The multiaxial assessment and the DSM-III : a conceptual analysis

Abstract

With the release of the DSM-III, multiaxial assessment, which was a new concept, was introduced to daily clinical practice. This article will review the history and the development of the concept of multiaxial assessment and will focus on the its relationship to the DSM-III. In conclusion I will discuss different critiques of the concept.

Keywords

DSM-III DSM-IV DSM-5 multiaxial assessment multiaxial diagnosis

Methods

The search terms multiaxial diagnosis, multiaxial, DSM-III, DSM-IV and DSM-V were used in the following databases: Pubmed, Google Scholar, PsycInfo. Relevant articles were chosen and read. Further relevant data sources were manually picked and evaluated.

Introduction

The term multiaxial diagnosis was first introduced into general clinical practice with the publication of the DSM-III (APA, 1980). Multiaxial diagnosis was conceived to enable the independent evaluation of several different clinical aspects in order to perform a more comprehensive assessment of patients. The release of the DSM-III along with its newly developed multiaxial system was defined as a ‘fateful’ point in psychiatry (Klerman et al., 1984; Strauss, Yager and Strauss, 1984). However, the more experience the clinical community gained with the multiaxial system, the greater the criticism of it. Therefore, the DSM-5 has introduced radical changes to the concept (APA, 2013). In this article I will review this concept and its ramifications.

The history of the multiaxial assessment

In 1918 Kretschmer published his famous monograph Der Sensitive Beziehungswahn (Kretschmer, 1918). He described a case-series of schizophrenic patients while considering not only their clinical state but also their social environment, their past experiences, their personality variants and possible aetiological factors. This holistic approach was a novelty at the time and planted the seeds for future developments.

The first published attempt to introduce a multiaxial diagnosis was made by Essen-Möller and Wohlfahrt (1947), and a similar proposal was also presented by Lecomte et al. (1947). The work of Essen-Möller and Wohlfahrt had its roots in an earlier work in which the first author addressed the problem of differentiating between organic mental disorder and functional mental disorders (Essen-Möller, 1943). In this earlier paper, he indicated that the differentiation between a functional disorder and an organic disorder is not based on a single homogenous definition but rather on a series of definitions, which are derived from different concepts, namely a descriptive concept and an aetiological concept. For example, in order to diagnose amphetamine intoxication, a drug screen has to be positive. However, in order to diagnose schizophrenia, only a descriptive approach is used. This mixing of different methodologies (contamination) was to him illogical and damaging (Essen-Möller, 1961).

In their joint work Essen-Möller and Wohlfahrt (1947) addressed this problem and suggested a biaxial system in which one axis is dedicated solely to the aetiological factors (if known) and the other is dedicated to the descriptive nature of the disorder, thus establishing a consistent and contamination-free classification system. Their suggestion had a moderate impact on the development and integration of the multiaxial system into daily clinical practice. In 1952 the Danish Psychiatric Association adopted a biaxial system which distinguished between symptoms and aetiologies (Dickmeiss et al., 1952).

In years to come the World Health Organization (WHO) encouraged researchers to develop the concept further using a multiaxial system. Other authors expanded the concept of the biaxial system into a triaxial system (Bilikiewicz, 1951; Lemes Lopes, 1954). However, very little progress was made until the early 1970s, and no specific proposals were made regarding the incorporation of the multiaxial approach into the existing classification systems (Williams, 1985). A noteworthy proposal was offered by Essen-Möller (1961). He expanded his biaxial model into a triaxial model and suggested splitting the descriptive axis into two different axes. One axis would be the ‘syndrome specified’ axis, which would describe the current state of the patient. The other axis would be known as ‘clinic’, and would describe habitual abnormalities. Thus, a personality disorder would be placed on the ‘clinic’ axis while an amphetamine intoxication would be placed on the ‘syndrome specified’ axis. Later, Essen-Möller (1971) suggested that each disorder should be coded using numbers for statistical purposes. During the 1970s a series of specific proposals for the establishment of a multiaxial diagnosis began to appear in the literature (Helmchen, 1980; Kato, 1977; Ottosson and Perris, 1973; Rutters et al., 1969; Strauss, 1975; Von Cranach, 1977; Von Knorring, Perris and Jacobsson, 1978; Wing, 1970). In 1980 the DSM-III was published and was the first official classification to fully develop and integrate a multiaxial system.

