Having spent time wandering thru the FDA's website, reading warning letters etc, I'm all in favor of an FDA-type agency. Remember the Lilly drug NIAU for HBV, that worked like a charm , then ~ 3 months in, caused catastrophic kidney failure? http://www.healthforyou.org/m/medical/fialuridine.htm
Me, I'm all for an FDA.
All a new drug candidate has to do is show its: more efficacious than the existing standard of care, or otherwise beneficial to the patient.
"Otherwise beneficial" includes:
Works no better but fewer daily doses needed
works no better but better tolerated
works no better but less risk of serious side effects
works no better but less expensive
Cancer drugs can get approved if they can show an average increase in life-expectancy of days vs the existing standard of care.
Cancer drugs havent taken nearly the hit that contraceptives etc have. The Dalkon Shield lawsuits much shut down that industry. I cant remember the name of the drug, but it turned out to cause cancer in the female children of women who had taken it - not a happy outcome. Still - do you conduct a 2 generation study?
Clinical trials are incredibly expensive & many many drugs dont get thru the three trial process. Still, whining about the money is silly. If you dont think your drug is good enough, you wont put it on trial. There is even an 'orphan drug' status, for drugs to treat diseases that are sufficiently rare that there is no profit potential.
Safety trials start with 10-25 people, typically. That's a Phase I clinical trial.
It tests whether the drug will kill people outright, essentially.
Phase II are dosage etc determining trials. Different doses and regimens are used, and the company determines which is the best for the final test.
Phase III trials are big. This is where the drug has to prove itself better than the competition in sort of real-world conditions. Often, this is the first time a large number of people get the drug. Interesting things happen here.
One thing I hope is changing - often studies in the past have used 'surrogate markers' in place of endpoint markers, with the accepted assumption that the surrogate correlated with the true endpoint (ex: reduction in serum cholesterol equates to reduction in death from heart disease). Turns out quite a few surrogates are not as predictive as expected. I think trials are going to become more expensive as they turn to the actual markers.
Over prescription - amen to that! I'm seeing a lot of people (finally) told to do sinus flushes for infections, in place of the automatic antibiotics they took with each cold, 'just in case'.
Altoid - curiously strong.