Wednesday, March 07, 2007

Antipsychotics for Everything!

It is amazing what passes for evidence these days in psychiatry. A December article in the journal Current Psychiatry is titled “Antipsychotics for patients without psychosis? What clinical trials support”. It’s a doozy. As you can see from the table above, the authors are in favor of antipsychotics for many conditions. More on the table later.

The authors point out a positive feature of second generation antipsychotics (SGAs; Risperdal, Zyprexa, Seroquel, Abilify, Geodon). The sentence reads:

In schizophrenia patients, antipsychotics have been shown to improve psychotic and nonpsychotic symptoms [long list]…, insomnia, poor appetite, …

Sure, it might lead to diabetes, but enjoy your increased appetite in the meantime! Sure, you might feel entirely zonked out for much of the day, but that’s part of improving your insomnia. How’s that for spin?

Another quote:

SGAs do offer clinicians unique tools; no other class of psychotropics can claim efficacy in psychotic disorders, bipolar disorder, depression, and other disorders we describe in this review.

This is quite reminiscent of the attempts of Janssen to label Risperdal as a “broad spectrum psychotropic agent” and of Lilly to label Zyprexa as “the safe, proven solution in mood, thought, and behavior disorders.” The evidence for these medications as treatments for depression is quite weak, especially in the context of their heavy side effect burden.

The article goes on to discuss the “evidence,” nearly all of it quite meager, that allegedly supports the use of SGAs in obsessive-compulsive disorder, PTSD, social anxiety disorder, developmental disorders, and personality disorders. In no case is it ever mentioned how large of an effect these medications have versus a placebo. The authors almost entirely discuss positive findings, no matter how small the positive effects were in the clinical trials they cite. Non-positive findings are rarely mentioned, as is par for the course in most review articles in psychiatry.

So, let’s have a brief review of what the clinical trial data actually say about the use of SGAs for these various nonpsychotic conditions.

OCD: For people with OCD who have not shown much improvement after receiving an SSRI (Zoloft, Paxil, etc.), adding an SGA may bring, on average, a small benefit. On the other hand, the side effect profile, as mentioned earlier, is likely to not be friendly. Assuming that an individual stops taking the antipsychotic at some point, it is quite likely that whatever benefit was gained from the medication will be lost. Nowhere mentioned in the article is the prospect of behavioral psychotherapy, which has a pretty good (though not great) track record in treating OCD without the, um, “improvements” in sleep and appetite brought on by SGAs.

PTSD: Evidence for SGAs so weak it barely merits comment. Please let me know if I’m missing something here.

Depression: Weak evidence at best, as witnessed by the ARISE-RD study regarding Risperdal and a recent study on Zyprexa.

Social Phobia: No kidding. If you give people a major tranquilizer like Zyprexa, they will probably feel somewhat less anxious. Again, check out the side effect profile and determine if a lifetime of Zyprexa makes sense in treating social anxiety. Again, there are nondrug treatments that often work pretty well, so why would someone prescribe an SGA for social anxiety? Oh, and there is very little data to actually support that SGAs actually work for social anxiety.

Borderline Personality Disorder: Again, throwing a big tranquilizer at people who are impulsive may calm them down a bit. But has anyone ever heard of a personality disorder being changed significantly by the administration of an antipsychotic? The evidence is pretty clear that SGAs yield only modest effects for borderline personality.

Alzheimer’s/Dementia: Here’s a quote you cannot miss, because it is featured in white text against a black background on the fourth page of the article:

SGAs remain the first therapeutic option for psychosis and agitation in Alzheimer’s patients.
Yeah, so what if they are linked to a higher rate of death among patients in this group? The authors acknowledge this point, and also acknowledge the findings of a recent study (CATIE-AD) which found poor efficacy for SGAs in Alzheimer’s patients. Yet they make the statement about SGAs being the “first therapeutic option” – oh, to be a “thought leader” and to make such statements…

Back to The Table: Let's come back to the table (top of page) one more time. The authors' definition of SGA uses being "supported" means that on at least one measure (out of several), the SGA was "significantly" better than a placebo.

What if the difference versus a placebo was small? What if patients still had moderate to severe symptoms after taking the SGA? What if the side effect profile was poor? What about unpublished studies that did not yield any positive results? These questions were, of course, not taken into consideration. It's all about cherry picking the most positive findings.

I could write more about this so-called article, but it is making me cranky and causing my head to spin. Fortunately, I’m sure that SGAs are also effective for dizziness and crankiness – perhaps I should go grab a “safe, gentle psychotropic” and take a nap.


Anonymous said...

IMHO the only way in which these drugs help is by knocking people out and that's only for the short term. In the long term these drugs make people worse on all these fronts including psychosis by "sensitizing the brain", causing chemical and morphological changes that when the drug is removed make things a lot worse than they ever were to start with (according to Grace Jackson and others who have researched the matter). So on that point it would be interesting to know if the clinical trials upon which all these claims are based are mostly short term trials -- (to say nothing of pointing out that the way SGA trials have traditionally been conducted is by taking treated psychotic patients, withdrawing them abruptly from their current treatment and then putting some on placebo and some on SGA -- now it's pretty obvious which arm is going to do better in that scenario -- withdrawal and recovery from withdrawal takes months, if ever, not ten days or two weeks).

CL Psych said...


The trials are almost always short-term. When they stretch them to show "good" long-term results, it is by, as you mentioned, switching some to placebo and keeping some on meds. The withdrawal effects are then interpreted as "the disease" returning. The funny thing is that while this is bleedingly obvious to you and I, most "key opinion leaders" in psychiatry don't seem to understand the idea.

Monica Cassani said...

Grace Jackson is one of my heroes. I'm withdrawing from multiple neuroleptic among other drugs and she helped me tremendously by speaking to my psychiatrist. He was clueless as to what I was taking on, but has become very helpful once being enlightened by Grace. I recommend her book, "Rethinking Psychiatric Drugs: A Guide to Informed Consent" to anyone dealing with these drugs in any way...whether they are taking them or prescribing them.
thanks for bringing her up!