The DSM-III and the multiaxial assessment

The concept of multiaxial assessment was conceived 36 years before it was integrated into the DSM. Moreover, multiaxial assessment was not discussed in US psychiatry until the 1970s (Williams, 1985). Despite the lack of evidence and experience, the APA was the first organization to fully adopt the multiaxial assessment. In order to understand this haste, one has to understand the difficult situation of American psychiatry at the time.

After World War II, it was based on a psychosocial model with insight from psychoanalysis, psychodynamics and sociology (Wilson, 1993). The psychosocial model assumed that: (1) the boundary between mental illness and normality is fluid, and anyone can become mentally ill if exposed to a severe enough trauma; (2) mental illness is a continuum from neurosis to borderline conditions to psychosis; (3) a mixture of environmental and psychic conflict causes mental illness; and (4) the mechanism by which mental illness emerges is psychological (psychogenesis) (Grob, 1987). Karl Menninger, a prominent spokesperson for the psychosocial model, wrote in his book, The Vital Balance: ‘Instead of putting so much emphasis on different kinds and clinical pictures of illness, we propose to think of all forms of mental illness as being essentially the same in quality, and differing quantitatively’ (Menninger, 1963: 32–3). One of the major problems of the psychosocial model is that it does not clearly demarcate the well from the sick. By the late 1960s the psychosocial model was attacked both from within and outside the profession.

The antipsychiatry movement

The antipsychiatry movement led by Thomas Szasz claimed that mental illness is in fact nothing but a ‘myth’ (Szasz, 1961). Supported by the fact that psychiatry at the time could not clearly demarcate the sick from the well, Szaz claimed that mental illness was used only to label ‘the career deviant’ (Lemert, 1951: 75–6) and served as a control instrument of the politically dominant over the weak. The antipsychiatry movement managed to protest effectively against the mainstream psychiatry of the 1960s (Dain, 1989; Talbott, 1974). The claim held by the antipsychiatry movement was reinforced by the controversy over the status of homosexuality. The diagnosis of homosexuality seemed to be more of a political issue than a scientific one (Bayer, 1981). Thus in 1973 the diagnosis was omitted from the DSM. This omission has contributed greatly to the delegitimization of psychiatry as a scientific discipline.

Lack of diagnostic validity

In 1953 the APA held a 3-day conference, which focused on developing a research programme for the evaluation of psychiatric therapies. The goal of the conference was to develop comprehensive and sound criteria, methodology or standards under which the validation of a psychiatric therapy can take place. The conference concluded that the efficacy of a treatment was impossible to assess owing to the lack of standardized criteria for both diagnosis and treatment (Pressman, 1987). However, it seemed that American psychiatry in the 1950s and 1960s had little interest in the diagnosis of mental illness. Due to the influence of the psychosocial model, clinicians believed that a descriptive diagnosis was irrelevant and might damage clinical work (Eisenberg, 1986). Robert Spitzer, chairman of the Task Force on Nomenclature and Statistics of the DSM-III, recalls that during the 1960 lectures the field of descriptive diagnosis had very little audience (Wilson, 1993). This lack of interest had a direct negative impact on American psychiatry. During the early 1970s, comparative studies between American and British psychiatry revealed that the former had a bias towards the diagnosis of schizophrenia. Due to the very broad definition used in American psych-iatry, many affective disorder patients were diagnosed with schizophrenia and consequently received inappropriate treatment (Kendell, Cooper and Gourley, 1971). The tendency to diagnose patients with severe diseases was eloquently shown by Rosenhan (1973) in his famous study ‘Being sane in insane places’. In this study, 19 normal subjects admitted themselves to psychiatric hospitals with the single complaint of hearing the word ‘thud’, and during their hospitalization all the subjects behaved ‘normally’. Despite the lack of any visible psychopathology, all were released from the hospitals with the diagnosis ‘schizophrenia in remission’.

