Research That Matters (January 17 - 20, 2008)

Purpose: Employment serves as an important aid in the recovery process for persons with mental illness, helping mental health consumers focus on positive aspects of their lives and reduce negative behaviors, such as substance use (Mueser, Noordsy, Drake, & Fox, 2003). Supported Employment has been recognized as an effective method of assisting persons with severe mental illness secure and maintain competitive employment (Bond, 2004). Less is known, however, about the impact of Supported Employment among individuals with co-occurring mental and substance use disorders, the focus of this paper. Data are presented on the relationship between consumers' characteristics (demographic, socio-economic, illness and wellness) at baseline and consumers' subsequent service usage, employment status, and patterns of employment. Methods: Consumers in the study with co-occurring substance and mental disorders were drawn from Integrated Dual Diagnosis Treatment (IDDT) programs in four community mental health agencies. Study participants included 103 consumers who were referred for Supported Employment services. This study utilized baseline demographic, illness, and wellness measures collected through consumer questionnaires, case manager assessments, and administrative records. Data on hospitalization service utilization, employment status, and patterns of employment were collected monthly from agency staff. Results: The mean age of the sample was 38.5 years. Sixty-five percent were male and 65% of the sample were Caucasian. One-third had less than a high school education. Eighty-five percent of consumers reported previous work experience at baseline. Of consumers referred to Supported Employment, 83% engaged in services, averaging 64 minutes per month. Analysis of employment status and patterns was limited to only those consumers who received services (n=85). Two-fifths (41%) of consumers entered competitive employment, averaging $7.98/hour at 20 hours/week. Variables that were positively associated with employment entry included lower drug and alcohol use, less psychiatric distress, and less hospitalization. Employment entry also varied by agency location. A logistic regression model to predict employment entry was designed utilizing the significant variables. A test of the full model versus the model with intercept only was statistically significant χ2 (7, N = 85) = 32.1, p < .001. Employing a .05 criterion of statistical significance, agency location had a significant partial effect in the direction of predicting employment entry. Implications: The rate of Supported Employment service utilization was somewhat lower than the average rate reported in nine previous clinical trials of Supported Employment, 41% vs. 59%, though none of these previous trials focused specifically on adults with co-occurring substance and mental disorders (Bond, 2004). As compared to these previous studies, the employment rates for the four Supported Employment programs in the current study showed that one agency was substantially above the average rate, one agency was somewhat below the rate, and the other two agencies were substantially below the rate. Differences in employment rates among the agencies may be a result of differences in fidelity to the model or possibly a reflection of differences in clinical skills among the Employment Specialists at different agencies (Drake, Bond, & Rapp, 2006). These findings suggest additional research on staff practices would be informative.