Customize your JAMA Network experience by selecting one or more topics from the list below.
Author Affiliation: Department of Psychiatry,
Johns Hopkins School of Medicine, Baltimore, Md.
Psychiatrists are reconsidering how to classify the conditions they
treat. What they decide will affect who receives psychiatric services and
what those services will be––matters surely of interest and concern
to all physicians and health care workers. But the American Psychiatric Association
has delayed a new revision of its classificatory manual so as to confront
what historically has been a most obstinate challenge––how to
bring its diagnoses and explanations together.1
In the mid-20th century, Freudian psychoanalysts led the discipline
by teaching that unconscious mental conflict produced psychiatric disorders.
Because psychiatrists then wrapped all disorders into a single explanation
(and treated them similarly), they saw little point in sharply distinguishing
one manifestation of conflict from another. Although psychoanalytic influence
waned in the late 1960s, the diagnostic competence of psychiatrists, needed
for progress, did not improve. Even elementary census reports such as the
number of patients with schizophrenia who were admitted to mental hospitals
A clamor for reliable research criteria from Washington University’s
psychiatry department3 captured the attention
of Robert Spitzer at Columbia, who was editing the impending third edition
of the American Psychiatric Association classificatory manual for psychiatry.
Because he thought a criterion method for diagnosis would advance psychiatric
practice generally, not just in research, he led the process of building such
a method into the Diagnostic and Statistical Manual of Mental
Disorders, Third Edition (DSM-III) of 1980.4
Specifically, Spitzer asked expert clinicians and investigators what
features they used to identify, rather than to explain, the disorder they
studied. He then picked those features that were observable, such as hallucinations
and delusions, and insisted they be so defined that any psychiatrist could
confidently recognize them. Finally, he tested delineated features as diagnostic
criteria in field trials to uphold those criteria that helped psychiatrists
make replicable diagnostic decisions.
Because DSM-III provided psychiatrists with
a standardized diagnostic nomenclature, their enterprises flourished. Geneticists,
pharmacologists, and brain imagers became research partners with investigative
psychiatrists. Clinicians could compare experience, debate treatments, and
demand equity for their patients in the health system. Since 1980, DSM has gone through several revisions.
Why should anyone complain now? Quite simply, the process got out of
hand. Diagnoses based on descriptive criteria burgeoned. The DSM currently admits close to 300 mental and behavioral disorders.
Given that clinical appearances forge diagnoses, a particular patient can
satisfy the criteria for several disorders and many dissimilar patients can
meet criteria for the same disorder. Because the manual fails to identify
what underlies the symptomatic expression of a condition, it cannot suggest
intelligible principles relating one disorder to another or illuminate why
certain of them bunch together. For these reasons, faith in the criterion
method has gradually faltered.1 However, many
psychiatrists from long experience with misguided explanations are reluctant
to change the approach of DSM.
What exactly is this approach and how is it different from the approach
in general medicine? Physicians who come to psychiatry from another specialty
note how DSM diverges from the internists’
manual, the International Classification of Diseases (ICD). In the ICD, there are separate
sections devoted to each bodily organ (heart, gastrointestinal tract, kidney)
and physicians find disorders organized around etiopathies, such as vascular,
neoplastic, infectious, and autoimmune processes, which alter the organ’s
function and singly or more often in understandable combinations explain the
physical signs and symptoms of their patients.
The DSM is not systematic in that way. Being
appearance driven, it is similar to a naturalist’s field guide with
the advantages and disadvantages of such. Just as Roger Tory Peterson’s A Field Guide to the Birds5 distinguishes
a prothonotory from a yellow- or blue-winged warbler by the bird’s coloring,
voice, and range, the DSM distinguishes and then
arranges mental disorders by their appearance––“on their
shared phenomenological features.”6 Even DSM’s “decision trees” for differentiating
mental disorders mirror those in bird and botanical field guides. They enhance
accuracy of identification; therefore, they are reliable but do not explain
In fairness, ignorance over how the brain, the primary source of all
aspects of mental life, evokes either healthy or unhealthy psychological events
hinders a more essential etiopathic classificatory system for psychiatry.
Physicians know how other bodily organs work in life and appreciate the direct
roles of structural and functional pathology in producing their disorders.
But the brain is unlikely to produce mind, like the kidney produces urine,
or the liver produces bile. And, as all psychotherapists know, the mind, unlike
the urine, can shape and sustain troubles generated from many sources.
The situation faced today by psychiatry, however, is remarkably similar
to the situation of medicine a century and a half ago and is susceptible to
a similar solution. In the 19th century, physicians advanced beyond their
practice of identifying patients from appearances (eg, fever patterns, wet/dry
gangrene) to grouping, diagnosing, and tentatively explaining their conditions
by the etiopathic (eg, vascular, infectious, neoplastic) agencies comprehended
at the time.
In this way, physicians built a coherent and ultimately progressive
infrastructure for interpreting the disorders afflicting their patients and
opened their descriptive skills to the discerning of pathogenic processes.
Classification was synthesized from explanation, expecting that regular reciprocating
exercises between these 2 methods of thought would lead to progress in both.
The success of this enterprise is the very history of the dynamic maturation
of medicine in the 20th century. Perhaps the most obvious and recent example
of its power was its reclassifying and reinterpreting peptic ulcer, a condition
whose signs and symptoms were long known, from among the inflammatory disorders
to the infectious disorders.
