Therapist effects in outpatient treatment of social anxiety disorder

Abstract

Background: The study examined therapist effects in the psychotherapy of social anxiety disorder using three methodological approaches: 1) variances of random effects, 2) effects of sociodemographic therapist variables and 3) disaggregation of therapeutic relationship and movement synchrony into between-therapist and within-therapist components. Material and methods: The patients (n = 267) were treated either with manually guided supportive-expressive therapy (n = 54), manually guided cognitive behavior therapy (n = 102) or naturalistic cognitive behavior therapy (n = 111). Patient self-reports of interpersonal problems (inventory of interpersonal problems, IIP-32), depression (Beck depression inventory, BDI) at the end of treatment and ratings of the therapeutic alliance (helping alliance questionnaire, HAQ) were collected. Movement synchrony was quantified by means of motion energy analysis and time series analytical procedures. The data analyses were carried out with multilevel models. Results: The age and sex of the therapist were not predictive for IIP-32 and BDI values at the end of therapy. Therapist effects modelled as random effects (intraclass correlation, ICC = 23.4% and 30.1%, respectively) explained much more outcome variance than the therapeutic approach (ICC = 6% and ICC = 1.6%, respectively). With respect to movement synchrony, the between-therapist and within-therapist parts were associated with a reduction of the IIP-32 scores. The between-therapist variance of the HAQ was predictive for a reduction of the BDI. Conclusion: In the psychotherapy of social anxiety disorder, therapist effects are more important than the therapeutic procedure. Similarly, the therapist part of the process-relevant variables therapeutic alliance and movement synchrony are more important than the patient or dyad-specific part.