Relation of patient rated early session process to treatment outcome.

摘要

Understanding psychotherapy outcomes and how the process of psychotherapy influences these outcomes has proven to be difficult (Stiles, 1980). The Session Evaluation Questionnaire (SEQ: Stiles, 1980; Stiles & Snow, 1984; Stiles et al., 1994) was developed to facilitate the investigation of this relationship. While the SEQ has been found to be related to different techniques and to other process measures (Stiles et al., 2002), there is still little known regarding how it is related to psychotherapy outcomes. Additionally, most of the research examining the relationship between the SEQ and treatment outcomes has used composite scores to assess outcome, which combined the scores of several outcome measures. This makes the limited support for the SEQ in relation to outcome more ambiguous because the specific types of outcome (e.g., symptom change, interpersonal functioning, etc.) that are related to SEQ ratings is less clear. The current study aims to provide a clearer picture of how patient SEQ ratings are related to psychotherapy outcomes by using independent measures to assess various types of outcome (i.e., symptom change and interpersonal functioning). Furthermore, the current study examines the therapeutic alliance and therapist technique as clinically relevant mediators. In addition, therapist effects were accounted for using multi-level modeling techniques. The sample consisted of 73 outpatients at a university based clinic where they were seen in psychodynamic psychotherapy by 28 graduate student therapists in a clinical psychology Ph.D. program. A trend (p = .08) was found for patient SEQ Smoothness (smooth/rough, comfortable/uncomfortable, easy/difficult, pleasant/unpleasant, relaxed/tense) ratings to be related to patient rated symptom improvement (BSI-GSI) both before and after accounting for therapist effects. Additionally, it was found that patient SEQ Positivity (happy/sad, confident/afraid, pleased/angry, definite/uncertain, friendly/unfriendly) ratings are significantly related (p .05) to a broad band scale of patient rated improvement in global functioning (PEI) across many domains (i.e., psychopathology, interpersonal, occupational, and social) both before and after accounting for therapist effects. However, after accounting for therapist effects, this relationship was found to be mediated by patient rated alliance. Theoretical and clinical implications regarding technique, treatment, and training are discussed.