Hacker News new | past | comments | ask | show | jobs | submit login

Sketchy stuff. The two drugs that come to mind are Xigris [0] and Celebrex [1]. There is so much money to be made and so few incentives to be rigorous in the trial process. It's no surprise we get cases where big pharma companies cook the books on efficacy and go to great lengths to ensure there is a market for their latest "blockbuster."

[0] https://en.wikipedia.org/wiki/Drotrecogin_alfa#Marketing_con...

[1] https://en.wikipedia.org/wiki/Celecoxib#Fabricated_efficacy_...




The unintended consequence of the onerous FDA approval process is that only a few extremely large players have the resources to develop and bring new drugs to the market. Therefore they need to make a lot of money back to cover their costs.

On the other hand, in earlier times we had a lot of drugs that were in the "snake oil" category, making a lot of claims but doing nothing at all (or that were actually harmful).


But in a bid to mitigate risk (from their shareholders) big pharma effectively outsources lots of compound discovery to biotechs. This has helped create a large pool of biotech startups that never intend to run a phase 1 study but instead prepare for acquisition upon promising preclinical.


IMO the only criteria should be safety. Medicine doesn't have to work for everyone to be useful for a lot of people.


A great many drugs are grossly unsafe if you even minorly exceed the clinical dosage threshold, but are still regularly used by people because we haven't found better replacements.

A much larger set of drugs fall into a gradient of "as far as we can tell you'll just excrete it if you take too much" to "it's probably not _good_ if you unnecessarily take it, but it won't really be detectable in 4 weeks" to "oh god why would you ever prescribe that".

So if we were purely grading on safety, we'd be out quite a lot of pharmacology.




Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: