Classifying psychosis: when is the time ripe for changes?
Year: 2007
Type: article
Abstract: The question about whether psychotic symptoms have an underlying taxonic (categorical) or dimensional structure is not an easy one. Even though diagnostic categories are widely used and form the basis of current classifications, it can be argued that a dimensional conceptualization of psychosis has more powerful empirical support, since there is evidence that symptom dimensions predict treatment response and outcome better than categorical diagnoses (1-3). However, even if the explanatory power of dimensions were better than what it is, there would still be a problem with their use in epidemiology, where the interest is in counting, and certainly in their use as robust phenotypes for biological studies. It would therefore seem that a parsimonious approach to the classification of psychosis is what can be supported by current evidence (4). Combining categorical classification with dimensional assessment offers the possibility of improving management of the individual patient, but the parsimony offered by categories is likely to be attractive for a long time to come.
Kraepelin's dichotomy was developed on the basis of what he saw among patients in asylum. Suggestions for revision of the dichotomy are being made on the basis of what we can see among samples which are more varied and diverse than those seen by Kraepelin. Given the common occurrence of psychotic symptoms in the community, with several people experiencing them not making contacts, can we be sure that whatever revision of our current classification we carry out will be applicable across the totality of the psychotic experience? Many previous studies of schizophrenia spectrum disorders (5-9) have proceeded from the standpoint of their symptoms being qualitatively and quantitatively different from common experience. If we conceive of the psychotic experience as a dimensional one or a continuum (10), then it may be necessary to measure the symptoms as a continuous phenomenon. That would require the study of a much broader profile of psychotic symptoms than are presented in settings. The same sort of suggestions has often been made for dealing with the continuity controversy in depression (11).
Nevertheless, several of the properties of a classification system identified by Craddock and Owen are indeed desirable. For example, the development of clinical, research and primary care versions of ICD-10 would seem to have been done in order to make the classification flexible for different needs (12). The multiaxial format was an attempt to use multiple descriptors to capture the status of the individual patient. However, the suggestion that clinical utility requires biological validity is probably a futuristic goal for psychiatry. If utility is defined mainly by what the clinician can do for the patient that waits for treatment in front of him, then it is difficult to see how current biological validations of the different types of schizophrenia spectrum can affect treatment choice. Isn't it the case that antipsychotics remain the treatment of choice for these disorders? And isn't it the case that current knowledge suggests that there are more similarities than differences between those medications, irrespective of the claims made by pharmaceutical companies? Indeed, the difficulty in following recent lines of evidence suggesting a re-classification of psychotic disorders, as competently reviewed by Craddock and Owen, is the fact that utility, that is, the hereand- now management of patients, is not necessarily influenced by that evidence. The controversy over whether to lump or to split is not going to be resolved until psychiatry has more than the blunt instruments it currently has to cut nature at its joints (13). Any classification that is based on aggregation of symptoms and signs without a validating set of biological features will be prone to arbitrary shifting of the of rarity, because such points will be defined in so many different ways by people with different needs for classification. Still, there is a recognized and subsisting need to derive phenotypes to aid in the search for susceptibility genes and in the identification of robust biological features that could help in identifying risk, targeting treatment, and offering more reliable prognosis. That process will be along the path of identifying specific neuropathologic processes underlying specific clusters of symptoms and signs. The process is likely to be greatly enhanced by the identification of endophenotypes which are likely to have better discriminating power than the phenotypes we currently employ.
Source: Europe PMC (PubMed Central)
Author: Oye Gureje
Institutions: University of Ibadan, University College Hospital, Ibadan
Language: English
Subfield: Psychiatry and Mental health
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W1827463789
