Read on Twitter
Remarkably, the FDA Commissioner doesn’t seem to understand elementary statistics.
If 13.7% die with low titer convalescent plasma vs. 8.9% with high titer, that could mean there is a 35% reduction in mortality [1-(8.9/13.7)]. (Citation: https://www.medrxiv.org/content/10.1101/2020.08.12.20169359v1)
1/3
If 13.7% die with low titer convalescent plasma vs. 8.9% with high titer, that could mean there is a 35% reduction in mortality [1-(8.9/13.7)]. (Citation: https://www.medrxiv.org/content/10.1101/2020.08.12.20169359v1)
1/3
It does NOT mean that for every 100 people given the drug, 35 survive who wouldn’t otherwise (which is what @FDACommissioner said: http://ow.ly/wOW450B6Tw6 ).
It means that for every 100 people given the drug, 13.7%-8.9% (≈ 5%) are saved by the drug. Not quite 35%.
2/3
It means that for every 100 people given the drug, 13.7%-8.9% (≈ 5%) are saved by the drug. Not quite 35%.
2/3
And that’s if you believe the non-randomized data (which you shouldn’t). How could this pass muster at FDA/HHS/White House?
This is what happens when science is devalued. 3/3
This is what happens when science is devalued. 3/3
