> BTW, if you only talked to a few doctors, and one of them supports it, that's actually a fairly high percentage: 20-33%.
This seems wrong, because it's not taking into account sampling bias
This seems wrong, because it's not taking into account sampling bias