Looking at the top settlements here I think as a society we would be better off without these able to be this big. Generally speaking they all seem to relate to

Looking at the top settlements here I think as a society we would be better off without these able to be this big. Generally speaking they all seem to relate to off-label use (which I think we should generally be more permissive of) and false claims. But looking through one case, https://www.seegerweiss.com/drug-injury/rezulin-lawsuit/ https://www.seegerweiss.com/drug-injury/rezulin-lawsuit/ it seems that liver damage was a known side effect. I don't see any real indication that this was hidden, it was a part of the FDA filing. It seemed like a "net positive", and then later it was decided "no, probably not". The issues here seem more related to binary FDA categories. It basically either: 1) Banned 2) Approved and must be covered by insurance When there really can and should be more categories of risk, like: 1) Banned because we know this is super bad 2) Not banned, but not recommended 3) Not banned, seems ok but there are bad side effects or other concerns that warrant attention 4) Not banned, seems good, lots of good data and this should definitely be covered This would allow companies to put drugs in the market to help people, but set realistic expectations. There is no fraud if the data is there and the categories are clear.