Cutbacks in budget, and coping with the difficulties

The consequences of the delegitimization of American psychiatry were beginning to take their toll financially. From 1965 to 1972, the National Institute of Mental Health’s funding decreased by 5% per year (Wilson, 1993). The Federal Employees Health Benefits Program had also made major cutbacks in the resources assigned to psychiatric treatment (Anon., 1977a) due to the apparent lack of reliability and efficacy of psychiatric diagnosis and treatment (Anon., 1977b).

The difficult situation in which American psychiatry found itself in the early 1970s demanded a radical change, and the Task Force apparently understood that this was necessary (Klerman, 1977). Thus the psychosocial model was abandoned and the influence of the psychodynamic movement was greatly reduced. However, the significance of the DSM-III Task Force was in understanding not only that old practices must be abandoned but also that something deeper was flawed in the general approach to psychiatry. The Task Force of the DSM-III understood the immense lack of systematization and methodology in psychiatry. The multiaxial system was the perfect answer to this major problem. It supplied an instrument to assess methodologically and document every aspect of the patient and thus provide an answer to many of the criticisms mentioned above.

The designing and the planning of the multiaxial system of the DSM-III

The first draft of the DSM-III, published in March 1976, delineated a multiaxial system for the childhood disorder:

A multi-axial approach to the classification of the psychiatric disorders of childhoods under deliberation by the committee …. The first axis would describe the clinical psychiatric syndrome. The second axis would describe the intellectual level of the patient. The coding on this axis would indicate the individual’s current level of intellectual functioning without regard to its nature or its causation. On the third axis specific development disorders would be coded. … The fourth axis would denote the associated biological factors which might be present. The fifth axis would code associated psychological factors. No well-developed classification scheme has been worked out for the coding of psychosocial factors. (APA Statistics, 1976)

The introduction of the multiaxial diagnosis to the task force was made by Dennis Cantwell, MD (Williams, 1985). He worked with Michael Rutter, a pioneer in the field of multiaxial classification and later an advisor to the committee responsible for the multiaxial system of the DSM-III. After the preliminary draft, Spitzer decided to adopt this approach to the adult psychiatry section, and so a committee for the development of a multiaxial diagnosis was appointed. It is noteworthy that J.B. Williams, a member of the committee in charge of the multiaxial system of the DSM-III, noted that the majority of the members of the committee were only introduced to the concept of multiaxial diagnosis in 1976, i.e. only four years prior to the release of the DSM-III (Williams, 1985).

Although some field trails regarding the use of the multiaxial system were completed before the publication of the DSM-III (Spitzer and Forman, 1979), the multiaxial system of the DSM-III was mostly based on theoretical considerations. The committee decided on a pentaxial model: axis I for the clinical syndrome, axis II for personality disorders and specific developmental disorders, axis III for physical disorders and conditions, axis IV for psychosocial stressors, and axis V for global functioning. This multiaxial system was designed in order to ensure that different areas of important information, according to the newly adopted biopsychosocial model, are systematically evaluated and recorded (Williams and Spitzer, 1982). The committee considered that the model should not be too complicated for everyday clinical practice and at the same time should represent a reasonably comprehensive assessment of patients (Williams, 1981).

Critique of DSM-III’s multiaxial assessment

The concept of a multiaxial diagnosis per se was generally not refuted, but the design of the multiaxial diagnosis was and still is a topic of an ongoing debate. During the early 1980s there seemed to be a consensus that the dichotomy of axis I and axis II was justified (Frances and Cooper, 1981; Nakdimen, 1981). The critique directed towards axis I and axis II did not challenge the dichotomy but rather the content of each axis (Kendell, 1980; Pichot, Guelfi and Kroll, 1983; Rutters and Shaffer, 1980). Axis III was dedicated to general medical conditions that are potentially relevant to the understanding and management of the individual’s mental disorder – a task with ill-defined borders (Roth, 1983).

