One of the main advantages of classificatory systems is that they enable doctors and psychiatrists to communicate easily about problems, prognosis and treatment (Gelder et al, 1989); taxonomic systems also have predictive value, i.e. they set up categories the mere membership of which makes the possession of certain attributes highly probable (Kleinmutz, 1980). If a psychiatrist tells a colleague that a patient is, for example, a manic depressive according to DSM III, the colleague will know what symptoms and behaviour to expect from the patient, how to treat him and how the illness is likely to develop. Such communication skills are essential in treating patients, and without a classificatory system this would be very difficult.
Classification systems also have practical and scientific usefulness. They enable many important decisions to be made about patients: in legal determination of insanity, in ward and treatment assignments, as a screening device in the military and industry, and in the compilation of census figures (Kleinmutz, 1980). Eysenck (1960) says that before one can look for the cause of dysfunction, one must isolate, however crudely, the dysfunction in question, and be able to differentiate it from other syndromes: this is greatly facilitated by the use of a classification system. Meehl (1959) also emphasises the practical importance of classification - "there is a sufficient amount of aetiology and prognostic homogeneity among patients belonging to a given diagnostic group, so that the assignment of a patient to his group has probability implications which it is clinically unsound to ignore". Without classificatory systems, it would be impossible to make important practical decisions about patients, and this consideration alone must outweigh any disadvantages of the application of the system.
There are, however, several disadvantages to classificatory systems. One of the most often cited is their focus on labelling patients. Labelling theory, as put forward by some sociologists, states that to allocate a person to a diagnostic category is simply to apply a label to a deviant behaviour (Scheff, 1963), and that such labelling only increases a person's difficulties, because some psychiatric terms attract stigma. Simply naming a disorder does not help to cure it. Labelling also detracts from an understanding of each patient's unique abilities, and indeed many patients do not fit neatly into the available categories. Szasz (1960) suggests that we classify in order to gain control: by labelling, the psychiatrist legitimises the social aspirations and roles of others. Some labels, such as psychosis, are actually only used as convenient terms for disorders that cannot be given a more precise definition (Gelder et al, 1989), and would seem to serve only to dehumanise patients further. It is more important to consider each case individually, taking into account the history and experience of the patient, his interaction with society and his needs.
Kleinmutz (1989) criticises the artificiality and reification of classification, saying that categories often correspond poorly with the actual experience of individual patients, and that clinicians are apt to make assumptions that alter their perception of, and behaviour towards, patients - that this leads the patient to act out the part in which they have been cast, i.e. a self-fulfilling prophecy. A patient when told, for example, that he is schizophrenic, may 'suddenly' begin to experience auditory hallucinations, as he is aware (consciously or subconsciously) that these are the symptoms of schizophrenia. Goffman (1961) saw this self-fulfilling prophecy at work, through the assignment of troublesome patients to 'bad' wards, which led to worse, psychotic behaviour. Szasz (1960) points out that as we create the categories, how can we be sure that we have them correctly placed? From this point of view, classification systems would appear to be rather artificial; it is highly possible that psychiatrists, once they have assigned a patient to a particular category, tend not to 'see' symptoms which would refute their decision, and this may well be one of the causes of the high unreliability of diagnostic rates that have been identified.
This unreliability of diagnosis between psychiatrists is perhaps the main criticism of classification systems. Eysenck (1960) puts the problem like this - "The psychiatrist typically bases his view on unsystematic observations of small and accidental samples of an undefined population - these observations are then integrated in accordance with unformulated canons of procedure along lines open to influence of bias and preconceived opinion, and not subject to external safeguards and checks." This view is supported by evidence. In a 1956 study, a pair of psychiatrists independently diagnosed 426 patients using APA 1952 (American Psychiatric association). It was found that the more specific the diagnostic categories, the less reliable the diagnosis: schizophrenia had a correlation of 51%, neurosis a correlation of only 16%. Cultural differences have also been found: schizophrenia is diagnosed twice as often in the USA than in the UK, and vice versa for depression (Cooper et al, 1972). Ward et al (1962) concluded that of overall disagreement, 62% was due to inadequate use of diagnostic criteria. However, such studies have been criticised (e.g. Buss, 1966) because they have not been adequately controlled; they only demonstrate that classification systems are not used properly, and do not point out difficulties in the systems themselves.
Such difficulties become manifest in practical situations, as three 1970 papers demonstrate. Kendell and Gourlay (1970b) state that although the distinction between affective psychoses and schizophrenia has been the cornerstone of psychiatric classification, it is recognised that some patients have symptoms of both: do such cases constitute a third group of psychoses (degeneration psychoses), or are they genuine mixed states? Classificatory systems presently in use seem to provide no clear answer to such a problem. The case regarding psychotic and neurotic depressions is similar (Kendell and Gourlay, 1970a): a study reported a misclassification rate of 14% between the two, and the argument as to whether they are separate entities of a continuum is not solved by reference to classificatory systems. Eysenck (1970) is against categorical allocation, on the basis that he found neuroticism and psychoticism to be continuous variables; he would prefer a dimensional system. These papers help to illustrate the practical difficulties faced by psychiatrists in diagnosing cases for which the present systems provide little help.
The case in favour of classificatory systems would seem to be rather weak in the light of the above evidence: the labelling problem, artificiality, unreliability and their practical repercussions would appear to be insurmountable difficulties of the use of classificatory systems. However, these problems are usually the result of misuse of such systems, rather than the systems themselves: there is no contradiction between labelling a patient and also attempting to understand him as an individual, and the inability of some psychiatrists to spot certain abnormalities is not strong evidence against the existence of a condition (Meehl, 1959). The answer to the problems of classification systems lies in improving procedures for reaching a diagnosis, perhaps by the use of functional (factor) analysis - to view the conjunction of symptoms with respect to their relation to one another (See Eysenck 1960, and Owens and Ashcroft, 1982). However it is achieved, it is essential to have one or another classificatory system in use, if only for communication value; psychiatrists would otherwise not be able to decide on diagnosis, prognosis and treatment with any degree of confidence.
Approx 1,500 words.
Bibliography
Boehm in Coleman (1972) Abnormal Psychology and Modern Life, Illinois.
Cooper et al (1972)
in Gelder et al (1989) Oxford Textbook of Psychiatry, Oxford.
Eysenck (1970) 'Classification of Depressive Illness', British Journal of
Psychiatry, vol. 117.
Goffman (1961) in
Kleinmutz (1980) Essentials of Abnormal Psychology, New York.
Kendell and Gourlay
(1970a) 'Clinical Distinction between Psychotic and Neurotic
Depression', British Journal of Psychiatry, vol.117.
Meehl (1959) in
Kleinmutz (1980) Essentials of Abnormal Psychology, New York.
Spitzer and Wilson (1975)
in Coles (1982) Clinical Psychopathology, London.
Szasz (1966) in
kleinmutz (1980) Essentials of Abnormal Psychology, New York.
Ward et al (1962)
in Gelder et al (1989) Oxford textbook of Psychiatry, Oxford.
