An empirical typology of seriously mentally ill patients using symptom and social functioning factors.

Abstract

There is general agreement that, despite the common experience of suffering that the seriously mentally ill share, they are not one group but many different subtypes with varying levels of disability and need. In the last several decades, mental health researchers and service providers have been refining methods of classifying psychiatric patients into homogeneous subgroups based upon their symptom profiles. Criticisms of traditional psychiatric classification and mental health policy demands for specific knowledge regarding the service needs of mental patients have focused greater attention upon the problems of classifying and identifying the mentally ill. Improvements in the specificity of psychiatric treatments as well as innovations designed to match patients with the most appropriate treatment modality have encouraged this effort. The focus of the study was to follow a recently-generated line of research that uses an empirical and statistically-based approach to classifying psychiatric patients. By using the exploratory multivariate techniques of factor and cluster analysis, and applying confirmatory procedures to test the prognostic utility of a given classification scheme, an empirically-based classification scheme was generated to provide prognostic information about patients that may not be available when traditional, symptom-based diagnostic procedures are used alone. The study investigated the effects of using social functioning dimensions as well as symptom data to subgroup seriously mentally ill patients for the purpose of discerning differences between the subgroups on clinically-relevant criteria, such as needs for specific types of treatment. A cluster analysis of seven symptom and social functioning factors yielded five meaningful patient subtypes: "Young Adult Chronic," "High Functioning Chronic," "Mainstream Chronic," "Socially Isolated/Withdrawn," and "High Functioning Psychotic." The groups differed significantly on three of six areas of treatment needs: for residential services, outpatient services, and crisis services. Statistically significant differences were also found on the Resource Associated Functional Level (RAFL) rating. The five subtypes did not differ significantly in terms of DSM-III diagnosis. On sociodemographic variables, the groups differed in terms of education and employment status. When comparisons were made between the DSM-III Axis I diagnostic categories on these same external criteria, no differences were found on any of the six treatment needs or on prior service utilization. The Axis I groups did differ in terms of RAFL rating and employment status. Thus, the clustering scheme yielded information about the patients' treatment needs and service utilization that was not provided by the DSM-III diagnostic categorization. The implications and limitations of the findings are addressed and suggestions are made for further research.