Ever since psychiatry emerged as an independent field of medicine, it has stood in the shadows of other specialties, never progressing scientifically as fast as neurology, oncology, or cardiology. In many developed countries, the decline of mental hospitals has brought psychiatric services into closer professional contact than ever with these and other medical domains. Yet psychiatry continues to inhabit medicine's netherworld - within the mainstream clinically, outside it scientifically.
Psychiatry is separated from other medical specialties by its lack of an objective basis for diagnosis. Other medical fields have refined the diagnostic process to the point where computerized laboratory tests have virtually replaced clinical examination of patients. Psychiatry remains dependent on interpretation of detailed case histories. These can be obtained only through careful examination and direct questioning of patients. No universal diagnostic tests exist for the most frequent mental disorders, such as depression, anxiety, and schizophrenia.
Of course, computer programs can now process data derived from patients' symptoms and generate psychiatric diagnoses. But these apply rules that only replicate the way one or more psychiatrists interpret a patient's symptoms. There is nothing absolute about the program's output, although it does at least do the same thing every time, which cannot be said for psychiatrists. After all, interpretations of a case may differ, and the outcome cannot be resolved other than by an appeal to authority - "I am more senior and experienced than you, so my diagnosis takes precedence."
Beyond diagnostic disputes in individual cases, psychiatry's absence of objective diagnostic tests implies something more radical. The history of the field is littered with definitions of entire disease categories that have been influenced by diagnostic fashions, politics, and the availability of treatments.
Diagnostic fashions reflect prevailing views of what causes mental illness. From the 1950's to 1980's, for example, psychoanalysis held sway among US psychiatrists. But the psychoanalytic approach, with its reliance on personal interaction with patients, led to a broad conception of pathology. During the 1970s, two international studies demonstrated that US psychiatrists diagnosed about twice as many patients as having schizophrenia as psychiatrists in other countries, where psychoanalytic training was relatively uncommon.
The pendulum has since swung in the US, with biological theories of mental disease displacing all other causal contenders. Parallel to this shift away from psychoanalysis, the diagnosis of schizophrenia has shrunk dramatically, and standard psychiatric practice in the US now incorporates a narrower definition of schizophrenia than is used in Europe.
Psychiatric diagnosis has also been vulnerable to politicization, nowhere more so than in the former Soviet Union, which locked up political dissidents in mental hospitals on what amounted to a charge of schizophrenia. I once interviewed three former Russian dissidents who had been diagnosed with schizophrenia and forcibly treated. In no case did I find any evidence that they had ever suffered from the disease as it is defined in the West.
Soviet authorities relied on an idiosyncratic definition of schizophrenia, introduced by a professor of psychiatry (A.V. Snezhnevsky) in Moscow. Anyone who believed that the political system should be changed was suffering from "reformist delusions." In deference to psychiatry's diagnostic conventions, he emphasized the history of the "condition." If a dissident demonstrated against the Soviet system at least twice - say, by distributing illicit writings - the "delusions" signified schizophrenia.
But we should not be complacent about the potential for politicization in the West. In 1987, political pressure in the US led to the elimination of homosexuality as a diagnosis from the nationally accepted guidelines. In fact, the inclusion of homosexuality as a pathological condition suggests the extent to which psychiatric diagnosis has been shaped by social and political currents. By contrast, no one - whatever a society's characteristics - has ever seriously questioned whether cancer is a disease.
Moreover, how psychiatrists treat mental illness profoundly influences how it is defined. Depression is a recent diagnosis - and now one of the most common. A study of diagnoses from 1949-1969 at the London psychiatric hospital where I work suggests why: a major shift from diagnoses of anxiety to diagnoses of depression coincided with the introduction of the first antidepressant pill. Psychiatrists began altering their diagnoses to a condition for which they had an effective treatment.
So is psychiatric diagnosis akin to pseudo-science? International comparisons carried out by the World Health Organization show that as long as psychiatrists from different countries are trained to interview patients using a standardized format, they can agree on who is suffering from schizophrenia. The WHO studies revealed that schizophrenia defined narrowly - using a particular group of unusual symptoms - occurs with similar frequency throughout the world.
Still, a broad definition produces wide variation in frequency across countries. Yet defining schizophrenia broadly may be unavoidable, for it is probably not a single disease with a distinct cause. Indeed, Eugen Bleuler, the Swiss psychiatrist who introduced the term, referred to the condition as "the group of schizophrenias."
Until biological markers for psychiatric illnesses are identified, diagnosis will remain in a 19 th century time warp, relying on the equivalent of broad and outdated medical terms like "fever" and "dropsy." But patients and their families and friends need not worry excessively. Most psychiatric treatments are effective across a wide range of diagnoses, whatever their origin. Psychiatry's success - like that of all medical specialties - should be judged by its ability to relieve human suffering, regardless of disagreements about causes.