Axis IV was for psychosocial stressors which might contribute to the current clinical state. Since different people respond to stressors in different ways, it is not clear what should be taken into consideration when addressing the patient’s problems (Frances and Cooper, 1981; Kendall, 1983). For example, a job promotion might be desired by one patient but would be a heavy burden for another. Axis V was dedicated to assessing the highest level of functioning using the Global Assessment of Functioning. The axis was criticized for not taking into account the duration of the symptoms, but was rather a superficial reflection of the patient’s current condition (Kendell, 1980). A patient with a dementia at its early stage might have a higher GAF score than a psychotic patient, due to medication. However, the prognosis of the former is bleaker than that of the latter. Over the years, the reliability and clinical utility of the GAF have been subjects of intense research, which has shown mixed results (Jones et al., 1995; Roy-Byrne et al., 1996; Startup, Jackson and Skodol, 2002).

Among the advisors to the multiaxial committee, J.S. Strauss, MD, advocated a much more cautious approach regarding the introduction of the multiaxial assessment. Strauss believed that instead of using a categorical approach, a continuum or a dimensional approach would be preferable. Strauss also thought that the introduction of the multiaxial assessment should be delayed until more research data could be gathered (Strauss, 1975). The changes that were made in the DSM-5 make one wonder where present-day psychiatry would be if the APA had heeded Strauss’s advice.

The changes in the DSM-5

The DSM-5 has removed the multiaxial assessment, an action which has prompted mixed feelings. DSM-5 has united axes I, II and III into a single entity, thus abolishing what was regarded as one of the important achievements of the multiaxial assessment. This approach is consistent with established WHO and ICD guidelines that consider the individual functional status as separate from clinical symptoms. Axis V will now be assessed with V-Codes, which are an earlier version of the ICD-10 Z-Codes. The DSM-5 has abandoned the use of the GAF and has adopted the WHO Disability Assessment Schedule, which has shown much more promising results (Garin et al., 2010; Janca et al., 1996). This might also be explained by the tendency to harmonize the ICD and the DSM.

The trend towards harmonization of classification systems and better patient care might be the official explanation as to why the multiaxial system was discarded. However, as this article proposes, other powers were at play during the compiling of the original multiaxial assessment; therefore one must consider the likelihood that other powers were involved in its removal.

Critique on multiaxial systems in general

The clinical necessity to provide a thorough evaluation and treatment plan using the multiaxial system is often impeded by considerations of practicality (Mezzich, 1988). In spite of the much more extensive research and planning that were invested in developing the multiaxial system of other psychiatric classifications such as the ICD-10, similar critiques regarding the nature of each axis were mounted (Maj et al., 2002: 163–77). In a large international survey held by the WHO, 68% of the respondents found the multiaxial system to be of value (Mezzich, 1999). A concern raised among the participants in the survey was the lack of formal training needed for using a multiaxial system.

Conclusion

In spite of the criticisms and the extensive changes that were made to the original multiaxial assessment, its introduction signified a giant step in terms of methodology. The multiaxial assessment represents a transition from a narrow perspective of the patient’s problem to a holistic approach. The introduction of a systematic evaluation had a great influence on psychiatry as well as on other fields. In reviewing the history of the multiaxial assessment, the current changes in the DSM-5 are not clear, and only time will tell if they were justified.

References

American Psychiatric Association [APA] (1980) Diagnostic and Statistical Manual of Mental Disorders (3rd ed.). Washington, DC: APA.

American Psychiatric Assocation [APA] (2013) Diagnostic and Statistical Manual of Mental Disorders-5 (5th ed.). Arlington, VA: APA.

American Psychiatric Assocation [APA] Statistics (Task Force on Nomenclature and Statistics) (1976) Progress Report on the Preparation of DSM-III (Archives, DSM Coll.).

Anonymous (1977a) Burden of MH coverage questioned at conference. Psychiatric News.

Anonymous (1977b) Javits says MH hurt by role confusion. Psychiatric News.

Bayer

(1981) Homosexuality and the American Psychiatry: The Politics of Diagnosis. New York: Basic Books.

