I imagine those numbers wouldn't be all that reliable for a bunch of reasons - companies would have vested interests in over or under reporting costs for various reasons, technology progressions mean previously expensive to develop drugs may be cheaper if they were to be redeveloped now, and sample size probably wouldn't be large enough for the raw data to be used when bankrolling pharma research ventures.

Edit: I should contextualise this by noting I have basically no experience with pharma so grain of salt blah blah

I think your points are fair. Sample size for their analysis was about 100 drugs. FDA approves about 25 per year.

My main point is that drug development is more expensive than most realize. Beyond R&D, a lot of that cost is in postmarketing (after approval) safety studies and surveillance- to catch side effects not observed in clinical trials. This process can be made somewhat more efficient as you suggested, but at the end of the day a human is reviewing every side effect report and reporting those back to regulatory agencies etc.

I am not qualified to discuss the economics re: your point on pharma r&d - drug discovery is it’s own thing aside from actual drug development.

I dont think you should be downvoted because thats valid, yet hard to use as an argument, paranoia.

The basic assummption in cryptography is “assume the target system is already compromised” so i would wager you are partially correct. Its just that its not the whole story, especially if u think that private persons get massssive discounts if they try to privately pay (from reading below comments about quotes to insurance companies vs private people).

Its a sad dynamic thats probably ruining more loves than it should.