Desktop version

Home arrow Psychology arrow Psychiatric Diagnosis Revisited: From DSM to Clinical Case Formulation


From Prototype to Checklist: A Brief History

The checklist-based approach to diagnosis that can be found in the DSM-5 is relatively new to the discipline of psychiatry. Before the DSM-III, psychiatric diagnosis usually started from prototypes. At that time, psychiatric handbooks did not contain lists of items, but “clinical pictures”: narrative descriptions of different forms of psychopathology. Indeed, the prototypical approach is as old as psychiatry itself and can already be found in the work of the alienist Philippe Pinel. In his book A Treatise on Insanity, Pinel (1806) not only reviews key principles of his new discipline, including the moral and medical treatment of insanity, but also provides a study of different “species” of “mental derangement” (Pigeaud 2001). Guided by Immanuel Kant’s (1764) suggestion that the diagnostic notions of general medicine might be applied to insanity, and inspired by Thomas Sydenham’s (1624-1689) suggestion that diseases should be studied with the method that botanists use in their study of plants, Pinel supposed that kinds of insanity could be discerned, hence his belief in the project of collecting a nosology of mental maladies.

In Pinel’s view, the insane is subject to all kinds of passions. By observing and listening to patients, writing up their history, and treating them in institutions, he aimed to grasp how passions play a role in different conditions (Pigeaud 2001). Along this way he made distinctions between curable and incurable patients, and between kinds of derangement “depending upon diversities of temperament, habits, intellectual ability, the faculties principally affected, and other causes” (D.D. Davis in: Pinel 1806, p. ii). More specifically, Pinel discerned five forms of mental derangement: melancholia, mania without delirium, mania with delirium, dementia, and idiotism. Melancholia and mania are conditions that were already described since antiquity. Pinel added dementia and idiot- ism to the list, thus making a classification based on which patients could be categorized. He details these so-called forms of derangement qua the symptoms patients reported as well as clinically observable characteristics. Each specific condition he discerns is marked by distinctive modes of behaving and relating, which he illustrates by means of clinical cases from the Bicetre Hospital in Paris, as well as historical cases drawn from the literature. Moreover, Pinel aimed to grasp broader patterns underlying specific troubles, such as the typical course and prognosis of a given condition. Obviously Pinel’s approach to mental illness focused on describing the original form, or basic type, of the five disorders he discerned; hence the idea that he engendered a prototypical approach to diagnosis.

Just like the early botanists made drawings and pictures of the species they discerned, Pinel and his followers described how different kinds of deranged patients characteristically behave, interact, and how they physiognomically looked like.

After Pinel, the prototypical approach long dominated the field of psychiatry. In handbooks outlining different mental illnesses, attention was often paid to the patient’s present mental state, patterns at the level of etiology and illness history, and typical prognosis and outcomes. Clinical vignettes of particular patients whose problems fitted in very well with the overall clinical picture were included. Some authors, like Eugen Bleuler (1934), who generally followed an inductive approach to the classification of disorders (Moskowitz and Heim 2011), strongly emphasized the psychological mechanisms connected to specific conditions, like disturbances of association in schizophrenia. Others, like Emil Kraepelin (1921), who followed a more deductive approach, started from biological conjectures (Weber and Engstrom 1997).

The main difference between such a prototype-based approach and the checklist method of the DSM is that diagnosis with prototypes focuses on sets of characteristics. Particular behaviors and complaints are not evaluated separately, but examined in terms of patterns that make up a person’s functioning. Moreover, prototypical diagnosis usually starts from demonstrative cases and clinical vignettes, based on which the clinician evaluates whether a given person’s problems match those described in the literature. DSM checklist diagnosis, by contrast, focuses on individual symptoms and signs, leaving aside the question as to how these may relate to one another along an underlying structure. Consequently, the DSM seems to assume that diagnostic criteria are independent of one another and are therefore additive in nature. The DSM’s additive view on diagnostic criteria is particularly reflected in the DSM-5’s severity assessment. Whereas previous versions of the DSM held on to a strict categorical approach, assuming that patients either have a diagnosis or not, the DSM-5 suggests that some disorders manifest in various degrees, depending on the amount of disorder characteristics an individual has. For example, this is illustrated in the DSM-5’s (pp. 490-491) category

“alcohol use disorder”: the more criteria1 a person meets, the more severe the condition is thought to be. What such an approach leaves out of consideration is the qualitative weight certain characteristics might have: individuals with few symptoms might nevertheless suffer deeply from their condition. Given that mental disorders are defined as disturbances that are associated with “distress or disability” in the DSM-5 (p. 20), the qualitative aspect of subjective suffering in relation to mental symptoms should not be neglected.

