Media Discussion, March 2012

Author(s): CBS News


Treating Depression: "Antidepressants or Placebos?"



I am so glad that there is a discussion on this topic. I watched that 60 min episode as a routine. I have found myself shaking as I cannot believe that Dr. Kirsch could draw a conclusion of placebo effect from ALL antidepressants. As a pharmacologist and psychiatrist, our research and clinical practice have demonstrated benefit for our patients. Even his conclusion is from his DATA, and there was an impression that drug trial data were driven by pharma, it should be reminded that many drug trials initiated and conducted independently from pharma also showed efficacy to depression. I wonder if Dr. Kirsch could have cured depressed patients without any antidepressant. I recall that Dr. Thase mentioned that "antidepressant may not be effective to mild depression, but effective to severe depression". I do not completely agree, as many trials showing effects in patients with mild to moderate depression, and in practice, mild to moderate depression at out-patient settings also have often been effectively treated with antidepressants. We do need new treatment strategies, but it is a shame to consider current antidepressant non-effective. I have not had one single patient questioned to me due to this 60 min episode, but I have prepared to address it if it comes up.
Dr. Xiaohua Li, M.D.,Ph.D.
Posted: Mar 20, 2012 5:04 PM
We all seem to be in agreement that the average drug-placebo difference for antidepressants is small. It is well below the threshold for clinical significance used by the National Institute for Health and Clinical Excellence (NICE), which establishes treatment guidelines for the National Health Service in the UK. Does it make sense to prescribe these medications, despite side effects and various other risks, when other treatments provide the same benefits without the risks? Michael Thase is correct in saying that there may be a substantial effect for a relatively small subgroup of patients. That was the whole point of the meta-analysis conducted by my colleagues and I in 2008 ( Specifically, there seems to be a clinically meaningful effect for patients with scores of 28 or above on the Hamilton Rating Scale for Depression (HRSD). This amounts to approximately 10% of patients diagnosed with major depressive disorder (MDD) in clinical practice. However, cognitive behavior therapy (CBT) also works better on more severely depressed patients, and has the advantage of being less risky. Finally, the maintenance trials discussed by Thomas Laughren leave out an important piece of data. The relapse rate among patients who have improved on placebo and are kept on placebo in continuation trials is as low as that of patients who have improved on drug and kept on drug in maintenance trials (see Andrews et al., Frontiers in Psychology, 2. doi: 10.3389/fpsyg.2011.00159). The Andrews et al. analysis also shows the relapse rate of patients given antidepressants increases as a function of the degree to which serotonin and norepinephrine are enhanced by the medication. Here again, the data on CBT are particularly important, as it capable of reducing relapse rates substantially over follow-up periods as long as six years (Fava et al., 2004, American Journal of Psychiatry, 161, 1872–1876).
Irving Kirsch
Posted: Mar 20, 2012 12:03 PM
I don’t agree with many of the inferences that Kirsch draws from his analyses; to begin with there are severe limitations on the extent to which one can generalize from clinical trials to clinical practice But I do think that Kirsch raises an important and legitimate question about the true value of antidepressants and which depressed patients benefit from them. These questions deserve thoughtful consideration and should not be dismissed out of hand. The idea that antidepressants offer little or no advantage over placebo for patients with the less severe variants of depression is neither new nor the bombshell that Kirsch and Sixty Minutes imply. In 1970, Jonathan Cole wrote: “In my judgement both the antianxiety agents and the antidepressant agents suffer from over promotions . . . On the one hand , a lot of depressions, particularly neurotic depressions, get better anyway, and severe retarded endogenous depressions did well with electric shock therapy before the antidepressant drugs ever came along .. . .On average, the superiority of these drugs over placebo is not impressive . . .” Forty years and at least a billion dollars worth of studies later Cole’s comments-made at the dawn of the antidepressant era-seem eerily prophetic. Prominent among the raft of studies that support this contention is the 1980s NIMH sponsored collaborative treatment of depression study which found no significant difference in outcome for an antidepressant, two types of psychotherapy and placebo in both less severely depressed patients and the entire sample of depressed patients spanning the gamut of severity. It may also be worth noting that placebo is not equivalent to no treatment. In fact, most studies comparing either pill placebo or “psychotherapy placebos” to psychotherapies like CBT, widely believed to be effective treatments for depression, find less of a difference in outcome between psychotherapy and placebo than between drug and placebo. And, as Jerome Frank suggested, the active ingredients of the psychotherapies may be delivered in placebo treatment. Accordingly, one could argue that placebo controlled antidepressant clinical trials are in effect comparing an antidepressant to an alternative treatment widely recognized as effective in all but the severest forms of depression.
Walter A Brown
Posted: Mar 18, 2012 11:56 AM

* indicates a required fieldAdd a Comment

Name *
Email *
Your email address will not be published.

Enter Code Above:
Comment *