Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Treatment motivation in adults with severe and persistent mental illness (SMI) is a common treatment focus for practitioners working with these clients. Studies among patients with many kinds of chronic illnesses have shown that motivation plays an important role in treatment adherence. However, little is known about the measurement of motivation in the SMI population. The current study analyses the validity and reliability of responses to the Treatment Motivation Questionnaire (TMQ) in a large sample of persons with SMI living in the community. The TMQ is a 25-item instrument that was originally developed with alcohol-dependent clients to assess internal and external motivation for treatment, along with help-seeking characteristics and confidence in treatment.

The TMQ was administered to a quasi-random sample of 469 persons with SMI who were receiving treatment in the public mental health service system in North Carolina. Two exploratory factor analyses (EFA) with minimized residuals were performed in order to assess the number of factors measured by the TMQ in this population, and to identify the specific items measuring these factors. Following the EFA, two separate confirmatory factor analyses (CFA) were performed with polychoric correlations and asymptotic covariances and a weighted least squares estimator, in order to confirm the underlying factors and find the best theoretical and statistical fit of the model in people with SMI. Akaike information criterion (AIC) was used to compare these non-nested models.

Four factor- and five factor-confirmatory analyses were performed in order to compare a theory- based factor solution versus a statistically-based factor solution, respectively. The two models revealed similar factor patterns; however, the overall models differed from published TMQ results from studies of persons with alcohol problems. Specifically, in the SMI sample factor patterns, “Internal” and “External” motivation divided into “Autonomous” and “Guilt” categories in the four factor solution; however, the five factor solution provided an additional dimension of “Supportive Pressure”. This factor consisted of items reflecting treatment pressure from family, friends, and treatment. The AIC indicates that the five factor model (AIC=363.2) provides a better fit than the four factor model (462.6). The reliability of the four primary subscales of “Autonomous Motivation” (α=.77), “Guilt Motivation” (α= .77), “Help Seeking” (α=.87), and “Confidence in Treatment” (α=.82) have respectable to good Cronbach's alpha scores whereas, the fifth subscale, “Supportive Pressure” has poor reliability (α= .47).

The current analyses appear to indicate that people with severe mental illness manifest a somewhat different pattern of treatment motivations than people with alcohol dependence issues. Specifically, the domains of external and internal motivation for the original population were not validated in this study. Instead, motivators such as feelings of guilt and obligation to attend treatment and pressure from supportive social network members factored together to capture a somewhat distinctive pattern of motivation among these consumers' treatment motivations. Indeed, although this scale needs further validation and development, clinical social work practice may be aided through the use and understanding of motivational factors that influence treatment within this population.