The first two editions of the DSM (DSM-I and II) also provided prototypical descriptions of mental disorders, but these descriptions were elaborated very poorly. For example, in the DSM-II (p. 33) schizophrenia is characterized with a paragraph of only 128 words, in which no reference to clinical case material is made. Likewise, subtypes of schizophrenia are described with very brief explanations. For example, the specific description of the hebephrenic type of schizophrenia offers a mere 35 additional words.[1] [2] It mentions clinical characteristics like “disorganized thinking” or “silly and regressive behavior and mannerisms,” but doesn’t document or operationalize how these typically manifest in clinical practice. References to other authors who might have discussed such characteristics in previous work are also lacking. Compared to the detailed clinical analyses in classic handbooks like that of Bleuler (1934), who devoted 72 pages to his discussion of schizophrenias, the description of specific psychiatric conditions in the early versions of the DSM is extremely concise, and in fact more stereotypical than prototypical in nature.

One reason why the first versions of the DSM were elaborated so poorly is that initially American psychiatrists didn’t believe in the relevance of elaborating mere classificatory diagnostic systems (Strand 2011). On the one hand, they had a holistic mental health concept, and assumed that mental illnesses “were precipitated by a combination of psychological and environmental etiological factors that were mediated by the constitution or predisposition of the individual” (Grob 1991, p. 422). On the other hand, they conceived illness in individual terms, which made them focus on the patient rather than on the illness qua abstract category (Grob 1991). For example, Karl Menninger et al. (1958, p. 6) argued that diagnostic labels might well engender an artificial feeling of understanding, but actually don’t grasp illness dynamics in patients: “names do not create illness forms; they only comfort the doctors and impress the relatives.”

If categorical classification was only of marginal interest to US psychiatrists in the first part of the twentieth century, then why did the American Psychiatric Association decide to publish a handbook like the DSM? The answer to this question cannot be found in the field of clinical psychiatry, but in the work of epidemiologists and statisticians, who aimed to quantify mental distress. In order to make population overviews of different psychiatric conditions they needed straightforward, clearly delineated disorder categories that could be assessed relatively simply.

Serious interest in psychiatric statistics emerged in the mid-nineteenth century, and as large-scale census studies took off in the early twentieth century, the need to develop standardized classification schemes was placed high on the agenda. In 1918 the American Medico-Psychological Association and the US National Committee for Mental Hygiene published the first Statistical Manual for the Use of Institutions for the Insane. This manual counts 37 pages and provides a basis for quantitative data collection. It contains sample forms for the administration of patients as well as brief narrative descriptions of 22 mental diseases, spread across 16 pages. Notably, these descriptions are brief, amounting to less than a page. They mainly consist of summaries and descriptions of symptoms and differential diagnostic remarks. Clinical psychiatrists criticized the manual, yet “Despite such criticisms, the Statistical Manual became the definitive nosology of the interwar years and went through no fewer than ten editions between 1918 and 1942” (Grob 1991, p. 426). A growing desire to develop a standard nomenclature for psychiatry motivated these revisions. At the side of clinical psychiatry, an impetus to elaborate such classification was found in the work of Emil Kraepelin. While describing and delineating illness entities was a major objective for him, his nosological descriptive approach nevertheless remained strongly narrative in nature.

World War II dramatically changed the overall view of mental disturbance. During these years, severe distress was observed in substantial numbers of soldiers and veterans. Psychoanalytic therapy outside traditional clinics proved to be highly successful in treating these young men, recognizing the link between their distress and the extreme environmental stressors they had been living through (Grob 1991). However, existing classification schemas, like the Statistical Manual for the Use of Institutions for the Insane (which was strongly oriented toward residential psychiatry), were of little use in this context: “Only 10 % of the total cases seen [by military psychiatrists] fell into any of the categories ordinarily seen in public mental hospitals” (DSM-I, p. vi). Since neurotic problems and trauma-related psychopathology were not represented in the manual, the US Army and Navy took to developing their own classifications to get hold of the magnitude of mental disorders among soldiers and veterans (Strand 2011). Inspired by both this new classification system and the Statistical Manual for the Use of Institutions for the Insane, the American Psychiatric Association made it their business to develop their own statistical manual: the DSM-I was published in 1952. Note that whereas some people seem to believe that the concept “statistical” from the DSM’s title refers to statistical evidence that the handbook is based on, this is far from the case. It only indicates that the DSM disorder classes might serve as a basis for categorical classification, and thus provide nominal data for statistical calculations.

The development of the DSM was obviously not really motivated by clinical dilemmas, but by administrative concerns: bureaucrats and social scientists alike wanted to obtain statistics about mental health problems. The net effect of adopting such classificatory systems was that the concept “diagnosis” was narrowed down quite dramatically. Etymologically the concept “diagnosis” (from the Greek “diagignoskein”) refers to “discern” as well as to “know thoroughly” (Harper 2011). In the context of the DSM, diagnosis is not so much concerned with developing thorough knowledge about an individual’s mental health conditions, like traditional clinical psychiatrists were. From the outset, it simply focused on distinguishing between different disorders. Indeed, from its inception the DSM largely equated diagnosis with classification, leaving aside the elaborate narrative description of a patient’s global functioning. This is a reduction that remains neglected. This reduction is productive if one wants to quantify human problems, as it opens the possibility of statistical computation. However, this leaves us with the very serious question as to whether such a leaning toward categorization facilitates an accurate characterization of an individual’s mental suffering.

