Posted by Larry Hoover on August 11, 2003, at 5:51:58
In reply to For Lori and Larry, posted by BobS. on August 10, 2003, at 20:14:10
>However, my point is that clinical trial studies are designed to make money for drug companies not help your child, regardless of what Larry and Dr. Kahn say. Just be wary of data as the article indicates.
I'm not supportive of the pharmaceutical industry, nor am I happy that for-profit enterprise has such a key role to play in mental health care. I have no doubt that information is selectively released by private enterprise. They develop the drug, conduct animal and laboratory experiments, and do three levels of human trials, all at great expense. Those are necessary processes, but they are done to provide a return to shareholders, not for anyone's health. Shareholders want profit, not social responsibility. Big tobacco (cancer), big pharm (side-effects), same thing, IMHO. Still, in a world where a woman can spill a hot coffee in her lap because she was unwise enough to take the lid off while driving, yet still successfully sue McDonald's ("They sold me hot coffee!" "Ummm, duh!"), I can't really blame them.
I don't know what message I was thought to have brought to the forum, but let me expand a little bit.
I am a professional scientist, last employed (pre-depression) in a job which required me to assess the published scientific work of others. I looked at methodology (subject selection, variables manipulated and controlled, etc.), data collection, data analysis (statistical methods), and then I very closely looked at the discussion and conclusions reached. One of the key determinants of good science is that the conclusions are supported by the data, but that is less common than you might believe. Many papers failed in the methodology alone, but that didn't stop them from being able to publish (unreasonable) conclusions, many times nothing more than a theory restated now as fact. I'm a skeptic, in general. Prove it to me. In my opinion, perhaps 10% of published work meets quality standards. And, that presumes that the data are accurate and complete. Still, I'd say that 60% of published papers are useful.
Just looking at the subject selection process for antidepressant clinical trials, it has been estimated that only about 8% of those diagnosed with major depression would qualify for a clinical trial, due to restrictions based on things like comorbidity, disease severity, age, predicted compliance, and so on. One of the limitations of selecting the trial population in this way is that the results are not generalizable to the broader, less-selected depressed population without considerable uncertainty. That needs to be more emphasized than it commonly is. You can't know what a drug really does (or doesn't do) until it's released for use by John/Jane Q. Public. We're all guinea pigs, but we can't do better than that if we want any new drugs to ever make it to market. When you get into "off-label" use of a drug, there is no data, period.
Another issue with clinical trials is the placebo responsivity. It has been argued (quite persuasively), in the highly specialized literature dedicated to clinical trial methodology, that the conditions applied in antidepressant clinical trials serve to maximize the placebo response (and thereby reduce the difference between the drug and what is commonly and mistakenly taken to be "no treatment" (something (i.e. no treatment whatsoever) that is no longer considered to be ethically acceptable in clinical trials, by the way)). It's important that people understand that the placebo response arises from more than the supply of a sugar pill. The level of care during a clinical trial is superb. (In the clinical trial I took part in, and in those I helped develop, that was the case, anyway.) There is no reasonable generalization from the level of care in a clinical trial setting and the 3 minutes with an uncaring HMO doctor that is more typically offered a depressed patient. No reasonable comparison possible.
I could show you a statistic derived from meta-analysis of active antidepressant treatment vs. placebo that shows that active treatment is more effective, with a p value of greater than 10 to the minus 30. But, so what? The real issue, and the only one to consider by individuals, is "Will it work for me?" There's only one way to find out. To take some. And cross your fingers.
That said, a patient taking a drug should have all the key information provided to them (true informed consent), and also have appropriate support during treatment (e.g. monitoring for suicidality during the first four weeks). Those are care issues, not drug issues.
What the science really says, and the conclusions (opinions) which have been publicized, are not the same, IMHO.
Lar
poster:Larry Hoover
thread:248910
URL: http://www.dr-bob.org/babble/20030807/msgs/249984.html