On 29 September Stephen Black wrote:
> Another detail is that they didn't include an important control group 
> on the grounds of cost and artificiality. This would be a group 
> receiving CBT and a placebo, not Prozac. But without that group, 
> there's no way to tell whether the Prozac was having a drug or a 
> placebo effect. Given the faith most of us have in the power of 
> pills, I think it's at least possible that giving CBT with a big fat 
> sugar pill and strong suggestion would work as well as CBT plus 
> Prozac without the risk of adverse effects. In fact, it raises the 
> interesting question whether all behaviour therapy would go better 
> with a placebo pill [I'll leave the ethics of doing this for another 
> discussion].

I agree with Stephen that it would have been valuable to study the effects
of CBT with a placebo, but not when he writes says that giving CBT with a
sugar pill would be �without risk of adverse effects�. It may well be with
some severely depressed patients that there would be adverse effects of
not prescribing Prozac (or another SSRI), namely the continuation of
excruciating symptoms beyond what might have occurred had Prozac been
used. (It�s the nature of the situation that the discussion here deals
with hypotheticals.) I appreciate that this is not what Stephen meant by
adverse effects, but it is something whose seriousness may not be fully
appreciated. The continuation of such symptoms may seem like a �neutral�
result to the observer, but for someone living through the nightmare it
counts very much as an �adverse� result of having sugar pills rather than
an SSRI that might have reduced the symptoms. Please note that I am making
no assumptions here about the efficacy or otherwise of SSRIs, nor arguing
against the additional component to the study that Stephen suggests, only
noting that his saying that the sugar pill alternative has no risk of
adverse effects would ring hollow to someone suffering from severe
depression if there were no relief of symptoms.

On Stephen�s remark that �it raises the interesting question whether all
behaviour therapy would go better with a placebo pill�, it may well be a
good idea to include this component among the alternatives in many
studies. But it should be remembered that CBT is used for a number of
rather different mental disorders/illnesses. Would it be sensible to treat
someone with bipolar disorder, when suicide is a very real risk in the
depressive phase, with CBT plus a placebo when there is good evidence that
lithium is effective in controlling the symptoms?

On the whole issue of the effectiveness of SSRIs, the publicity tends
revolve around those who argue for their deficiencies (of which no one
would deny to some degree or other). The critical literature on the
studies in question tends to get less publicity. See, e.g,  F. M. Quitkin,
J. G . Rabkin, J. Gerald, J. M. Davis, and D. F. Klein:
�Validity of Clinical Trials of Antidepressants�
Am J. Psychiatry 157: 3, March 2000, pp. 327-337

Abstract

Objective: Recent reports have criticized the design of antidepressant
studies and have questioned their validity. These critics have concluded
that antidepressants are no better than placebo treatment and that their
illusory superiority depends on methodologically flawed studies and biased
clinical evaluations. It has been suggested that the blind in randomized
trials is penetrable-since clinician's guesses exceed chance-and that only
active placebo can appropriately camouflage the difference between drug
and placebo response. Furthermore, evidence has been cited to suggest that
psychotherapy is as effective as antidepressants in both the acute and
maintenance treatment of depression. These positions are often accepted as
valid and have been broadly discussed in both the lay press and scientific
literature. The purpose of this review is to reassess the cited data that
support these assertions. Method: The authors examined the specific
studies that were cited in these reports, evaluated their methodology, and
conducted aggregate analyses. Results: Analyses of the original sources
failed to substantiate 1) that standard antidepressants are no more
effective than placebo, 2) that active placebo offers an advantage over
inactive placebo, or 3) that substantial evidence of a medication bias is
suggested by raters' treatment guesses exceeding chance. The authors also
note that some researchers have suggested that the interpretation of
psychotherapy trials can be complicated by "allegiance effects."
Conclusions: The issue of bias or allegiance effects for both
antidepressant and psychotherapy research is real. Investigators of all
orientations must guard against potential bias. However, studies cited as
supporting the questionable validity of antidepressant trials fail upon
closer examination to support assertions that these trials are invalid.

There�s also an interesting article on meta-analyses that I came across
while Googling, in Prevention & Treatment, Volume 1, Article 0006c:
http://www.journals.apa.org/prevention/volume1/pre0010006c.html 
�Listening to Meta-Analysis but Hearing Bias�
Donald F. Klein, M.D. 
Department of Psychiatry, Columbia University,
Department of Therapeutics, College of Physicians and Surgeons, and the
New York State Psychiatric Institute

One criticism of meta-analyses that I recall is that they don�t
distinguish between the prescribing of SSRIs to severely depressed people,
for whom the potential benefits are more clear-cut, as against that for
the moderately depressed.

Two things seem clear to me. The intricacies of studies and meta-analyses
on anti-depression drugs are immense, and not easy for the layperson to
navigate. And the current state of our knowledge does not warrant the
notion that such drugs are no better than sugar pills.

Allen E.

---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to