This chapter will deal with issues of classification most important to liaison psychiatrists. At the time of writing, the two main psychiatric classification systems, one developed by the American Psychiatric Association (APA) and the other by the World Health Organization (WHO), were undergoing a period of transition, with the development of new classification manuals that are due to be published in the near future. The emphasis in this chapter will not be on describing the current classificatory systems in detail but highlighting problems with selected categories and some proposed solutions. Thus it is hoped that the chapter remains relevant as a record of underlying tensions and problems even when the new classificatory systems are published.

Psychiatric classification

Psychiatric classification, like other medical classifications, aims to carve nature at the joints. Earlier psychiatric classificatory systems, akin to the botanical and zoological counterparts of the day, were lacking in a priori classificatory principles and classified disorders according to whatever presenting characteristics appealed to the classifier (Kendler, 2009). Scientific rigour was injected into psychiatric classification with the classification of psychoses by Emil Kraepelin based on course and outcome; when a group of biologically minded American psychiatrists revolutionised psychiatric classification with the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III; American Psychiatric Association, 1980), the members of the group were described as neo-Kraepelinians (Klerman, 1978).

The WHO has also published a psychiatric classification, the International Classification of Diseases, currently in its tenth version (ICD-10; World Health Organization, 1992). This was initially developed for coding purposes for research, but it has developed into an internationally accepted classificatory system (Stengel, 1959). There has been a great deal of convergence between the DSM and ICD systems in their latest versions and both provide operational definitions of disorders. The advantage of operational diagnosis has been in an improvement in reliability, while the criticism is that it encourages a field-guide approach to psychiatric diagnosis (McHugh ' Clark, 2006).