This longitudinal study was designed to refine the search for reliable predictors of premature termination from adult psychotherapy by examining therapy-specific and contextual factors that may predict clients' decisions to leave treatment early. These variables included client's stage of change prior to treatment, client's level of self-determination or motivation for therapy, client and therapist expectations and any discrepancies between them, the therapeutic alliance, and practical barriers to treatment participation. To facilitate comparison with previous research, premature termination was operationalized in two manners, duration of treatment and type of termination decision (unilateral versus mutual agreement with the therapist), although type of termination decision was considered a better measure of premature termination. These variables were examined with self-report data from 107 client/therapist dyads at a community mental health clinic. Data were collected prospectively and retrospectively from both client and therapist perspectives before therapy began, after the third treatment session, and after therapy ended.
Logistic regression analyses showed that barriers to treatment participation, measured at the end of therapy, and working alliance, measured after the third treatment session, predicted type of termination from the client perspective, and barriers predicted type of termination from the therapist's perspective. The therapist's perception of barriers to treatment added incremental predictive value to that of the client's. The client's perspective on barriers to treatment also enhanced the predictive value of the therapist's perception of barriers. The discrepancy between client and therapist expectations for therapy duration did not predict the nature of the termination decision. When a client had a low self-determination index score, or was non self-determined, but had expectations of longer therapy, he or she tended to make a unilateral decision to end therapy. If a client, who perceived a poor working alliance, was not far along in the stage of change process (e.g., more contemplative than active), s/he was more likely to make a unilateral decision to end therapy. The strengths and limitations of the study and implication of these results for future research and clinical application are discussed. The findings support the overall contention of this study that it is therapy-specific and contextual factors, and not client or therapist factors alone, that predict the likelihood of premature termination.
| Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/29042 |
| Date | January 2003 |
| Creators | Best, Marlene |
| Contributors | Hunsley, John, |
| Publisher | University of Ottawa (Canada) |
| Source Sets | Université d’Ottawa |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Format | 227 p. |
Page generated in 0.0014 seconds