Bilikiewicz

(1951) Próba ukadu nozograficznego etioepigenetyczengo w psychiatrii [Layout attempt in the nosography and etioepigenetics of psychiatry]. Neurologia i Neurochirurgia Polska 13: 68–78.

Dain

(1989) Critics and dissenters: reflections on “anti-psychiatry” in the United States. Journal of the History of the Behavioral Sciences 25: 3–25.

Dickmeiss

P et al.

(1952) Udarbejdet af Dansk Psykiatrisk Selskabs Diagnoseudvalg [Prepared by the Danish Psychiatric Association Diagnostic Committee]. Nordisk Psykiatrisk Medlemsblad 6: 277–295.

10.

Eisenberg

(1986) Mindlessness and brainlessness in psychiatry. British Journal of Psychiatry 148: 497– 508.

11.

Essen-Möller

(1943) Über den Begriff des Funktionallen und Organischen in der Psychiatrie. Acta Psychiatrica Scandinavica 18: 1–44

12.

Essen-Möller

(1961) On classification of mental disorders. Acta Psychiatrica Scandinavica 37: 119–126.

13.

Essen-Möller

(1971) Suggestions for further improvement for the international classification of mental disorders. Psychological Medicine 1: 308–311.

14.

Essen-Möller

Wohlfahrt

(1947) Suggestions for the amendment of the official Swedish classifcation of mental disorders. Acta Psychiatrica Scandinavica 22 (Suppl. 47): 551–555.

15.

Frances

Cooper

(1981) Descriptive and dynamic psychiatry: A perspective on the DSM-III. American Journal of Psychiatry 138: 1198–1202.

16.

Garin (2010) Research validation of the “World Health Organization Disability Assessment Schedule, WHODAS-2” in patients with chronic diseases. Health and Quality of Life Outcomes 8: 51.

17.

Grob

(1987) The forging of mental health policy in America: World War II to the New Frontier. Journal of the History of Medicine 42: 410–446.

18.

Helmchen

(1980) Multiaxial systems of classification. Types of axes. Acta Psychiatrica Scandinavica 61: 43–45.

19.

Janca

A et al.

(1996) The World Health Organization Short Disability Assessment Schedule (WHO DAS-S): a tool for the assessment of difficulties in selected areas of functioning of patients with mental disorders. Social Psychiatry and Psychiatric Epidemiology 31: 349–354.

20.

Jones

S et al.

(1995) A brief mental health outcome scale-reliability and validity of the Global Assessment of Functioning (GAF). British Journal of Psychiatry 166: 654–659.

21.

Kato

(1977) Multiaxial diagnosis in adult psychiatry. Paper presented at the Sixth World Congress of Psychiatry, Honolulu.

22.

Kendell

(1980) Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. American Journal of Psychiatry 137: 1630–1631.

23.

Kendell

(1983) DSM-III: A major advance in psychiatric nosology. In: Spitzer

Williams

JBW

Skodol

(eds) International Perspectives on DSM-III. Washington DC: American Psychiatric Press, 122–139.

24.

Kendell

Cooper

Gourley

(1971) Diagnostic criteria of American and British psychiatrists. Archives of General Psychiatry 25: 123–130.

25.

Klerman

(1977) Mental Illness, the medical model, and psychiatry. Journal of Medical Philosophy 2: 220–243.

26.

Klerman

G et al.

(1984) A debate on the DSM-III. American Journal Psychiatry 141: 539–553.

27.

Kretschmer

(1918) Der Sensitive Beziehungswahn. Berlin: Springer Verlag.

28.

Lecomte (1947) Essai d’une staistique synoptique de médicine psychiatrique. Techniques Hospitaliers 18: 5–8.

29.

Lemert

(1951) Social Pathology: A Systemic Approach to the Theory of Sociopathic Behavior. New York: McGraw-Hill.

30.

Lemes Lopes

(1954) As dimensões do Diagnóstico Psiquiátrico. Rio de Janeiro: Agir; in Portuguese.

31.

Maj

M et al.

