Diagnosis: Same as It Never Was

In the last 55 years, psychiatry has seen two extremes in the approach to diagnosis—one was individualized, improvisational and impressionistic; the other has been reliable, but too constraining. In 1952 and 1968, when the first and second editions of the DSM were published, psychoanalytic principles were ascendant and psychiatric practice had largely moved out of the asylum and into medical schools, outpatient clinics and private offices. In contrast to earlier views that mental illness was either hopeless or the product of a moral failing, the period after World War II was a time of great optimism in psychiatry. Illness could be understood as the result of derailed human development, or arising from conflict between biological drives and the environment. Such problems were thought to be correctable through psychotherapy.

DSM-I and DSM-II, unlike later editions, did not include detailed symptom lists, because disorder was presumed to be the result of processes that could not easily be seen. They could be deduced in a specialized therapeutic conversation. A prevailing view, and still an attractive ideal, was that each illness—and therefore each treatment—was one of a kind.

That was the good and the bad news. We like to be treated as individuals, but if each treatment is the first of its kind, it is impossible to garner evidence that might help you choose the most effective treatment. DSM-III, published in 1980, was an attempt to correct this, and it did so emphatically. By the 1970s, the limitations of psychotherapy (along with its benefits) were becoming clearer and new treatments were starting to take hold. With the rise of psychopharmacology, psychiatrists sought more dependable categories of illness. Following an era when diagnosis was heavily influenced by the theory and experience of the practitioner, the authors were determined to establish more reliable diagnoses based on strict descriptions alone. DSM-III and its successor purposely left theory out.

But this system did not make diagnosis more valid. After a quarter century of basic research, we know much more about how the brain works and which treatments are effective, but we are not yet close to mapping psychiatric disorders to specific biological or environmental causes.

What's next? Here's a prediction: DSM-V authors will approach their work with a generous attitude toward human nature, and will create a diagnostic system consistent with today's scientific knowledge. They will offer it, not as the last word, but as a tool for testing hypotheses about mental suffering. After all, good science is about getting it both right and wrong. And wisdom—with all due respect to the Greeks—is about appreciating how much we do not know.

Miller is editor in chief of the Harvard Mental Health Letter. For more information go to health.harvard.edu and health.harvard.edu/newsweek.

© 2007