Working alliance and stability of therapeutic success in the treatment of depression: Feasibility study for a process-outcome study

Prof. Dorothea Huber

In empirical psychotherapy research, the working alliance is considered the best-studied construct; its statistical correlation with treatment outcomes has been proven to be robust in meta-analyses. For this reason, the working alliance is considered to play a central role in the complex structure of specific and non-specific factors.

In an exploratory design, the following questions will be investigated:

• Is the working alliance at the beginning and/or in the middle of treatment a predictor of therapeutic success at the end of treatment and of the stability of therapeutic success?
• Do the results differ depending on the perspective of the measurement: patient, therapist, and external examiner?
• Are there differential differences between different forms of therapy in the prediction of therapy success through the working alliance?

A completers sample of 72 patients (26 PA, 24 PT, 22 VT) from the Munich Psychotherapy Study (MPS) was examined using the Beck Depression Inventory and the Inventory of Interpersonal Problems (IIP). The Working Alliance Scale - Observer Rating (WAI-O-S) with a revised scale construction and guidelines was used as a predictor. It is based on a 12-item short form of the WAI with its three components: affective connection between patient and therapist, agreement on common therapy goals, and agreement on therapeutic tasks. In addition, the therapeutic alliance was assessed using the Helping Alliance Questionnaire, version for patients (HAQ-P) and therapists (HAQ-T), an 11-item measurement instrument that captures two facets of the working alliance: the perceived help provided by the therapist and the cooperation with or commitment to the therapist.

Multiple regression analyses are calculated to examine the research questions. One of the two outcome variables (BDI, IIP) at one of the three target time points (end of therapy, 1-year follow-up, 3-year follow-up) is used as the criterion. The following measures are considered as predictors: 1) type of therapy, 2) respective outcome variable at the start of therapy, and 3) working alliance from a specific measurement perspective (HAQ-P, HAQ-T, or WAI-O-S) and at a specific measurement point in time (6 and 12 months).

Although intensive rater training was conducted with the WAI-OS, it was not possible to achieve acceptable inter-rater reliability. This was attributed to the predominantly behavioral operationalization of the scale points. Therefore, no further application was submitted.

Original language: German

Project Start: 01/2014
Project End: 01/2021