Posted by jhj on September 14, 2007, at 0:26:09
In reply to placebo vs. antidepressant, posted by Larry Hoover on September 13, 2007, at 13:51:57
I believe these conclusions to be unsupported by the available evidence. If these statements were true, what would we find when we looked at the placebo-controlled clinical trial data?
If we assume equivalence between an antidepressant drug and placebo, then under the statistical assumptions that govern what we call 'significance', 19 times out of 20, there would be no significant difference between the two experimental groups. When we look at all available clinical trial data, including all the studies that were never published, is this the case? No.
Under those same assumptions, assuming equivalence, both placebo and the antidepressant should have similar frequency of being found to be statistically superior to the other. Is this the case? No.
But there are other ways to look at the data. We can rank them, even those studies where significant superiority was not obtained. If equivalent, then placebo and the antidepressant should have a similar likelihood of being superior. Is that the case? No. Mean, median, t-test, I don't care how you look at it, antidepresants and placebo are not found to be equivalent.
I've previously argued that you cannot form conclusions about equivalence from obtaining null results in clinical trials. There are a number of reasons why that is so. One is that you cannot know if you conducted an otherwise valid study, but used the wrong subjects. Nor can you know from the results if your methodology was sound. Nor can you exclude chance. You don't know if there wasn't a difference to be found, only that you failed to find one, under the conditions employed. Again, the scientific aphorism, "The absence of evidence is not evidence of absence."
There is no disorder known to man that exhibits a more robust placebo response in clinical trials than does major depression. In a clinical trial you get validation, attention, a chance to be heard, empathy. So maybe some people do just need more love, and that's a big reason why they're exhibiting depression. For others, the difficulty might be more 'mechanical', i.e. biochemical in nature. Lumping both groups together in a clinical trial is going to dilute the effect of a biological intervention. It just stands to reason.
I am not saying it that placebo work as well as antidepressant.I am merely quoting from the one previous post to this same thread.
"but I can argue that placebos work as well as antidepressants in most clinical trials becuase that is fact based.
Linkadge"I am neither saying that placebo work as well as antidepressant nor am i saying that ADS incerase suicide risk.I am merely saying that if the statement i have quoted from previous post is assumed true then people should not be prescribed ADs.My belief is that ADs are more effective then placebos and they do not increase the suicide rates.
poster:jhj
thread:781684
URL: http://www.dr-bob.org/babble/20070911/msgs/782786.html