If you read more closely you'll notice they're talking about a particular kind of thrombosis[1] which is accompanied by low platelet counts. This particular symptom is usually sign of a particular autoimmune response and has an incidence of about 4-5 per million people per year.

[1]: <warning, contains images of heavy bruising> https://en.wikipedia.org/wiki/Thrombotic_thrombocytopenic_pu...

so basically, base rate fallacy strikes agan.

If the risk turns out to be 1 in 50.000, which would be on the conservative side if the observations from Norway are correct, it's not a clear-cut question in a country with low incidence of Covid. Especially if other vaccines are available with a small delay.

The British tool for calculating covid risk puts risk of death as 1 in 500.000 for a healthy 25-year old, and risk of hospital admission as 1 in 13.000.

Must also add to this the public's willingness to vaccinate in the face of these numbers, both now and in the future. It's a shifting of risk from high-risk groups to low-risk groups.

> risk turns out to be 1 in 50.000

It's more like 1 in 500,000.

The numbers are three healthcare workers out of 120,000 vaccinated in Norway, with one death, so the risk is 1 in 40,000 from what we know.

On top of that we can speculate that healthcare workers are more likely to seek treatment than the average vaccine recipient and they are healthy adults of average age.

Everyone would seek treatment, the clotting has like a 40% fatality rate.

That's an odd sample to use unless you have some sort of agenda. The 40 out of 17MM is much more statistically significant.

If you give a vaccine to 17 million people and 40 of those are children and all 40 children dies then should we just ignore that fact because the "40 out of 17MM is much more statistically significant"? Of course not.

That's the epitome of a gross exaggeration.