Positive Relationships: A Comparison of Consumer Perceptions of Working Alliance with Case Managers and Therapists

Background/Purpose:  Research consistently demonstrates that a quality working alliance (WA) between a consumer and therapist predicts a positive outcome (Horvath & Symonds, 1991).  Because case management is the primary psychosocial treatment for persons with severe mental illness (SMI) in the community (Rose & Moore, 1995), researchers have studied the impact of the WA within this type of practice.  An assumption in most of the case management research is that what constitutes a positive alliance is the same as in therapy, which suggests that both groups should be equally able to develop positive WAs.  This study explored how consumers in case management and individual outpatient therapy rated the quality of the working alliance. 

Methods:     Data were collected from two groups of consumers living with a SMI; one group receiving intensive case management (n = 121), and the other group receiving psychotherapy (n = 45). Each consumer rated their relationship with their provider by filling out the Working Alliance Inventory (WAI).  The researchers used the average item score of the WAI as the dependent variable.  Additional data collected from consumer included race, age, sex, length of time in treatment, and level of education.  Consumers were nested in providers.  To account for the nested data structure and to the potential threat of autocorrelation (Luke, 2002), the researchers used multi-level modeling.  Consumers were level one (n = 166) and providers were level two (n = 37).  The average number of consumers per provider was 4.5 and the range was 2 to 15.  Type of practice, case management or psychotherapy, was set as an attribute of providers.  There was a small amount of missing data; the researchers replaced the missing data using a multiple imputation.

Results:  The researchers estimated three models.  The first model explored the effects of providers on consumer perceptions of WA.  Providers accounted for 20% of the variance in WA scores.  In the second model, the researchers included all consumer level variables, and this model accounted for 30% of the variance in WA scores.  Lastly, the researchers estimated a model that included type of practice.  The third model accounted for 53% of the variance in WA scores, and type of practice (b = -.87, df = 35, p < .05) was significantly related to WA scores.  Consumers rated their WA with therapists more positively than with case managers.

Conclusions/Implications:  Results showed significant differences in consumer perceptions of relationship quality between case managers and therapists.  This finding is important, as previous research has assumed that both groups are equally capable of developing positive alliances with consumers.  Future research should use caution when assuming that results related to the therapy relationship are transferable to the case management relationship.  Further, the results suggest that case managers, most of whom are not professionally educated, need to be specifically trained to develop positive relationships with their consumers.  Future research should explore differences in WA scores between professionally trained case managers, such as social workers, and on-the-job-trained case managers.