The American Journal of Psychiatry (AJP) has not published a single letter to the editor regarding a highly publicized study by Dr. Gregory Simon, et al, "Suicide Risk During Antidepressant Treatment," claiming that data from this uncontrolled epidemiological study demonstrated that antidepressant treatment reduced suicide attempts. See: AJP,163:1 January 2006 1-6.
(Of note, the online title is different, suggesting a bit of editorializing propaganda: "Is the FDA Warning About Antidepressants Wrong?")
The study has been widely criticized. The director of the National Institute of Mental Health acknowledged the Simon study provides no valid information: “Since there was no placebo control group, we do not know how many of these patients would remit without active drug treatment.”
Other critiques have been made on both sides of the Atlantic: ours is at: http://www.ahrp.org/cms/content/view/13/28/ ; another by Charles Medawar of Social Audit (UK), who corresponded with Dr. Simon seeking clarification about inconsistencies in the published report is at: http://www.socialaudit.org.uk/6060521.htm#ANTIDEPRESSANTS;
a synopsis at: http://www.ahrp.org/cms/content/view/174/94/
But not a single letter or commentary sent to the AJP about the study has been published.
Below is a thougtful letter to the editor by Ms. Sara Bostock in which she has identified an important bias in the Simon study that makes antidepressants appear to be effective in preventing suicide attempts. She states that: "Choosing as a benchmark the point in time when antidepressant therapy commences is artificial."
In other words, there will be a high rate of suicide attempts just before a prescription for antidepressants because a suicide attempt will lead to antidepressant therapy.
Similarly, the rate of ear infections will be very high in the weeks before getting an antibiotic than during the weeks after getting an antibiotic, even though antibiotics have, if anything, only a modest impact on ear infections.
A more appropriate strategy would be to exclude the "index" period that led to antidepressant therapy, and then compare the rates of suicide attempts before and after the index period.
As shown in figure 6 by Simon et al, their data do not indicate lower rates of suicide attempts after starting antidepressants. See: http://www.ahrp.org/files/ssri06/Simon2.ppt
To the contrary, for the largest time window shown of up to 12 weeks, the rates of suicide attempts are actually GREATER during the period AFTER starting antidepressant therapy than during the period before starting therapy. Indeed, Simon et al acknowledge: "Our data certainly do not exclude the possibility that antidepressants may precipitate increased suicidal ideation or attempts…"
It is misleading, therefore, to suggest that these data demonstrate effectiveness of antidepressants in preventing suicide attempt. We know from controlled clinical trials that antidepressants increase the risk of suicide attempt, and the data from Simon et al. are consistent with this hypothesis.
One would think that Ms. Bostock’s insightful explanation of how the interpretation of the published data is misleading would be worthy of some deliberation. Unfortunately, the readers of the American Journal of Psychiatry will not get the chance to consider her comments because the editor, Dr. Robert Freedman, declined to publish Ms. Bostock’s letter apparently for fear of ruffling feathers (see correspondence below).
His stated reason is that the letter "disparages" those who do these kind of studies. Even if this statement were true–which is not the case–the intellectual value of Ms. Bostock’s comments would outweigh the need to protect the fragile egos of (presumably) seasoned professionals. But Ms. Bostock’s letter commented on the authors’ methods and interpretation of the data, which she demonstrated was clearly slanted in favor of the drugs when the data are at best ambiguous.
Perhaps Dr. Freedman is censoring letters (and critical articles) that pose a threat, not so much to authors’ feelings as they threaten the conventional practice of modern psychiatry and the marketing interests of the journal’s primary advertisers–the drug manufacturers.
Thus, at every opportunity, Dr. Freedman sings the praises of drugs advertised in the AJP. Indeed, even as research findings contradict those praises. For example, Dr. Freedman’s editorial in the New England Journal of Medicine, accompanying the CATIE (schizophrenia outcome) study which sent shock waves throughout biological psychiatry as the patients rejected currently prescribed drugs: "Nearly three-quarters (74%) of the patients discontinued the study medication before 18 months."
Indeed, the CATIE study authors were shaken by their own finding. Dr. Jeffrey Lieberman, the principal investigator, and chairman of psychiatry, Columbia University stated:
"The biggest surprise of the study was that the older medication, perphenazine, was comparably effective to at least three of the new medications and not much worse than the new drug that did the best–olanzapine."
