Exploring the diversity of dual diagnosis: utility of cluster analysis for program planning

This study demonstrates the utility of using cluster analysis to explore the heterogeneity of dual diagnosis populations so as to facilitate planning and implementation of individualized treatment programs. A sample of 467 persons admitted to a state psychiatric hospital with DSM-III-R psychiatric diagnoses and substance abuse problems were interviewed on the Addiction Severity Index (ASI) and other measures to assess psychological, social, and community functioning. Scores on seven ASI severity ratings (medical, employment, alcohol, drug, legal, family, and psychiatric functioning) were used to group patients into seven homogeneous subgroups using cluster analysis: best functioning, unhealthy alcohol abuse, functioning alcohol abuse, drug abuse, functioning polyabuse, criminal polyabuse, and unhealthy polyabuse. Cluster reliability and validity were demonstrated using split-half tests as well as cross-sectional and longitudinal analyses. Results illustrate the extreme heterogeneity of dual diagnosis and are suggestive of how individualized treatment programs can be matched to the particular needs of patients with dual diagnoses.