The pause in AstraZeneca vaccination based on dubious associations makes me think of 👇: a randomized trial where the treatment was… nothing. By random chance, the treatment group, who, again, had nothing done to them, died much more. You need to have a theory as well as data. https://t.co/Y0kUsqebfs

— Ethan Mollick (@emollick) March 15, 2021

With a low background rate, doesn’t take much of anything to skew results. People have a lot of trouble reasoning about things like this.