Should we accept the analysis of a journalist who has not treated a patient or implemented a study and reaches conclusions that run counter to well-established practice guidelines?
I’ve been thinking about the “occupy” movement lately. It occurs to me that we've witnessed it through our annual APA meetings when sign-carrying people (Scientologists?) protesting the use of ECT declared, “Psychiatry Kills.” I simply ignore them because I don’t really “get” what they are communicating.
More recently, I have been witnessing a different attempt to occupy psychiatry led by journalist Robert Whitaker, the author of Anatomy of an Epidemic. Whitaker has articulated a position that I think we should not ignore.
Whitaker's polemic, which combines ideology with evidence, reaches the conclusion that a marked increase in disabled mentally ill persons in the United States over the past 20 years is due to psychiatric medications. He writes hundreds of pages describing select studies containing hints that after antidepressants, antipsychotics, and other psychiatric medications are used to treat the acute presentations of illness, the medications backfire via specific mechanisms of action that result in worsening of illness and increased disability claims. The ideological part of his argument assumes that his hypothesis has been a closely guarded secret among researchers, clinicians, organized psychiatry, and the pharmaceutical industry. Clinicians without industry ties are not immune from this conspiracy; they are presumably motivated to employ these medications for revenue.
The psychopharmacologist in me prompted a 2-part response to Whitaker's book in a journal for which I had been writing a monthly column.1,2 I had become aware of an interview that the journal had published in which the authors (non-psychiatrists) accepted Whitaker's research efforts outright and used his work to highlight criticisms (many justified) about the way psychiatry delivers care to patients. I was also aware of Marcia Angell's embarrassingly uncritical review of Anatomy of an Epidemic in The New York Review of Books.3
The fatal flaw in Whitaker’s work is his conclusion that psychiatric medications have caused a rise in disability claims. Using garden-variety journalistic reductionism, Whitaker ignores or is unaware of the complexity that accounts for psychiatric disability. To prove that drugs cause disability, multiple factors need to be considered. Study designs and statistical analyses need to account for such factors as confounding, reverse causation, and bias. It is equally possible that the reported increase in psychiatric disability claims results from many sociocultural stressors that have emerged over the time frame in question, eg, increased separation of wealth, globalization and resultant loss of jobs, reduced stigma against psychiatric illness, change in definitions of disability, and shifts in service delivery policies.4
Is the prescription of psychiatric medications (event A) causally related to the rise in disability claims (event B)? Only journalistic speculation coupled with ideological fervor would reach such a conclusion. A properly controlled study would help identify the contribution that psychiatric medications may have made to disability claims relative to the sociocultural factors that I mention. I would love to see a retrospective multivariate cohort study of the association among bio-psycho-social factors and psychiatric disability claims between 1990 and 2010. Mark Olfson, get to work.
Another point to consider here: If the increase in disability claims were due to psychiatric medications that caused an increase in the occurrence and severity of illnesses such as schizophrenia and depression, as Whitaker claims, we’d expect to see supportive epidemiological evidence. In fact, available studies show no real change in the rates or morbidity of serious mental illnesses.5,6
In the absence of epidemiolog-ical support, Whitaker chooses to doggedly pursue the highly speculative notion that medications cause psychiatric disability by reviewing select psychopharmacological and neuroimaging studies that he thinks might explain the notion. This is where it gets sticky and cumbersome.
The studies he selects, which are legitimate efforts by reputable researchers, involve efforts to examine the long-term effects of psychiatric medications. The average practicing clinician (not to mention the layperson) is not prepared to critically assess Whitaker’s presentation.
Whitaker’s descriptions of the studies he selected are fairly accurate, but his interpretations stray from the level of available evidence and prevailing expert opinion. He consistently comes to the same erroneous conclusion: that psychiatric medications cause long-term morbidity. Over hundreds of pages, Whitaker presents these studies as though he has uncovered clues to a mystery . . . and has solved it. He covers theories such as tardive psychosis, aka dopamine supersensitivity and tardive dysphoria, and treats them more like proven realities than the speculations that they are.7,8 Thus far, concepts such as these have not swayed prevailing opinion, despite considerable academic consideration.
The studies that Whitaker incorporates in his analysis are too severely limited-when considered individually or en masse-to justify reaching such a definitive, general conclusion. They involve small samples of fewer than 1000 subjects, do not employ randomized designs, and are mostly conducted at single centers, which make them difficult to compare or generalize. Of the studies considered, I’ll focus on just one for the purpose of characterizing Whitaker’s analytical style.
Whitaker refers to the research of Harrow and Jobe9 in his contention that long-term antipsychotic treatment causes increased pathology. The Harrow and Jobe study is a marvelous 15-year follow-up of 64 schizophrenic patients-not an easy study to perform, in fact, a true labor of love. Twenty of these patients did not receive antipsychotic medications, and the researchers noted that their outcomes were more likely to be better than what was seen in the 44 patients who received antipsychotics.
Is it the disease or the treatment?
