Hacker News new | ask | show | jobs
by derefr 2777 days ago
The drug didn't cost $100 million to make. It cost $100 million to prove that it works in order to get it approved for sale in Europe.

In a world where you could just put a drug on the market (where "the market" is, say, hospitals), the drug would have cost far less, and so the inventor would need to recoup far less.

Note that I'm not arguing for complete deregulation of drug manufacturing. Testing for product integrity through factory audits and the like is the original—and still core—purpose of bodies like the FDA and its EU equivalent(s). You want to know that your "ibuprofen, 20mg" indeed contains 20mg of ibuprofen in each pill, and nothing else.

But that's cheap! That doesn't cost a hundred million dollars to do. Proving product integrity is something that any food processing plant manages on the regular, not to mention manufacturers of generic/non-patented drugs.

If we stop with the need to prove—with overwhelming statistical power—the safety-and-efficacy of a molecule that we're planning to dispense to fewer than 1000 people (and even then only through doctors who can carefully monitor them for adverse reactions), the pharma industry would go back to being just another quiet industry that makes stuff, instead of the crazy behemoth it is.

Yes, certainly, there should be a stringent process before you stick something on a store shelf for OTC purchase. But most drugs never seek to be OTC anyway. Prescription drugs require an interaction with a prescribing doctor, a dispensing pharmacist, and possibly (in in-patient settings) a bunch of watchful nurses. Why, also, must they require an interaction with a standards body regulating the molecule they're composed of? The patients who were in the clinical trial itself got effectively the same treatment, minus the standards body, and they were all fine.

Maybe that's what I'm saying: why not just allow any random chemical to be given to people (without intention of standards-body registration) as long as the prescribing doctor is willing to do clinical-trial-like monitoring for adverse side-effects?

4 comments

> (without intention of standards-body registration) as long as the prescribing doctor is willing

Unfortunately, the doctor (or scientist) cannot be allowed to make that decision on their own. There are far too many terrible examples in the not-very-distant past where doctors testing a new type of treatment drifted away from legitimate research into unethical experimentation[1], patient abuse, even torture and eugenics. Yes, most doctors wouldn't but the unusual-long length of the Wikipedia page [1] is why it's important to say "this is why we can't have nice things" require external approval.

That said, this approval are primarily concerned with preventing unethical experimentation and limiting the damage when the testing goes horribly wrong. Oversight of ethical issues and experimental technique should, in theory, be a lot simpler[2] and cheaper than the higher standard of proving that a new drug is "safe enough".

[1] https://en.wikipedia.org/wiki/Unethical_human_experimentatio...

[2] I know IRB approval isn't going to be "simple" for any type of experiment involving humans. It's easier than getting full FDA or EU approval.

>Maybe that's what I'm saying: why not just allow any random chemical to be given to people (without intention of standards-body registration) as long as the prescribing doctor is willing to do clinical-trial-like monitoring for adverse side-effects?

Snake oil, first. Imagine if e.g. Herbalife could claim their pills cure cancer? Good luck getting signal through all that noise.

Second, the inevitable democratic anger of the masses when the next 'cure for the common cold' (which I'd be happy to sell you, btw) starts killing people.

Not to say there aren't (large) improvements possible to the regulatory process, but the process itself exists for some very important reasons.

So maybe it should be more like the UL model? Let the insurance companies fund the proof of safety/efficacy, but it's not mandatory.
If by UL you mean Underwriters Laboratories in the US or the other US NRTLs, compliance testing costs are borne by manufacturers, not insurance companies.
The number of drugs that have gotten FDA approval at a billion+ cost only to be later rescinded is pretty massive as well... so it's not like the current system is fool-proof.
There have been a number of FDA and even OTC approved NSAIDs which may have killed off millions... I don't know who profits from the present system, but it ain't patients.