Psychiatrists have not taken such a synthesizing step but eventually
they must. They may hesitate because they await further advances in the basic
sciences. But, in hesitating, psychiatrists squander opportunities to contribute
knowledge about human mental disorders in ways that would promote the very
progress sought, especially in psychology and neuroscience.
Psychiatrists now have, in part because of research prompted by the DSM-III, sufficient information to launch such a synthesizing
project. If the process begins simply as medicine did years ago, both what
is already known and how to advance diagnostic practice toward explanation
can be carried beyond dependence on a field guide.
Mental disorders can be separated into 4 simple and comprehensible clusters.
Each cluster has its salient, identifying (ie, necessary) feature. But, such
salient features and the clusters they identify are not mutually exclusive
and may, as with medical etiopathies, combine in understandable (ie, comorbid)
ways in some patients. Each cluster derives coherence from contemporary psychological
knowledge, discerns common modes of treating the encompassed disorders, and
directs clinical scientists intelligibly toward systematic investigations
with psychology, neuroscience, genetics, and development.7
The first cluster comprises patients with brain diseases that directly
disrupt the neural underpinnings of psychological faculties, such as cognition,
emotion, and perception. This family includes those patients with Alzheimer
disease, schizophrenia, and bipolar disorder––patients who “have”
(or are proposed to “have”) structural or functional pathology
affecting their brains and thus disrupting particular psychological faculties.
The second cluster encompasses those patients vulnerable to mental unrest
because of their psychological make-up. These patients deviate to some extreme
along one or several of the universal, psychometrically measurable dimensions
of human disposition, such as intellect, extraversion, and emotional stability.
The deviance from the mean primes these patients to respond with greater distress
and discouragement to life’s challenges. They do not necessarily “have”
neuropathology, although they can, but experience distress because of how
their brains and minds matured––making them who they “are.”
The third cluster comprises patients who adopt a behavior that has become
a relatively fixed and warped way of life (ie, patients with alcoholism, drug
addiction, sexual paraphilia, or anorexia nervosa). They are patients because
of what they are “doing” and how they have become locked into
doing it. What these individuals “have” and what they “are”
may enhance their vulnerability to the behavior, either by triggering or sustaining
it, but primarily their conditions are tied etiopathically to what they chose,
how they responded to that choice, and why, for them, such choices led to
a driven habit.
The fourth cluster encompasses patients with distressing mental conditions
provoked by events thwarting or endangering their hopes, commitments, and
aspirations. For example, they experience grief, situational anxiety, homesickness,
jealousy, or posttraumatic stress disorder; states that derive from what they
encountered in life and from the assumptions and expectations drawn from those
encounters. Diseases these patients have, dispositional deviance making them
who they are and even events provoked by what they are doing, can contribute
to the frequency, character, and duration of these disorders, but the necessary
etiopathic feature in this group is the encounter itself and what it has come
to mean to the patient.
Clustering psychiatric conditions according to what patients have, what
they are, what they do, or what they encounter, and noting how each etiopathic
cluster generates psychological unrest, helps distinguish causes from symptoms.
For instance, depressive moods or anxious states are psychological symptoms,
just as fevers or coughs are physical symptoms. A depressive mood can derive
from a life encounter or may represent the habitual emotional response of
a person with an unstable personality. But depression may accompany an addict’s
intoxication or can present as the prominent feature of bipolar disorder.
The key to treatment and prognosis is not the depressive symptom but its underlying
Other advantages in grouping mental disorders into these clusters include
(1) because psychiatrists already use these modes of explanation, rendering
them more explicit dispels mystery from this field; (2) each cluster encourages
research directed at uncovering the hypothesized causal elements in the examples
(for diseases: pathogenesis; for dimensions: psychometric validation; for
behaviors: mechanisms provoking and shaping habit; for encounters: life histories
and their contexts); and (3) distinctions identified by these simple clusters
render psychiatric practice more intelligible to all observers but particularly
to medical colleagues. This strengthened explanatory structure indicates why
recognizing symptoms, such as depression or anxiety, despite their reliable
definitions or even symptomatic remedies, is insufficient for full understanding
of a patient.8 One must strive to find the
generative sources of the symptoms. Given that alleviating diseases, interrupting
disordered behaviors, guiding patients with dispositional vulnerabilities,
and rescripting the meaning-filled assumptions of those patients troubled
by life-encounters are distinct therapeutic exercises, these clusters affirm
and identify the skills used by psychiatric specialists.
Leahy9 quotes a Chinese adage, “The
beginning of wisdom is calling things by their right name,” an adage
identifying the historic value of the DSM-III. To
be coherent, psychiatrists must explain to their patients, themselves, and
their medical associates what they have named. Psychiatrists could advance
toward this goal if, as the fifth edition of DSM is
planned, they consider organizing the disorders already reliably described
into etiopathic clusters, such as those suggested here.
Corresponding Author: Paul R. McHugh, MD,
Department of Psychiatry, Johns Hopkins School of Medicine, 600 N Wolfe St,
Meyer Bldg, Room 127, Baltimore, MD 21287-7413 (firstname.lastname@example.org).
Financial Disclosures: None reported.
McHugh PR. Striving for Coherence: Psychiatry’s Efforts Over Classification. JAMA. 2005;293(20):2526–2528. doi:10.1001/jama.293.20.2526
Coronavirus Resource Center
Create a personal account or sign in to: