An important thing to remember when thinking about this thing: the downside of approving a drug that was dangerous is bounded: it kills people or causes some unexpected side effect and then it gets pulled. Meanwhile, the downside of failing to approve something that could have saved/improved lives is at least potentially unbounded: if it never gets approved that's the entire futures worth of humanity who is unable to take advantage of it (although admittedly I'd like to imagine that _eventually_ tech will improve to the point that we either figure out we should approve the drug and/or invent something else that obviates the need).

The point is that case B is both harder to notice and at least potentially _much_ worse than case A. And it's the case that the FDA currently errs towards.

The incident around thalidomide is basically FDA mythos at this point.

Not just lives also DALYs from glaucoma.

We all love a good trolley problem

And I hope we've all seen Dallas Buyers Club which showed the real, historical problem of drug approval being slow-rolled because it was killing a controversial minority