Indeed, as I further explain in Chap. 3, it could be argued that from its inception the DSM approached psychopathology in a naturalistic way, thus neglecting the subjective experience of psychopathology and above all its contextual embedment. Eventually, the DSM-I contained 106 disorders, a notable increase from the 22 disorders outlined in the 1918 Statistical Manual for the Use of Institutions for the Insane. As is true for all subsequent editions of the DSM, arguments and evidence that substantiated this fivefold increase in the number of disorders were not published in the DSM or in related sources.[3] As a result, it remains unclear as to what motives and interests guided these revisions. Overall, the DSM-I distinguishes between brain disorders, mental deficiency, psychotic disorders, psychophysiologic disorders, psychoneurotic disorders, personality disorders, and adjustment reactions. These are defined by short descriptors across a total of 42 pages. Next to that, the handbook contains sections with guidelines and classification forms for statistical analysis, also spanning 42 pages.

As the DSM-I was adopted in the 1950s, diagnostic classification remained a marginal concern to the then dominant group of psychodynamic psychiatrists in the USA (Strand 2011). Often, generic psychiatric labels, like schizophrenia or melancholia, were seen to be too imprecise as characterizations for grasping the problems a patient suffers from. For example, Karl Menninger (1959) argued that mere classification is insufficient. In his view a psychiatric diagnosis “is always a complex set of statements - descriptive, analytic and evaluative. They have to describe a patient’s method of interacting with his environment, past and present [...] and the psychiatric diagnosis is always both polydimensional and multidisciplinary” (Menninger 1959, pp. 233-234). Such focus on the case-specific study of the meaning and function of symptoms changed only slightly when the American Psychiatric Association was asked to attune the manual to the standards outlined by the International Classification of Diseases (ICD). In 1948 the international community, under the auspices of the United Nations, created the World Health Organization (WHO). One of the objectives of the WHO was to standardize medical classification. Its tool for realizing this objective was the ICD, which provides a taxonomic list of all recognized forms of medical illness. At first, the ICD listed only somatic diseases, but in 1949 a section on mental disorders was added. Given the membership of the USA to the WHO, the American Psychiatric Association was asked to attune its diagnostic system to the ICD format. This gave rise to the DSM-II (1968), an elaboration of the DSM- I, which comprises 182 disorders. Indeed, between 1952 and 1968 the number of diagnoses again increased substantially, but overall, the DSM-II didn’t differ much from the elementary prototypical approach of the DSM-I.

Finally, in 1980, with the publication of the DSM-III, the American Psychiatric Association dropped its elementary prototype-based approach for a checklist-based system, a system that can still be found in the DSM-5. The main reasons for this switch is likely rooted in the harsh discussions on the fundaments and the practice of psychiatry during the 1960s and 1970s (Decker 2013). During these years, as mentioned earlier, psychiatry was in crisis. In response to this discontent, a restoration movement took shape that was somewhat opposed to psychodynamic theory and practice. A group of so-called neo-Krae- pelinian psychiatrists at Washington University in St. Louis responded to the malaise in the discipline by defining psychiatry more strongly as a strictly medical discipline (Decker 2007, 2013). This group heavily influenced the revisions that gave rise to the DSM-III and provided psychiatry with a redefinition of its object. During the decades in which psychoanalysis was the dominant framework, much attention was paid to unconscious conflict, object relations, and drive-related dynamics in different types of psychopathology. The neo-Kraepelinians shied away from the psychoanalytic discourse, firmly steering toward the use of biomedical concepts and methods with respect to the study and treatment of mental illness. Indeed, as M. Strand (2011) indicates, the principal motive in this shift toward the checklist-based system of the DSM-III can be found in an effort toward “ontological transmutation”: psychiatrists aimed to redefine the controversial object of psychiatry by firmly connecting it to medical discourse. More than ever before, diagnostic classification preceding psychiatric and psychological treatment became a central concern. Below I review how and why this transition took shape in greater detail, focusing on studies that demonstrated the unreliability of prototype-based psychiatric diagnosis.

  • [1] For alcohol use disorder the DSM-5 (pp. 490—491) lists 11 criteria, including: “Alcohol is oftentaken in larger amounts or over a longer period than was intended;” “Craving, or a strong desire orurge to use alcohol;” “Recurrent alcohol use in situations in which it is physically hazardous.”
  • [2] “This psychosis is characterized by disorganized thinking, shallow and inappropriate affect, unpredictable giggling, silly and regressive behavior and mannerisms, and frequent hypochondriacalcomplaints. Delusions and hallucinations, if present, are transient and not well organized” (DSM-II, p. 33).
  • [3] It could be argued that as such, an increase in the number of mental disorders is not necessarilynegative: it could indicate that the profession developed a more refined understanding of mentaldisorders, and thus made an evolution in the direction of more precise diagnosis. However, as wediscuss in Chap. 3, the validity of most DSM categories was, and is, problematic: the numericalincrease of diagnostic categories does not rest on solid grounds.
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >

Related topics