I agree with the push for better precision medicine. I think such a future will be invaluable. But the way NNT is being used in this post is a gross misrepresentation. NNTs for the top 10 drugs come almost entirely from short trials, usually 2 to 5 years, while people live with these diseases for decades. An NNT of 25 over a 3-year trial does not mean only 1 in 25 ever benefits. It means 1 in 25 benefited within that narrow time window. Extend the timeline and NNT improves, sometimes exponentially, because absolute risk reduction compounds with age and exposure time. So taking short-term NNTs and using them to imply that most drugs “don’t work for the majority of people” is deeply misleading. It's unfortunate because when someone with a large platform posts statistics this way, it puts them in a tough bind. Either: (1) they don’t actually understand how NNTs and time-bounded risk reductions work but speak on them with confidence, or (2) they do understand and choose to frame them in a way that misleads the audience about what the data really shows. Neither option is a good look.