Read again, the 'proportions' are literally 1 in NNT, I think. NNT is the number needed to treat in order to make a positive difference for 1 patient, vs contro

Read again, the 'proportions' are literally 1 in NNT, I think. NNT is the number needed to treat in order to make a positive difference for 1 patient, vs control group with no treatment. NNT between two studies not directly comparable if they are of different durations. I'm still curious then how such high NNT drugs are blockbusters. Restating what I wrote above, I guess it is worth trying many of these, even with potential side effects, for a 1 in 4 chance the drug will help (seems entirely plausible) to even a 1 in 50 chance (less so). I suppose then that precision medicine could largely have people trying fewer drugs, as for people similar to the patient, the NNT might be lower (indicating a definite try) or higher -- but OTOH if people are taking a 1 in 50 chance now, what odds will actually make them bypass a drug?