</font><blockquote><div id="quote"><font class="small">Quote:</font>
Some of the newer meds are causing problems--big problems in some patients. You may have heard of some anti-depressants such as Zoloft, Paxil, Wellbutrin, and that popular one with a "P" that got a bad name too. Causes some peeps to think or commit suicide. I don't think they do enough research on the newer ones and are too eager to give to patients because they are the patients are being used as guinea pigs.

</div></font></blockquote><font class="post">

There is a lot of misinformation in the press about what is really going on. I guess it sells newspapers if you come down on the drug companies. If you look at the underlying data, as I have done, the picture is not nearly so bad as we have been led to believe.

An example. Paroxetine (Paxil) was said to cause a seven-fold increase in suicidal ideation and suicidal acts in adolescents. That is actually not even close to true. The full clinical trial data, including individual patient reports, was made available on the drug company website. So I read it. All 500+ pages of it.

There was no adverse effect reporting category for suicidal acts, or thinking. They had a category for "emotional lability", so reports of a suicidal nature, and others unrelated to it, went there. For each of these incidents, complete records were included in an appendix.

At the outset, it was apparent that there were seven times more reports of emotional lability in the paroxetine group, compared to the placebo group. However, only about a half of those in the paroxetine group were considered "moderate to severe". And when you hear what was described in those, you'd have to wonder what might have been in the "mild" reports. Anyway, we're already down to four to one, and we haven't even looked at what happened yet.

I'll give two examples from the Paxil group. One girl quit taking Paxil, her mother gave her another drug for two weeks, then she overdosed on Tylenol. The report was coded (by a blind observer) to indicate no link to the drug being tested (correctly, as far as I'm concerned).

Another example was one girl in the study who couldn't count her meds properly. Participants were given foil-pack meds, twenty at a time, but told to take one pill twice a day. They were to turn in unused meds each week, when they were given the next week's meds. This one girl returned 1-2 fewer pills than she should have, three weeks running. She said she had trouble remembering if she took her pill or not. No adverse symptoms were reported, but because these incidents had occurred in consecutive weeks, they were defined in the study protocol as "intentional overdose". I don't think that fairly describes what happened here. If she had done the same thing every other week, there would have been no report filed.

In contrast, that incident in the placebo group, was a gravely serious incident. After being rushed to the emergency wing with severe blood loss, unconscious, this subject's blind was broken to determine if she was on any medication. She was not, and she was admitted to an inpatient psych facility for followup care. This was the only real suicidal event, and the only incidence where a blind was broken.

What I saw was not a seven-fold increase in suicidal ideation and/or acts over placebo, but instead, I saw no suicidal acts in the paroxetine group (attributable to paroxetine), and one in the placebo group.

A recent re-analysis by the FDA was conducted, to see if a "signal" for suicidal ideation or behaviours was "hidden" in the clinical trial data. I reviewed the actual reports the FDA panel considered. Even when they lumped all adolescent clinical trial data together, all the SSRIs as if they were a single drug, the results did not show a significant difference between drug and placebo. Only when they included venlafaxine (Effexor), which isn't an SSRI, and the data for trials not related to depression, did they get a significant measure on one estimate (but not on another supposedly identical estimate). Even still, the difference in suicidal acts or ideation was 1.7. That is a relative risk estimate, a comparison between drugs and placebo. If you believe that the suicidal gesture rate in the placebo group was very low (and it was), then taking 1.7 times that very small number is still a very small number. We're not talking about a huge surge here. It was so small that it could only be found by taking extreme liberties with the data. The assumptions required to perform the meta-analysis, and the equivocal results (barely significant and non-significant results), argue against concluding that SSRIs are a serious risk.

It seems to me the FDA was reacting to public pressure far more than they were making a scientific decision. All antidepressants *ever* used have the unfortunate side effect of potentially inducing suicidal behaviour early in treatment. The rate of that induction is unchanged across MAOIs, tricyclics, SSRIs, and the atypicals (with a few exceptions such as Wellbutrin). What is astounding to me is that anyone might have thought SSRIs would be different. One motive for developing them in the first place was to find a non-toxic-in-overdose replacement for the toxic tricyclics.

I think everyone should read everything they can about their drugs before taking them, but in some cases (such as severe depression), I can't imagine that someone might make sense out of the information in the drug monograph. Still, you can go to http://www.rxlist.com/ and read the same detailed information that's in the PDR. You'll find that detail linked to the generic drug name, not the brand name, though.

Lar