(2002) Psychiatric Diagnosis and Classification. Chichester, UK: John Wiley & Sons.

32.

Menninger

(1963) The Vital Balance. New York: Viking Press.

33.

Mezzich

(1988) On developing a psychiatric multiaxial schema for ICD-10. British Journal of Psychiatry 152 (Suppl. 1): 38–43.

34.

Mezzich

(1999) A preliminary report on the International Survey on the Use of ICD-10 (Paper presented at the WPA Classification Section Symposium on Empirical Assessment of ICD-10). XI World Congress of Psychiatry. Hamburg.

35.

Nakdimen

(1981) DSM-III’s multiaxial system: political finesse. American Journal of Psychiatry 138: 259.

36.

Ottosson

Perris

(1973) Multidimensional classification of mental disorders. Acta Psychiatrica Scandinavica 3: 238–243.

37.

Pichot

Guelfi

Kroll

(1983) French perspectives on the DSM-III. In: Spitzer

Williams

JBW

Skodol

(eds) International Perspectives on DSM-III. Washington DC: American Psychiatric Press, 155–173.

38.

Pressman

(1987) Uncertain promise: psychosurgery and the development of scientific psychiatry in America, 1935 to 1955 [1986]. Dissertation Abstracts International 47: 2716.

39.

Rosenhan

(1973) Being sane in insane places. Science 179: 250–258.

40.

Roth

(1983) The achievements and limitations of DSM-III. In: Spitzer

Williams

JBW

Skodol

(eds) International Perspectives on DSM-III. Washington, DC: American Psychiatric Press, 91–105.

41.

Roy-Byrne

P et al.

(1996) Evidence for limited validity of the Revised Global Assessment of Functioning Scale. Psychiatric Services 47: 864–866.

42.

Rutters

. (1969) A triaxial classification of mental disorders in childhood. Journal of Child Psychology and Psychiatry 10: 41–61.

43.

Rutters

Shaffer

(1980) DSM-III: a step forwards or back in term of the classification of child psychiatric disorders? Journal of the American Academy of Child Psychiatry 19: 371–394.

44.

Spitzer

Forman

(1979) DSM-III field trials: II. Initial experience with the multiaxial system. American Journal of Psychiatry 136: 818–820.

45.

Startup

Jackson

Bendix

(2002) The concurrent validity of the Global Assessment of Functioning (GAF). British Journal of Clinical Psychology 41: 417–422.

46.

Strauss

Yager

Strauss

(1984) The cutting edge in psychiatry. American Journal of Psychiatry 141(1): 38–41.

47.

Strauss

(1975) A comprehensive approach to psychiatric diagnosis. American Journal of Psychiatry 132: 1193–1197.

48.

Szasz

(1961) The Myth of Mental Illness: Foundations of a Theory of Personal Coduct. New York: Harper & Row.

49.

Talbott

(1974) Radical psychiatriy: an examination of the issues. American Journal of Psychiatry 131: 121–128.

50.

Von Cranach

(1977) Categorial vs multiaxial classification. Paper presented at the Sixth World Congress of Psychiatry, Honolulu.

51.

Von Knorring

Perris

Jacobsson

(1978) Multiaspect classification of mental disorders: experiences from clinical routine work and preliminary studies of interrater reliability. Acta Psychiatrica Scandinavia 58: 401–412.

52.

Williams

(1981) DSM-III: a comprehensive approach to diagnosis. Journal of Social Work 26: 101–106.

53.

Williams

(1985) The Multiaxial System of DSM-III: where did it come from and where should it go? I. Its origins and critiques. Archives of General Psychiatry 42: 175–180.

54.

Williams

Spitzer

(1982) DSM-III forum: focusing on DSM-III’s multiaxial system. Hospital & Community Psychiatry 33: 891–892.

55.

Wilson

(1993) DSM-III and the transformation of American psychiatry: a history. American Journal of Psychiatry 150: 399–410.

56.

Wing

(1970) Observations on the psychiatric section of the International Classification of Diseases and the British Glossary of Mental Disorders. Psychological Medicine 1: 79.