The American Psychiatric Association has recently published a new and radically revised nomenclature—The Diagnostic and Statistical Manual of Mental Disorders (DSM-III). This classification offers innovations in method and content which have stirred considerable controversy. The many psychoanalysts who were actively involved in writing and field testing DSM-III tend to like the new system, but others have expressed concern about its possibly detrimental effects on psychiatry and psychoanalysis. They contend that DSM-III institutionalizes a descriptive and anti-analytic bias and ignores the knowledge accumulated during eighty years of psychoanalytic practice. I will briefly outline some of the issues involved and suggest a perspective on them.
In order to be acceptable to psychiatrists of all theoretical orientations, DSM-III avoids etiologic and management considerations and bases its diagnoses on descriptive observations that require only a low level of inference. This procedure greatly improves the reliability of DSM-III and allows clinicians to agree upon a diagnosis even if they strongly disagree on its etiology and management. Unfortunately, this advantage has been purchased with the sacrifice of psychodynamic concepts and observations which were regarded as too unreliable and specialized for inclusion in the manual. The DSM-III diagnosis is thus intended to be only the beginning of a complete formulation and treatment plan.
The DSM-III method utilizes two innovations. The first—the multiaxial approach—shall be enthusiastically welcomed by psychoanalysts. The clinician is no longer compelled to reduce complicated observations to one simple label but can rather diagnose the patient along each of five axes. Axis II records the presence of the adult personality disorders and specific developmental disorders of childhood; Axis I includes all of the other diagnostic categories in the classification and, therefore, the majority of them. Axis III records the presence of relevant medical disorders; Axis IV is a rating and description of stresses in the patient's life; and Axis V records the patient's highest level of adaptation in the
year preceding evaluation. This five axis diagnosis provides a more rounded, specific, and prognostically important picture of the patient than did the single label of the previous classification.
The manual's second innovation is more controversial among psychoanalysts, but has long been recommended by experts in diagnosis. This is the provision of explicit criteria to spell out the grounds upon which each diagnostic decision is made. The field trials indicate that this system dramatically improves diagnostic reliability—a necessary first step toward rational clinical discourse, treatment planning, and the generalizability of research results. Of course, the improved reliability does not imply that the current diagnostic categories are valid. In most instances, the criteria are based upon clinical judgment alone and have not yet been validated with systematic data on important correlates like clinical course, outcome, family history, or treatment response. The clearly stated diagnostic criteria require explicit data and inhibit the use of hunches and inferences. This is experienced as a mixed blessing by psychoanalysts. We are accustomed to using our clinical experience, intuition and educated inference in making diagnostic decisions and will sometimes feel that valuable preconscious information is lost in following the seemingly simple-minded and atomistic DSM-III system. On the other hand, the requirement of explicit criteria may induce us to increase the clarity of our own fuzzy discourse and to articulate in operational terms the basis of our intuitive judgments. Certainly, the lack of a reliable psychoanalytic or psychiatric nomenclature has been a major stumbling block to progress in psychoanalytic research and discussions, and we stand to benefit from a serious consideration of the criteria method.
The particular criteria required for certain of the diagnoses will be unfamiliar and unpleasing to at least some psychoanalysts. The narcissistic personality disorder (which was not included in DSM-III) is good example of the contrasts between DSM-III and psychoanalytic diagnostic thinking. I was involved in writing this section and have a firsthand knowledge of the problems involved. As a psychoanalyst, I might want to include in the criteria for a narcissistic personality some statement about the patient's deficits of intrapsychic differentiation, his relative deficiency of specific intrapsychic structures, and his deficits in achieving a cohesive sense of self. Such concepts are enormously useful in psychodynamic clinical work and help us in making treatment decisions and in predicting prognosis. Even among analysts, however, there is no general agreement or reliability in deciding how to measure the relative cohesiveness of the self or the proportion of object as compared to self-object relatedness. These are not concepts that can, at this stage of our understanding, be easily translated into simple and readily observable behaviors, and such factors are especially difficult to evaluate based on the data obtained from only one or two diagnostic interviews. For nondynamic psychiatrists these concepts have little or no meaning at all. Thus, the patient who satisfies the criteria for narcissistic personality disorders as defined by DSM-III may or may not meet one or another of the psychoanalytic definitions of the disorder. This is true for a number of other diagnoses as well.
The most serious controversy surrounding the new system concerned the word neurosis. Those patients formerly considered as having neuroses are now classified in several newly developed categories, such as the anxiety disorders, somatoform disorders, dissociative disorders, and affective disorders. Neurosis as a classifying term was abandoned as an anachronism without any stable or specific meaning. This decision was endorsed by many nonanalysts but also, interestingly, by those analysts who felt the term had become too variously defined to retain usefulness, particularly in a descriptive sense. The label “Neurotic Disorders” appears in the body of the classification to announce in bold type the particular new categories under which the formerly unified neuroses are now separately subsumed. In addition, traditional labels appear in parentheses immediately following the revised DSM-III term—for example, “dysthymic disorders (or depressive neuroses).” The manual will allow either label to be used, although the new one is recommended. Each will be coded with the same digits for statistical purposes.
The furor over the term neurosis seemed like something of a tempest in a teapot from a scientific viewpoint, but was crucially important to many people's semantic and political sensitivities. The term neurosis has had two different meanings: (1) to describe and label several types of psychopathology—Freud's original four psychoneuroses, plus a few others; (2) as an etiological statement that significant aspects of character and symptoms arise out of wishes, fears, and conflicts operating, at least largely, on an unconscious level, and perpetuating derivatives of the conflicts of childhood which caused distortions or arrests in development. Neuroses in the second sense would account for the psychopathology not only of much of the anxiety, affective, somatoform, dissociative, personality, and psychosexual disorders but would operate to varying degrees in all psychiatric disorders. The etiological meaning of neuroses has been kept out of DSM-III just as have most other etiological considerations. In the current compromise, the limited descriptive use of neuroses has been perpetuated as an optional choice for those who desire to use it. The crucial aspect for analysts to study and to teach (especially to nonanalysts) is the ubiquitous presence of unconscious psychological conflict as at least a partial contribution to all psychiatric disorders and its central position in some of them.
What can we say on balance about this new manual? It was an enormously difficult task to create a classification system within a discipline precisely at a time of its greatest ferment. Although DSM-III does not draw deeply from the richness of psychodynamic understanding, it is nonetheless readily usable and useful to psychodynamic clinicians. For the future, it is incumbent upon us to clarify and make operational our own concepts and to study them systematically. Only as we begin to test the reliability and validity of our own convictions and observations will we continue to make a significant contribution to the increasing rush of scientific psychiatry. DSM-III is vastly superior to DSM-II and will certainly be superseded in its time by a vastly improved DSM-IV. It is important that analysts and analytic research be instrumental in this progress. It is also clear that a nomenclature intended for use by all psychiatrists can never encompass all of the observations and diagnostic refinements useful to psychoanalysts. It is surely unfortunate that we do not have our own reliable psychoanalytic nosology. The method used in DSM-III is worthy of our study and assimilation, even if we cannot agree with all of the decisions that have been made in it.