Recent book reviews by Marcia Angell, MD1—the highly esteemed, former editor in chief of The New England Journal of Medicine—have stirred great controversy within psychiatric circles, as evidenced by the heated exchange of letters published in The New York Review of Books.2-4 Perhaps—to paraphrase a line from the film Cool Hand Luke—“What we got here is a failure to communicate!”
Certainly, when someone of Dr Angell’s sophistication gets psychiatry (mostly) wrong, our profession has a daunting problem on its hands. I do not intend to explore the many erroneous claims proffered in Dr Angell’s reviews and letters, most of which have been rebutted by several prominent psychiatrists.2-4 (For a magisterial de-bunking of the “antidepressants are just expensive placebos” myth, I recommend the splendid monograph by Davis JM et al.5) Instead, I want to focus on one particularly troubling passage in Dr Angell’s rejoinder to Drs Richard Friedman and Andrew Nierenberg. She states therein that
“[Drs Friedman and Nierenberg] are simply wrong in asserting that psychiatry, in using drugs to treat signs and symptoms of illness without understanding the cause of the illness or how the drugs work, is no different from other medical specialties. First, mental illness is diagnosed on the basis of symptoms (medically defined as subjective manifestations of disease, such as pain) and behaviors, not signs (defined as objective manifestations, such as swelling of a joint). Most diseases in other specialties produce physical signs and abnormal lab tests or radiologic findings, in addition to symptoms. Moreover, even if the underlying causes of other diseases are unknown, the mechanisms by which they produce illness usually are, and the treatments usually target those mechanisms . . . [italics added].”2
In my view, Dr Angell’s assertions reflect both a serious misunderstanding of psychiatric diagnosis, and—equally important—a failure to address the core philosophical issues involved in her use of the terms “subjective,” “objective,” “behaviors,” and “signs.” The bright lines of separation drawn by Dr Angell—eg, between “behaviors” and “signs,” “psychiatry” and “general medicine”—are in fact far fainter than she acknowledges. But this is not surprising. Dr Angell’s analysis is part of a long-standing narrative—alas, sometimes embraced by psychiatrists themselves—that has had the effect, if not the intent, of marginalizing and denigrating psychiatry.6
Let’s focus on the linguistic and philosophical issues that underlie Dr Angell’s claims regarding psychiatric diagnosis. First, let’s dispatch the demonstrably erroneous claim that psychiatric diagnosis does not involve “signs.” To be sure, most psychiatric diagnoses are made primarily on the basis of the patient’s self-expressed reports, experiences, and complaints—in effect, his or her “symptoms.” And, unfortunately, DSM-IV itself contributes to the confusion by calling any feature of psychiatric illness a “symptom.” The framers should have been more precise in their terminology!
However, using the definition of “sign” common in general medicine—ie, “Any abnormality indicative of disease, discoverable on examination of the patient; an objective indication of disease, in contrast to a symptom, which is a subjective indication of disease”7—we find that a number of “signs” are indeed part of the criteria sets for several DSM-IV psychiatric disorders:
• Significant weight loss (major depressive episode, anorexia nervosa)
• Psychomotor agitation or retardation “. . . observable by others, not merely subjective feelings of restlessness or being slowed down” (major depressive episode)
• Increase in goal-directed activity . . . or psychomotor agitation (manic episode)
• Distractibility (ie, attention too easily drawn to unimportant or irrelevant external stimuli) (manic episode)
• Motoric immobility as evidenced by catalepsy (including waxy flexibility) or stupor (schizophrenia, catatonic type)
• Stereotyped movements, prominent mannerisms, or prominent grimacing (schizophrenia, catatonic type)
• Avoidance of eye-to-eye gaze (autistic disorder)
• Delay in development of spoken language (autistic disorder)
I have deliberately omitted signs of cognitive disorders (impaired recent memory, impaired calculation, etc), such as Alzheimer disease, so as to head off the objection that “those are really neurological conditions.” And I have listed only those findings I believe are unambiguously “signs,” according to the definition provided. However, one could make a credible argument that many other determinations made by psychiatrists are in fact observations of “signs,” eg, pressured speech, loose associations, and markedly elevated affect. Then there are the many cognitive functions we assess, such as the patient’s ability to perform calculations, recall words, and draw a clock face, all of which enter into our diagnosis of delirium, dementia, and the amnestic disorders.
