Psychiatrists and a popular science writer have collided over claims made
in a recently published book that revives the controversy about a 30-year-old
study questioning the validity of psychiatric diagnoses.
The controversy, which has been brewing for nearly two years since the book
was published, has more recently made its way into the scientific literature
in a remarkable exchange between the combatants in the Journal of Nervous
and Mental Disorders.
Robert Spitzer, M.D., of Columbia University College of Physicians and
Surgeons, and other psychiatrists and psychologists have called into question
claims made by science writer Lauren Slater in her 2004 book, Opening
Skinner's Box: Great Psychological Experiments of the 20th Century.
In that book, Slater describes a journalistic exercise in replicating an
experiment made famous by David Rosenhan in 1973 in which he enlisted healthy
individuals with no ostensible history of psychiatric diagnosis to pose as
patients hearing hallucinatory voices. Rosenhan's account was published in
1973 in Science under the title "On Being Sane in Insane
Places." All of the pseudo-patients were admitted to psychiatric
hospitals, and all but one were diagnosed as having schizophrenia; the
pseudopatients, who claimed to have stopped hearing the voices after
admission, were diagnosed at discharge with schizophrenia in remission.
In Opening Skinner's Box, Slater recounted the Rosenhan study and
his conclusion that psychiatric diagnoses are products of social context
rather than objectively verifiable scientific criteria, as well as the
enormous impact that the study made on the psychiatric profession and public
perception of the reliability of diagnosis.
Slater also reported the results of her own exercise in posing as a
patient: At nine separate emergency rooms, Slater said she presented herself
as a patient seeking treatment because of a voice she was hearing that was
saying the word "thud" (one of the same words Rosenhan's
pseudo-patients claimed to hear). She further reported being prescribed 25
antipsychotics and 60 antidepressants and said that she was diagnosed"
almost every time" with psychotic depression.
"It becomes fairly clear to me that medication drives the decisions,
and not the other way around," Slater wrote. "Rosenhan's point
that diagnosis does not reside in the person seems to stand."
Spitzer, whose conversations with Slater are quoted in the book, is
skeptical.
"When I read it, it just didn't make sense," he told
Psychiatric News. "Someone presents in the emergency room only
hearing the word `thud' and moreover denies being depressed—there's no
way that someone is going to say that is psychotic depression. You don't make
a diagnosis of psychotic depression on the basis of hearing the word
`thud.'"
Spitzer, who was editor of DSM-III and has been an important
figure in the evolution and refinement of psychiatric diagnoses, said he
requested to see evidence of Slater's experiment—including files from
her visits to the ERs or names of psychiatrists or hospitals—but that
she has not produced them. And he told Psychiatric News he is highly
skeptical about the medications Slater reported she was prescribed.
Spitzer is not alone in his reactions. "It would be nice if she could
assuage my concerns about her study by providing us with objective evidence
that she did it," psychologist Scott Lilienfeld, Ph.D., told
Psychiatric News. "It would be comforting if she could provide
us with files from her visits to the ER, the name of one attending
psychiatrist she saw, or the name of hospital she visited. She hasn't done any
of that.... I don't know what happened. But I don't think we should take it
seriously until she can provide us with documentation."
Lilienfeld is an associate professor of psychology in the Department of
Clinical Psychology at Emory University in Atlanta.
Others are less circumspect. "I believe the data were
fabricated," psychiatrist Mark Zimmerman, M.D., told Psychiatric
News. "As a researcher, I know that if anyone has a question about
my data, how easy it is to produce data files, and Slater hasn't done
that."
What Slater has done is to offer a response to Spitzer and colleagues in
the November 2005 Journal of Nervous and Mental Disorders.
In what must surely be one of the more unusual exchanges to take place in a
scientific venue, Slater says what she described in her book is not a"
study" and cannot be critiqued as such.
"The book features not only the experiments and the experimenters,
but me, as well, and my husband, and my daughter and my pet raccoon,"
she wrote. "In any case, such a colorful cast of characters and deeply
personal details.. .are in and of themselves enough to make abundantly obvious
to any and all readers that we are not here dealing with an academic inquiry,
or `study.' That Spitzer et al. have chosen to label my work as a study is a
silly and troubling mischaracterization.... The authors also fully understand
that my use of the word `experiment' is of course vernacular, as in, `honey,
let's experiment with this recipe tonight.' It would be far more useful, and
appropriate, if the authors would drop the bloated and obfuscatory remarks,
hunched and hiding as they are behind the veneer of `science,' and claim their
point in plain language. Their point, as far as I can see, is that they think
I lied, which probably would not bother them so much if my alleged `lie' did
not result in narcissistic injury."
Reached at her home, Slater told Psychiatric News that she had no
more comment on the controversy beyond the response she published in the
journal. A spokesperson for the book's publisher, W.W. Norton, also declined
to comment.
In that same journal, Spitzer, Lilienfeld, and Michael B. Miller, Ph.D.,
published a study in which they presented 74 emergency room psychiatrists with
a case vignette modeled on the description in Slater's book and asked them a
series of questions regarding diagnosis and treatment recommendations. In
contrast to what Slater reported, just four psychiatrists offered a diagnosis
of psychotic depression, and only a third recommended medication (see
box).
"Our study raises questions regarding Slater's results and
conclusions, and provides scant support for the claim that psychiatric
diagnoses are mostly the product of fashion or fad, as claimed by
Slater," they concluded.
They also published a rejoinder to Slater titled "A Response to a
Nonresponse to Criticisms of a Nonstudy: One Humorous and One Serious
Rejoinder to Slater." Finally, Zimmerman wrote a review titled"
Pseudopatient or Pseudoscience: A Reviewer's Perspective."
Spitzer and others said that the issue is a serious one involving not only
the veracity of an acclaimed science writer, but the trust of patients and
would-be patients in their doctors.
Spitzer was critical of the original Rosenhan study and published a
critique that appeared in the April 1976 Archives of General
Psychiatry. But he has acknowledged the profound impact it had on popular
perceptions of psychiatry.
Thirty years later, when the field has made enormous strides toward
refining diagnostic criteria, Slater's work threatens to undermine patients'
faith, he said. "There is a real possibility that there will be patients
who read this book and wonder if they can depend on their doctor's
diagnosis," Spitzer said. "She's resurrecting a flawed study that
will be used to make fun of psychiatry and affect the public image of the
profession."
Lilienfeld agreed. "This is a popular book and one that may well be
read by a lot of people in the general public," he said. "In this
chapter [on Rosenhan] she argues forcefully that psychiatric diagnoses are
arbitrary and driven mostly by medication choices. She communicates very
strongly the impression that much of psychiatry is pseudo-science. She doesn't
use that word, but any reasonable person would come away with the impression
that psychiatric diagnosis is much more of an art than a science, and that
psychiatric diagnoses are entirely a product of context.
"The book may lead people to take psychiatric diagnoses less
seriously," he said. ▪