Whitaker’s conclusion is that long-term use of antipsychotics is not necessary for or may be harmful to patients’ functioning. He dismisses the fact (brought out clearly by the study authors) that the untreated patients in this nonrandomized observational study were likely to have been self-selected (ie, they were healthier and elected not to take antipsychotics). Whitaker reverses cause and effect to sustain his argument. Standard clinical logic would conclude that the poor outcomes in the antipsychotic-treated patients were a function of their illness-not the treatment. (Interested readers may want to see E. Fuller Torrey’s excellent analysis of Whitaker’s “tortured logic” in the schizophrenia studies.10)
The Harrow and Jobe study, by virtue of its design, is severely limited in its ability to generate clear and conclusive evidence that antipsychotics cause or increase morbidity over time. Ask Harrow and Jobe if you don’t believe me. The value of the Harrow and Jobe study is at the level of hypothesis generation. It suggests that not all schizophrenia patients require long-term medications. The question is, which ones?
The same type of discussion can be applied to Whitaker’s treatment of antidepressants, antianxiety agents, and mood stabilizers. He consistently employs a reductionist mentality to reach the “astonishing and startling” (Amazon.com’s advertisement words for the book) conclusion that our drugs have caused an “epidemic” of increased psychopathology in the major psychiatric disorders.
What to make of all this?
Should we accept the analysis of a journalist who (1) to my knowledge, has not treated a patient or implemented a study and (2) reaches conclusions that run counter to well-established practice guidelines? Whitaker’s ideological viewpoint, which is implied throughout the book, is that our guidelines are inaccurate and driven by industry and our own need for income-that we are dishonest brokers. Beauty is in the eye of the beholder.
I would predict that if the authors of the studies on which Whitaker relied convened a meeting to critically discuss the book, the vast majority would vote against this journalist’s idiosyncratic conclusion.
No one will argue that there is a dearth of long-term studies, defined as longer than 2 years, of the enduring effects of psychiatric medications. If Whitaker’s work stimulates more long-term research, then it serves the field. But while we wait for that research, should we change our practice to fit this journalist’s thesis that our medications cause poor outcomes? My answer is no.
Occupying the doctor-patient relationship
As practitioners, we work with our patients and provide for them the best treatment that we can, based on their preference, available evidence, and our judgment. We work collaboratively with each patient to weigh the risk to benefit ratio of long-term prescriptions over time. Risk to benefit calculations are no easy task. It is difficult to identify whether one patient requires long-term treatment following a good response to short-term treatment. Ideally, these decisions proceed within a solid alliance that we establish with the patient and others. Successful alliances require trust.
Books (and blogs) that go out for public consumption alleging conspiracy among doctors that results in epidemics and offering a prejudiced review of select literature run the risk of distorting that trust. (I am interested in hearing about direct experiences that you may have had in which articles, blogs, or books of the kind that I have discussed have altered existing treatment plans-positively or negatively. Contact me at firstname.lastname@example.org or 305-293-3555.)
We and our patients do not need to have that trust hijacked by misguided, “astonishing and startling” conclusions presented in publications such as Anatomy of an Epidemic. It is our obligation to protect that trust.
1. Glazer WM. Rebuttal: questioning the validity of Anatomy of an Epidemic (Part I). Behavioral Healthcare. October 31, 2011. http://www.behavioral.net/article/rebuttal-questioning-validity-anatomy-epidemic. Accessed June 28, 2012.
2. Glazer WM. Rebuttal: questioning the validity of Anatomy of an Epidemic (Part II). Behavioral Healthcare. November 15, 2011. http://www.behavioral.net/article/rebuttal-questioning-validity-anatomy-epidemic-part-ii. Accessed June 29, 2012.
3. Angell M. The epidemic of mental illness and the illusions of psychiatry (2-part review). New York Review of Books. 2011.
4. Wilkinson RG, Pickett K. The Spirit Level: Why More Equal Societies Almost Always Do Better. London: Allen Lane; 2009.
5. Kessler RC, Demler O, Frank RG, et al. Prevalence and treatment of mental disorders, 1990 to 2003. N Engl J Med. 2005;352:2515-2523.
6. Substance Abuse and Mental Health Services Administration. Mental Health, United States, 2010. Rockville, MD: Substance Abuse and Mental Health Services Administration, US Dept of Health and Human Services; 2012. HHS publication (SMA) 12-4681.
7. Fava GA. Do antidepressant and antianxiety drugs increase chronicity in affective disorders? Psychother Psychosom. 1994;61:125-131.
8. El-Mallakh RS, Gao Y, Roberts JR. Tardive dysphoria: the role of long term antidepressant use in inducing chronic depression. Med Hypotheses. 2011;76:769-773.
9. Harrow M, Jobe TH. Factors involved in outcome and recovery in schizophrenia patients not on antipsychotic medications: a 15-year multifollow-up study. J Nerv Ment Dis. 2007;195:406-414.
10. Torrey EF. Anatomy of a nonepidemic-a review by Dr Torrey: How Robert Whitaker got it wrong. Treatment Advocacy Center. 2011. http://www.treatmentadvocacycenter.org/index.php?option=com_content&task=view&id=2085. Accessed June 28, 2012.