Of course, in addition to data obtained from the mental status examination, psychiatrists routinely assess a variety of external (non-subjective) validators of diagnosis—eg, the patient’s vocational function, prior hospitalizations, substance use, suicide attempts, family psychiatric history, and comorbid medical disorders. Finally, although not yet a part of routine clinical assessment, several biomarkers of psychiatric illness show considerable consistency across many studies—eg, enlarged cerebral ventricles and abnormal smooth pursuit eye movements in schizophrenia.8
“Philosophical problems begin when language goes on holiday.”
—Ludwig Wittgenstein, Philosophical Investigations
Now, I am (painfully!) aware of the kind of argument often marshaled against my list of putative psychiatric “signs.” It goes something like this:
Oh, come now, Doctor! With the exception of weight loss, you have really listed behaviors, not signs—at least, not as that term is understood by other medical specialties. In fact, you are really talking about certain “disapproved of” behaviors, such as an “increase in goal-directed activity.” Really, now! You are just making value judgments! Who is to say what degree of goal-directed activity is or is not normal? Don’t psychiatrists really just mean, “I don’t like how much social or sexual activity this guy is carrying out!”? And who is to say that there is anything pathological in maintaining certain postures for long periods? If you watch a group of people performing tai-chi, you could call them all “catatonic”! No—in general medicine, a “sign” is something objective, like swelling or pallor or hypoactive reflexes. And, these bona fide signs are usually reflected in abnormalities we can pick up in lab tests, imaging studies, nerve conduction studies, etc.
This essentially “positivist” view of medical diagnosis6—and of the distinction between signs and behaviors—has the ring of superficial plausibility. Indeed, I would guess that a number of psychiatrists would nod (somewhat ambivalently) in general agreement with it. But in my view, this analysis is quite wrong-headed. It creates a spurious distinction that, on closer examination, is not supported by the way several other medical specialties function in clinical practice. It also employs a sense of the term “objective” that harks back to the largely discredited school of logical empiricism (or logical positivism) that my colleagues and I have addressed elsewhere 6,9
Why, for example, does the neurologist’s statement, “The patient’s handgrip is weak,” describe an objective finding, whereas the psychiatrist’s statement, “The patient’s psychomotor activity is abnormally decreased,” entails a subjective judgment? (Could it be that neurologists simply don’t “approve of” certain degrees of muscle weakness? And that in a society that “valued” myasthenia, a weak handgrip would not be deemed pathological?) Why does the claim, “The patient’s finger appears swollen,” represent an objective finding, whereas the statement, “The patient’s speech is pressured,” represents merely a subjective judgment? Is there some scientific or epistemological principle that privileges visual or tactile over auditory data in deciding what counts as “objective”? One philosophically coherent meaning of “objective” stipulates 2 main requirements: careful and repeated observation and concurrence among multiple observers. On this view, psychiatric diagnosis is, to a considerable degree, an “objective” process.10,11
True: a swollen finger might well be correlated with, say, an elevated white blood cell count, whereas there may be no abnormal lab test associated with pressured speech. But it is not clear why that difference ought to render the determination of pressured speech merely “subjective.” (No general physician would dismiss a swollen finger as inconsequential or “merely subjective,” even if all the lab tests came back within normal limits.)
