Interpersonal Problems as a Predictor of Outcome in Psychotherapy for Depressive and Anxiety Disorders: A Multilevel Meta-Analysis

Objective: Although interpersonal problems are assumed to play an important role in the treatment of depression and anxiety, meta-analytic attempts to explore the association between interpersonal problems and outcome in psychotherapy for these patients are missing. This study aims to conduct a systematic review and meta-analytic synthesis of the association between patients' baseline interpersonal problems and outcome in psychotherapy for depressive and anxiety disorders. Method: We conducted a three-level meta-analysis (i.e., disaggregating sampling variance, within-study variance, and between-study variance) of the interpersonal problems-outcome prediction (IPOP; as measured by the total distress factor of Inventory of Interpersonal Problems). Results: We found 40 effect sizes (ESs) nested within 21 primary studies. The three-level model showed a significant aggregated effect size of IPOP, r = −.13, SE = 0.02, 95% CI [−.18, −.09]; t(39) = −5.71, p < .001; d = −0.27, with greater interpersonal distress being associated with worse psychotherapy outcome. Results showed significant heterogeneity across effect sizes with considerable variability across studies (I2 = .39) and to a lower extent across reported outcomes (I2 = .19). Graphical measures did not show evidence of a substantial publication bias. Conclusions: This meta-analysis showed that baseline overall interpersonal problems are a small but robust negative predictor of psychotherapy outcome. Therapists might benefit from exploring patients' interpersonal problems at the beginning of therapy to enhance their prognostic inferences. What is the public health significance of this article? This meta-analytic review showed that patients' interpersonal problems at the beginning of therapy are a small but robust predictor of psychotherapy outcome. Clinicians might benefit from systematically measuring patients' interpersonal problems at baseline to enhance their prognostic inferences.