There have been corporations who suppressed evidence that their drugs didn't work right. I'd describe that as publication bias, too, if not simply fraud...
I agree that there are long-term safety and efficacy studies that are fraudulently suppressed by some larger players in industry. It is an insidious problem that has horrible effects on patients. However, it is separate from my argument.
My point was that the registration of drugs is not effected by publication bias. The "suppressed evidence" typically comes in the form of post-marketing studies in significant patient populations, some of which aren't even randomized or placebo controlled (some are). NDA-directed studies are tightly controlled and reported. Then the drug is registered, and the FDA says, "we'd still like to see long-term safety data on this to see what things look like over time. We'll let you put this drug out there, but keep sending us more data."
I think the problem therein is related to the regs on post-marketing clinical studies. Companies have too much latitude to conduct post-marketing studies and then decide what to send FDA (or to publish), thus introducing the publication bias. FDA should have more control over Phase IV studies. The Prescription Drug User Fee Act (PDUFA, or pud-oof-a) was established to charge companies over $1million to register a drug, funneling money back into FDA so they could acquire more resources to review and grand NDAs. So the review times for NDAs went down, FDA got more efficient, and industry was happier. However, the act did not increase Phase IV controls as stringently as they should have, and companies have a lot of latitude on "additional clinical data" from studies conducted after the drug hits the market. That's why you see articles like the following:
So yes, there is publication bias and possibly fraud at work here, but it lies, in my view, early in the process (preclinical, basic science) and very late in the process (Phase IV studies). The registration process is pretty well controlled.
Note: My expertise is in preclinical pharmacology, toxicology, and NDA-directed studies (Phase I-III) not post-marketing studies (Phase IV), so if anyone has more experience here please speak up.
briggsbio|14 years ago
My point was that the registration of drugs is not effected by publication bias. The "suppressed evidence" typically comes in the form of post-marketing studies in significant patient populations, some of which aren't even randomized or placebo controlled (some are). NDA-directed studies are tightly controlled and reported. Then the drug is registered, and the FDA says, "we'd still like to see long-term safety data on this to see what things look like over time. We'll let you put this drug out there, but keep sending us more data."
I think the problem therein is related to the regs on post-marketing clinical studies. Companies have too much latitude to conduct post-marketing studies and then decide what to send FDA (or to publish), thus introducing the publication bias. FDA should have more control over Phase IV studies. The Prescription Drug User Fee Act (PDUFA, or pud-oof-a) was established to charge companies over $1million to register a drug, funneling money back into FDA so they could acquire more resources to review and grand NDAs. So the review times for NDAs went down, FDA got more efficient, and industry was happier. However, the act did not increase Phase IV controls as stringently as they should have, and companies have a lot of latitude on "additional clinical data" from studies conducted after the drug hits the market. That's why you see articles like the following:
GSK Suppresses Paxil Suicide Data: http://goo.gl/LCM94
AstraZeneca Suppresses Data on Seroquel: http://goo.gl/408fJ
Pfizer Suppresses Data on Edronax: http://goo.gl/fxLbW
So yes, there is publication bias and possibly fraud at work here, but it lies, in my view, early in the process (preclinical, basic science) and very late in the process (Phase IV studies). The registration process is pretty well controlled.
Note: My expertise is in preclinical pharmacology, toxicology, and NDA-directed studies (Phase I-III) not post-marketing studies (Phase IV), so if anyone has more experience here please speak up.