Disorders are different from diseases

Pincus et al correctly point out that what is often called "the co-morbidity problem" is unavoidable, because it is simply a fact of life in clinical psychiatry. They provide a useful discussion of the topic, but the very use of the conventional term 'co-morbidity' serves to hide the real nature of the problem. This is because 'morbid' means disease, and to have a disease is conceptually very different from suffering from a disorder. Strictly speaking, the terms 'diagnosis' and 'disease' are both best avoided in psychiatric discourse unless they are completely justified. Clinical psychiatrists make few diagnoses in the sense of identifying known abnormalities which underlie the presenting symptoms. Instead, for most patients they have to make do with identifying disorders by assessing the number and severity of individually non-diagnostic symptoms from an agreed list. Most currently recognised disorders are no more than symptom clusters, and there is no particular reason why most patients should be expected to have only one of these. Viewed in this way, it is clear that it would be more honest for psychiatrists to use other terms, such as 'co-existing disorders' or 'multiple disorders'.

On the basis of the points just made, it is natural to wonder why the inappropriate term 'co-morbidity' has become accepted usage. Probably it is a hang-over effect from the vitally important general medical training that all psychiatrists undergo, during which it is easy to develop the expectation that most patients have only one diagnosable disease. But things are different in psychiatry, and surely it is best to use more realistic terms that are a constant reminder that our knowledge of the nature of psychiatric illnesses is rather superficial.

Two other associated issues are: first, the frequent lack of clarity in clinical work about the purposes for which the information is being recorded, and second, the special needs of researchers.

In the patient's clinical case record, good practice requires that as many disorders should be recorded as are needed to describe the complete state of the patient, and this instruction is common to both ICD-10 and DSM-IV. In addition, clinicians should be encouraged always to give an order of priority for the disorders present, the reasons for this order, and the actions required by each disorder. If for some reason only one disorder can be recorded, it is up to those requiring the information to make clear the main purpose for which it will be used. Recording systems are now far more powerful than they were even a few years ago, and the old administrator's plea of 'no room on the form' has become a weak excuse.

The needs of researchers are often different from those of clinicians, since most types of research involve the restrictive selection of groups of individuals whose symptoms and other characteristics resemble each other in clearly stated ways. Whether or not it is appropriate to include patients with more than one disorder will be determined by the type and purposes of the study, and so a more flexible approach to exclusion criteria is needed in research than is the case for ordinary clinical recording. This is why, for ICD-10, the Diagnostic Criteria for Research (1) are published separately from the Clinical Descriptions and Diagnostic Guidelines. While on this topic of criteria for research, it needs to be pointed out that the comment of Pincus et al that "the ICD-10 diagnostic criteria for research were largely modelled on the DSM-III system" is a somewhat approximate précis of a long and complicated process. The many similarities between ICD-10 and DSM-IV in both general style and detailed content is the purposeful end result of an initiative started as long ago as 1980 by Gerald Klerman and Norman Sartorius, in the form of a 'Joint Project'. The final manifestation of this was a series of meetings around 1990 between World Health Organization (WHO) advisors and the Chairpersons of the Task Forces for DSM-IV at which many harmonising changes were agreed to the drafts of both the classifications.

As a thought for the future, new ways of recording multiple disorders should be tried out as new versions of the classifications are developed, rather than leaving the process of recording as an afterthought to be worked out only when the classifications have been finished.