I'm guessing this happens so rarely that it would have been missed in long term trials of much smaller groups of people.
You almost have to unleash it on 100 million people to have enough cases to identify the blood clots as statistically significant.
My opinion.
They still skipped long term testing and there have been a slew of side effects reported, many of which might have been caught.
https://media.tghn.org/medialibrary/2020/11/C4591001_Clinical_Protocol_Nov2020_Pfizer_BioNTech.pdf
Pfizer’s trial plan. Page 92 they talk about a 24 month follow up visit for people in the trial. I suppose that will happen in 2022.
Check out this article on the history of the drug Thalidomide:
https://en.wikipedia.org/wiki/Thalidomide
While it is not identical to the “rush to market” we have seen with these vaccines, there are some similarities.
Bottom line: Scientists didn’t really understand the drug until it caused a lot of damage, partially because it violated the belief at that time that drugs could not pass through the placental barrier.
Science is always learning—”statistically insignificant” may be in the eyes of the beholder.
This rare clotting issue occurs soon after vaccination. So for this situation, time isn’t an issue. Only sample size as you said.
Let’s say this clotting issue is unreported by a factor of 10. So in that case, the guestimate would be 1 in 100,000.
Estrogen-based birth control causes clotting in 1 out of 1,000. Just for perspective.
Long term trials most likely would have listed “clotting problems” as a side-effect and the vaccine would have been approved. 16 cases in 8 million is rare. But this is a public trial we are having and so the general public has a vote. that and one of the political parties appears to enjoy the control that fomenting a panic affords them.