really just how effective and how deadly. Who should have gotten it and who shouldn't have.
https://www.science.org/doi/10.1126/science.abm0620
Figure 2 shows the vaccine delaying infection by 6 months, but not preventing infection. Figure 3 shows older non-infected unvaccinated with 0.88 survival at 26 weeks vs 0.96 for vaccinated - so 3x more likely to die from not-covid.
These were people who got a test for some reason so maybe there was something wrong, but still that's a pretty huge difference and it shows that you can get a big result from something besides what you thought you were measuring.
The thing about Sweden is I'm sure you've seen the Swedish bus stop meme. But yeah they came out well.