Oof. If only the data actually looked like that. They quietly ended the placebo group during Christmas break after 2.5 months (about 2 years earlier than planned)! We will never know the true medium term efficacy of the vaccine, nevermind the long term. That made me so sad, and an indefensible decision IMO.
Luckily I think there's enough people who acquired immunity the hard way to save us.
First, thanks for digging up the source chart. You are right that the source matches the essence of the xckd chart. Very helpful and a pretty interesting and well made chart. I was loving these FDA reports and was so bummed when they ended the placebo group early.
"Always try to get data that's good enough that you don't need to do statistics on it."
Don't need to do statistics on that to see that xkcd added strong bias.
But more important than that, the problem is the right arrow on the X-Axis implies the placebo experiment is ongoing and did not come to an abrupt end.
You're saying we should have kept giving placebos to a bunch of people after the trial has calculated 90%+ protection for a widespread infectious disease? Would any review board actually clear such a study?
~30,000 volunteers with just a placebo against a disease with a >99% survival rate vs having NO CONTROL GROUP for a new class of vaccines that we inject in over 1,000,000,000 people. We have diseases with fatality rates of 50% that we still keep placebo groups for. This was not a hard decision.
We had a very clear scientific way to judge the long term efficacy of the vaccines but they dropped it. It was not to save lives—the math doesn't add up—it was because they made the sale and didn't want to "sell past the close".
Lets just presume you somehow sell that idea to the review board. How do you actually recruit for such a study? Who are these people that are willing to enter a lottery to receive the vaccine or placebo and won't just pop on over to the vaccination center and just get the vaccine for sure?
Sometimes it’s takes a human looking at a chart to see what a million tables of data can’t