This post was co-written with Dr Max Fink, professor emeritus of psychiatry and neurology, Stony Brook University, Long Island, NY.
There has been a lot of recent talk in the media about psychiatric diagnosis and what it's for. In the New York Times, psychiatrist Ronald Pies says that the purpose of diagnosis is to make people feel good. He argues that it the importance of diagnosis lies in its tranquilizing effect upon the patient, supposedly "eminently humanizing" in its consequence.
This is wrong. The point of diagnosis in mediicne, including psychiatry, is to give the patient a notion of his future course (prognosis: "Doctor, am I going to get better?") and to steer the clinician towards treatments appropriate for that diagnosis. In psychiatry there are not many treatments that are specific for given diseases, but there are a few (e.g. lithium in mania, lorazepam in catatonia), and downgrading the diagnosis to a kind of feel-good status loses track of these advantages. It's been difficult enough to get colleagues to accept such treatments as convulsive therapy for melancholia.
The medical model of diagnosis identifies the patients symptoms, examines for bodily signs, looks to independent markers of dysfunction, offers a preliminary diagnosis, and directs attention to treatments. Treatment-related relief validates the diagnosis. But here's the bad news: DSM-III in 1980 excluded all laboratory procedures from the diagnostic process. A naïve confidence was placed in the identification of symptoms alone, with each diagnosis based on the presence of symptoms from a menu-list.
Psychiatry has a long history of using diagnostic tests: the Wassermann test for neurosyphillis, tests for vitamin and endocrine deficiencies (beri-beri, hypothyroidism), and for known genetic disorders (Tay-Sachs, Huntington). Yet DSM specifically rejected diagnostic tests in getting at the major puzzles in psychiatry: major depression, bipolar disorder, and schizophrenia. There is an endocrine test, called the dexamethasone suppression test, that differentiates between melancholic and non-melancholic depressive illness. This test was discarded in 1986. Meanwhile, the number of diagnoses in the DSM series has increased inexorably, from 106 in the first edition of 1952 to 265 in DSM-III (1980) and 297 in DSM-IV (1994). For none of these diagnoses were laboratory tests offered as criteria, on the assumption that symptom patterns and course of illness alone would define the conditions.
The inadequacies of the present system fly in one's face. The Cross-Disorder Group recently found that major depression and schizophrenia share common genetic defects. So much for the idea – absolutely fundamental to DSM –of a firewall between the two disorders. Recent government-supported evaluation studies reported that drug treatments for schizophrenia, major depression, and bipolar disorder had not worked out well. The reason? The population samples selected for these trials using symptom criteria alone were highly heterogeneous and the effect of the medicalions only minimally better than placebos.
The medical model of diagnosis, which relies on verification by tests, has proven successful in defining at least three psychiatric conditions: melancholic depression, catatonia independent of schizophrenia, and self-injurious behavior in autism. Melancholia differs from other kinds of depression because it has a biology of its own (disrupted cortisol metabolism) as well as distinctive treatment responses to convulsive therapy and tricyclic antidepressants. Relief rates are greater than 80%, compared to 34% relief rates for the Prozac-styole medications popularly prescribed for major depression.
The proposed DSM-5 offers hundreds of ill-defined diagnoses which exist mainly in the eyes of the medical enthusiasts who advocate for them, and have been subjected to no verification. DSM-5 abjures laboratory tests, thereby guaranteeing the same failures in illness classificaiton and treatment selection that have marked the decades since 1980. For all of medicine, the diagnosis is a marker of effective pathways to clinical releif. It is more than a label of reassurance. While it is true that we have few biological markers for the present catatlogue of psychiatric disorders, the DSM-5 ossifies the present failure by continuing to insist of numbers of symptoms and course alone as sufficient for diagnosis. Either psychiatry is a medical discipline or it is not.