New at Reason

July 10, 2007

Ronald Bailey mulls over the challenge of the FDA: when to approve medicines and at what risk.

Comments

Dave W.|7.10.07 @ 8:27AM|#

Later that year, then-New York Attorney General Eliot Spitzer sued the drug company GlaxoSmithKline, alleging that it had hidden data showing that its antidepressant Paxil produced suicidal thoughts in adolescents. . . . In 2004, GlaxoSmithKline settled with Spitzer . . . . and agreed to publish the results of all of its clinical trials online.

Good lawsuit, good outcome. Naught PharmaCo!

soulrebel|7.10.07 @ 8:38AM|#

all one needs to do is smoke a joint. nature's antidepressant.

Dave W.|7.10.07 @ 8:55AM|#

They executed that Chinese guy in charge of product safety:

http://news.bbc.co.uk/1/hi/world/asia-pacific/6286698.stm

Jim|7.10.07 @ 9:34AM|#

all one needs to do is smoke a joint. nature's antidepressant.

No, it has been declared my that meddicaly qualified agency, congress, as having no accepted medical value and too dangerous to use, even under close medical supervision.

|7.10.07 @ 10:36AM|#

I've been critical of Mr. Bailey in the past but I found this to be a fair and insightful piece. He doesn't quite go so far as to explicitly sympathize with the FDA's difficult job of balancing the need for innovation with the duty to protect the public but he comes pretty close...

|7.10.07 @ 10:50AM|#

"It is also likely that the use of antidepressants is a marker for more severe depression and patients suffering from severe depression are more likely to be suicidal."

This has always been my take on the issue, which got quite a bit of coverage in Indianapolis (home of Lilly): if an individual taking antidepressants commits suicide, I am happily willing to believe that the reason is not that the drug *caused* that patient to commit suicide, but simply didn't prevent it. Not all drugs are effective on all people, if you can believe that.

|7.10.07 @ 10:55AM|#

barris: At the risk of falling out of your favor again, I have made some proposals for modernizing the FDA's approval process in my 2005 column, "The End of Old Regulatory Rituals."

|7.10.07 @ 11:02AM|#

Not all drugs are effective on all people, if you can believe that.

Nor are they immediately effective. Some anti-depressants can take several weeks to become fully effective. Which would correlate with the larger decline in suicides at least a month after beginning treatment.

|7.10.07 @ 11:09AM|#

Bailey makes the very good (and, in a perfect world, obvious) point that the FDA's incentives are to delay drug approval, regardless of how many people they fail to save.

I really wish everyone understood the existence of opportunity costs. It'd make the world so much better.

|7.10.07 @ 12:25PM|#

As insurance and veterans' medical records become more available for analysis, it becomes ever more possible to find possible adverse effects. Not to mention the fact that most drugs benefits are small with regard to ameliorating morbibity and mortality - when measured in clinical trials. But the scientific data from clinical trials, although the most scientific method for evaluating drugs effect, is a median. Drugs are prescribed for individuals - and the individual risk/benefit ratio can only be determined on a case by case basis. We will soon relearn the obvious - every prescription drug can kill someone, and every prescription drug can help someone.

|7.10.07 @ 1:22PM|#

MR. BAILEY!
FOR SAME CRIME FDA EMPLOYEE IS DOING AGAINST PEOPLE IN USA IN CHINA THEY ARE EXECUTED.
PLEASE PAY ATTENTION TO THAT.

Caps Lock Key|7.10.07 @ 2:50PM|#

RK,

Please pay attention to me.

|7.10.07 @ 3:43PM|#

MR BAILEY
DO YOU OWN ANY STOCK IN DRUG COMPANIES OR NOT.
DISCLOSURE: " I DON'T THINK" IS UNCLEAR

Caps Lock Key|7.10.07 @ 4:02PM|#

RK,

Really. I mean it. I'm right here next to the "A" key. Just give me a little tap.

Gahan|7.10.07 @ 4:40PM|#

"all one needs to do is smoke a joint. nature's antidepressant."

Getting high actually causes me more depression and anxiety, which is why I stopped quite a while ago. I'm aware that it doesn't have this effect on everybody, but it's not exactly rare, either.

Anyway, to each his own, I guess.

|7.13.07 @ 4:07PM|#

None of the studies cited provide any compelling reason to repeal black box warnings. In many of these studies antidepressant related suicidal events are higher than placebo related events and the improvement in depression on antidepressants, if it occurs at all, is based on very marginal changes in rating scales that are highly subjective, usually a list of vague questions requiring self-evaluation of mood. The ratio of antidepressant related events to placebo events is often vastly understated since so-called placebo events are often actually withdrawal events. Events that occur during run-in to a clinical trial in which patients are often abruptly withdrawn from prior treatments are labeled placebo events when they are in fact really withdrawal events. The same thing occurs after the trial period when patients are taken off the investigative medicine. Events that occur at this point are called placebo events when again they are withdrawal events. Clinicians rarely understand or acknowledge withdrawal. So the occurrence of placebo events is inflated relative to drug events. Many of the placebo events are really drug events (i.e. withdrawal). If these events were rated accurately antidepressant related events would be much higher relative to placebo events which would be almost non-existent. And when these researchers refer to "untreated depression" they are in fact often really referring to TEMPORARILY untreated depression, not NEVER treated depression. In fact "untreated depression" is often drug withdrawal or rebound effects from prior treatment.

Comparing correlation between suicide rates and antidepressant prescriptions is also a very weak way to evaluate safety as suicide rates could be going up because of skyrocketing stimulant and antipsychotic use even if antidepressant use is going down.

We urgently need to discourage indiscriminate and cavalier overuse of dangerous drugs in children and adolescents especially. Black box warnings are one means to achieve this. There really is no deep understanding of the mechanisms by which antidepressants "work" and we are using a whole new generation as guinea pigs.