Yet, Dr. Freedman claimed the opposite: "the introduction of antipsychotic medications has had a profound social impact: today the vast majority of patients with schizophrenia are able to live in their communities as opposed to institutional settings. Treatment of schizophrenia has come a long way…"
Dr. Freedman should be reminded that scientific knowledge depends on critical discussion and debate between competing ideas–not merely reiteration of accepted dogma. Thus, to censor such debate arrests learning and undermines the essence of the academic pursuit of knowledge and perpetuates outmoded thinking to the detriment of patients.
By censoring opposing views instead of airing them, Dr. Freedman has revealed the he and the American Journal of Psychiatry are more interested in PROMOTING a particular point of view than they are in fostering critical discussion. In doing so, Dr. Freedman reveals his agenda to be one not of science but peddling propaganda.
The public has a right to hear the information that the "authorities" do not want you to hear.
**NOTE: Am taking a break–there will be no AHRP Infomails and nr veracare e-mails for two weeks–**
Contact: Vera Hassner Sharav
RE: "Suicide Risk During Antidepressant Treatment" Am J Psychiatry 163:1 January 2006 41-47
To the Editors:
This article published in January generated a lot of publicity in support of the safety of antidepressant treatment as well as implicit criticism of the FDA black box warnings. As someone who testified at the FDA open public hearings in support of those warnings and who lost a beloved daughter to a paroxetine induced suicide in the early weeks of her treatment, I would like to register my objections to the methodology and conclusions drawn in this study.
The researchers betray their bias towards the merit of antidepressant treatment by stating that the warnings “do more to discourage effective treatment than to improve the quality of follow-up care.” This implies that effective treatment must include an antidepressant, yet the evidence for the safety and efficacy of SSRIs is not as sound as psychiatry or pharmaceutical manufacturers would have us believe.
Choosing as a benchmark the point in time when antidepressant therapy commences is artificial. Antidepressants are the treatment of choice for suicide attempters so naturally this is a point in time at which attempts will be high. The fact attempts are less after treatment begins says nothing about the causal role of antidepressants in that decline because there is no comparison with any other form of treatment. Among the group that had already attempted suicide, the rate surely fell sharply in the first month as someone who has attempted suicide, failed and sought help (of any kind) is unlikely to try again so soon. In fact the rate in the first month is significantly higher than it was in the second and third month before treatment, times at which the rate of attempts might be more in line with the normal rate in a depressed population. It is not clear from this study that antidepressants actually cause a decrease in suicide attempts below the rate in a drug naïve population.
The reasoning behind excluding patients with any diagnosis other than major depressive disorder, dysthimia or depressive disorder is unclear since it is known that the risk of suicide is higher in those diagnosed with anxiety than depression. Also antidepressants are frequently a treatment of choice for bipolar depression so why was this diagnosis ruled out? The group also excluded those who had any previous antidepressant prescription in the past six months but failed to clarify how many in the group were on other psychotropic drugs such as stimulants. This is a relevant variable. The absence of any narratives to corroborate the statistical impressions weakens the conclusions.
The study seems more concerned with protecting the blockbuster status of antidepressant drugs than with a real understanding of variables that affect suicidality. It is constrained and manipulated to get desired results. Until doctors really start to listen to patients and make a true effort to distinguish between adverse effects, chemical dependency, drug withdrawal, and underlying suffering, more vulnerable individuals will be put at risk with a paradigm of care based on weak assumptions about these drugs.
Dear Mrs. Bostock,
I have had the opportunity to re-read your letter along with others on our editorial staff in our review process. Dr. Simon and the other authors of this article share with you a desire to decrease the tragedy of suicide. We publish articles such as his because they contribute to the knowledge base of physicians who work with potentially suicidal persons. All of us share a commitment to examine data on the pros and the cons of specific treatments, both pharmacological and non-pharmacological interventions. The tragic fact that persons like your daughter commit suicide despite everyone’s best efforts gives this work paramount importance. We considered your letter because it lets us know what we unfortunately already know too well, that current efforts are not enough. However, some of the paragraphs of your letter disparage those who do the very kinds of work that you and I know is needed. Therefore, we have decided not to publish your letter. Be assured, however, that the Journal will continue to publish studies about the treatment of potentially suicidal persons, highlighting both successes and failures. Thank you for submitting the letter.
Robert Freedman, M.D.
American Journal of Psychiatry