Posted by Iansf on January 8, 2005, at 1:08:45
In reply to Re: SSRIs vs Placebo.... PLEASE READ » ed_uk, posted by SLS on January 3, 2005, at 9:20:20
> Hi Ed.
>
> > Using the 'bar' on the right hand side of the page scroll down to page 165. Have a look at how the efficacy of paroxetine (Paxil/Seroxat) compares to placebo. The results of various trials are presented in a table. As you can see, paroxetine is only slightly more effective than placebo. The MDD studies were especially disappointing.
>
> And how do YOU interpret all of this?
>
>
Several things are not taken into account with many, if not most, psych drug studies. First, just being in a study, getting regular care, being paid attention to and doing something "important" can give many people a temporary boost. I suspect that once a study ends, many of the placebo takers who reported relief from depression will sink back down.Second, if you ever try to get into a study you'll discover that most long-term sufferers are excluded. Most particpants are people reporting a first depression - which tends to mean people with situational depression who are likely to see the depression lift fairly quickly anyway. Also, many others would be more appropriately categorized as dysthymic., and in general dysthymia appears to be less responsive to antidepressants than full-fledged depression is.
Third, antidepressants often take much longer than than 8 weeks, the typical length of most studies, to take full effect.. If studies lasted 16 weeks or longer, the number of positive responders to the med would likely increase while the number of positive responders to the placebo would remain unchanged -- and possibly even drop..
Fourth, people taking the medication are almost certain to experience more (and more problematic) side effects, which can be discouraging and counteract the lift given by the drug. Yet with time some side effects would lessen, allowing the positive feelings created by the drug to become more evident.
Finally, there are shortcomings of statistics that come into play. For example, take 100 people and give them Paxil and another 100 and give them a placebo. Maybe 15 respond in each group. Now take all 30 responders out. Give Wellbutrin to the remaining 85 drug testers but continue giving a placebo to the 85 remaining in the control group. Maybe you get another 8 responders with Wellbutrin. But you get no more responders from the control group because their "treatment" hasn't changed -- those who were going to respond to placebo already have. Now take the Wellbutrin responders out and put the remaining 77 drug testers on Parnate. Perhaps you find another 5 responders. In the control group you still get no more responders. Keep trying different meds on the remaining nonresponding drug testers and eventually you might have 60 or 70 who find an effective drug. But the number responding to placebo would never go above 10. Thus you get: antidepressant effectiveness, 60-70%; placebo effectiveness, 10%.
poster:Iansf
thread:437036
URL: http://www.dr-bob.org/babble/20050103/msgs/439277.html