tag:blogger.com,1999:blog-8285737581949527996.post1668183438640541293..comments2023-10-11T09:14:08.345+01:00Comments on Pyjamas in Bananas: Misrepresenting sciencepjhttp://www.blogger.com/profile/06832177812057826894noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-8285737581949527996.post-73734321087457728172008-02-28T21:05:00.000+00:002008-02-28T21:05:00.000+00:00phil - none of this bad science crap, I wanna be a...phil - none of this bad science crap, I wanna be an evil scientist!<BR/><BR/>black duck - it is actually quite worrying that based on a single study of patients with mild depression and <I>extrapolating</I> a regression line from more severe depression the message has been sent to the public (and GPs) that antidepressants are useless.pjhttps://www.blogger.com/profile/06832177812057826894noreply@blogger.comtag:blogger.com,1999:blog-8285737581949527996.post-60185324498215126012008-02-28T20:06:00.000+00:002008-02-28T20:06:00.000+00:00Good post!Wouldn't it be great if you could test t...Good post!<BR/><BR/>Wouldn't it be great if you could test the effect of just the drug. Simply get a friend of the test person to slip the pill into his morning cuppa while he's not looking. You would then have three sets of results: just drug, just placebo and placebo+drug. You could then find out whether the relation is purely additive, whether one masks the other, or whether they somehow wonderfullly combine to together achieve more than they would individually. <BR/><BR/>Of course such a test would be unethical, and might spoil a good cup of tea ;-)<BR/><BR/> In the absence of the third test, we have to make do with the comparison between the drug and a placebo. The only relevant questions are "Does the drug help the patient?" and "Does the drug help the patient better than the placebo?". If both answers are yes, then the drug works and the effect is real.philhttps://www.blogger.com/profile/14342456012514447897noreply@blogger.comtag:blogger.com,1999:blog-8285737581949527996.post-20797312256628365042008-02-27T00:06:00.000+00:002008-02-27T00:06:00.000+00:00The whole thing is very interesting. Simplistic re...The whole thing is very interesting. Simplistic responses, such as 'the drugs don't work', do not look right.<BR/><BR/>In one sense, this result is pretty much what you might expect for a placebo responsive condition. For mild cases, a placebo may well be sufficient. The placebo + treatment is not additive because there is no where to go once a certain amount of benefit has been achieved. For more severe cases of depression then the placebo effect would expected to be diminished as 'belief systems' become more and more impaired. The effect of the drug is likely to be clearer. <BR/><BR/>What I think might well be harmful is if the results are taken to mean that me should not give these drugs to milder sorts of depression. In the absence of other therapies then the patient is doubly denied any small effect from the drug and any placebo response. <BR/><BR/>Interesting stuff.Le Canard Noirhttps://www.blogger.com/profile/07565056022495154803noreply@blogger.com