Contortionist conclusions (MiniMatters)
If your data don't support your biases, just twist yourself into knots
This one is pretty unimportant and old, but I saw it for the first time today and it made me laugh.
It is from this letter to the NEJM two years ago (right when we were getting the first inklings that the mRNA shots might not be the miracle 100% cure that Pfizer claimed).
Very short version: they looked at COVID infection rates among ~18,000 nursing home residents - 4,000 of whom were not vaccinated.
They found that the incidence of COVID infections was higher among the vaccinated. (Oh, and that two vaccinated residents went to the hospital with their symptoms - but none of the unvaccinated did.)
But, because they know it is patently impossible for the unvaccinated to fare better, this is their conclusion:
Did you get that?
The vaccine is so effective that it reduces cases among the unvaccinated… but not the vaccinated.
Also - the fact that some residents got COVID after being vaccinated means… we need more vaccination.
The only truly shocking fact is that the data were permitted to be published at all.