As for the objection that “you can measure the swelling in a finger, whereas pressured speech is just a judgment call,” I would reply thus: If psychiatrists wanted to devise a sophisticated recording device that computed the number and volume of words per minute spoken by the patient, we could easily do so. And, after all, how many general physicians “measure” their arthritic patient’s fingers before declaring that they are swollen? Of course, in our pathologizing of both terms—“swelling” and “pressured speech”—there is an irreducibly non-objective judgment. This is because each term requires general agreement on what degree of deviation from the norm constitutes “pathology.” This is not something that observation alone can tell us; it is ultimately an existential judgment, based on very broad concepts of “health,” “disease,” “impairment,” “suffering,” etc.
Furthermore, critics of psychiatric diagnosis need to acknowledge that a number of well-recognized medical and neurological disorders are essentially symptom-based constructs that do not necessarily or consistently “. . . produce physical signs and abnormal lab tests or radiologic findings.” Nor are the “mechanisms by which they produce illness” necessarily well understood. Consider the diagnosis of migraine headaches by our colleagues in neurology. As one recent review put it:
Migraine is a very common disabling brain disorder with unclear pathogenesis. . . . It is now generally accepted that the primary cause of migraine lies in the brain, but the nature and mechanisms of the primary brain dysfunction that leads to activation of the meningeal trigeminal nociceptors remain incompletely understood and controversial.12
Well, it’s good to know that the brain is involved! The International Headache Society criteria are the basis for migraine diagnosis, and these criteria are entirely symptomatic and “subjective” in nature, eg, complaints of unilateral, pulsating headache, often accompanied by nausea or photophobia.13 There is no “lab test” or neuroimaging study that is diagnostic of migraine headaches (although such tests may rule out a brain tumor or other pathology). Indeed, unless the physician witnesses the patient during an acute migraine attack, one could argue that the migraine diagnosis is made solely on the basis of a normal neurological examination and the patient’s verbal behavior, ie, the patient’s narrative history of symptoms.
Space limitations prevent discussion of conditions such as atypical facial pain and Meniere disease, but I would contend that these, too, are essentially symptom-based, clinical diagnoses. Even idiopathic epilepsy is regarded, to this day, as a clinical diagnosis, made primarily on the basis of the patient’s history. As neurologist Dr Anthony Marson puts it, “Epilepsy is a clinical diagnosis. . . . the diagnosis is not based upon the results of tests, but upon an accurate description of the attacks.”14 Furthermore, even a witnessed tonic-clonic seizure is, arguably, a series of “behaviors”: the patient falls to the floor, moves his limbs in a certain manner, bites his tongue, becomes incontinent, etc. In what fundamental sense do these behaviors differ from those of the patient with the catatonic subtype of schizophrenia, who assumes uncomfortable postures for prolonged periods and shows waxy flexibility, stereotypy, and automatic obedience?
If it is argued that the patient with catatonia is “producing” catatonic behaviors, whereas the person with epilepsy is involuntarily “experiencing” epileptic behaviors, I would reply that this amounts to a metaphysical speculation and a certain “theory of mind”—not a scientifically verifiable claim. Indeed, the distinction between “behaviors” and “signs” in both epilepsy and catatonia is largely semantic. Behaviors are, in effect, a subtype of “sign,” and conform to the definition of “Any abnormality indicative of disease, discoverable on examination of the patient.”7
As Dan J. Stein, MD, recently noted, “. . . both medicine and psychiatry rely on clinician-observed signs and patient-reported symptoms in order to reach a diagnosis” (personal communication, August 17, 2011). I would not go so far as to claim that psychiatry is “no different” from any other medical specialty. And, to be sure, most of the non-psychiatric conditions I have discussed here fall under the rubric of neurology. The data psychiatrists collect often do differ from those collected by, say, orthopedists or infectious disease specialists. We are less interested than these specialties in x-rays or urine cultures. We are more interested—as we must be—in the patient’s phenomenology: the structure and contents of his “inner world.”15 But the wall of separation implied by Dr Angell’s comments is, in truth, riddled with holes. Indeed, psychiatry bears more similarities to other medical disciplines than many clinicians either understand or acknowledge.