Purpose: This study sought to undertake an exploratory analysis of the impact of therapeutic alliance and dysfunctional attitudes (perfectionism and need for approval) on outcomes for participants receiving therapist-delivered and eHealth interventions for comorbid depression and alcohol/cannabis use problems. These factors have been shown in previous research to moderate response to psychological treatment for depression and related disorders.

Method: Participants (N = 274) with concurrent depression and alcohol/cannabis misuse were randomized to 10 sessions of therapist-delivered cognitive behavior therapy/motivational interviewing (CBT/MI), computer-delivered CBT/MI with brief therapist assistance (SHADE CBT/MI), or supportive counseling (PCT). Follow-up occurred at 3, 6, and 12 months post-baseline. Exploratory moderator analyses examined changes in depression, alcohol use, and cannabis use over the 3-12-month follow-up timepoints, adjusting for baseline, as a function of treatment allocation, and the hypothesized moderators of therapeutic alliance, perfectionism, and need for approval.

Results: The sample size and number of comparisons in the analysis mean that the results are considered preliminary and need replication in larger trials. The analysis revealed that "client initiative," a subscale of therapeutic alliance, moderated change in depression scores between 3- and 12-month follow-up for the PCT group, with higher scores associated with decreases in depression over time. Higher therapeutic "bond" early in treatment for SHADE CBT/MI participants was associated with reduced cannabis use over time. Participants with higher "perfectionism" scores at baseline who received therapist CBT/MI reported increases in depression over the follow-up period, but reductions in depression if they received SHADE CBT/MI. Therapist CBT/MI participants high on "need for approval" at baseline reported better alcohol use outcomes over time.

Conclusion: The preliminary nature of these results do not justify firm conclusions. However, the specific variables of perfectionism, need for approval, and client initiative show promise as moderators of treatment efficacy for comorbid depression and alcohol/cannabis use problems. Further research is justified to determine whether these factors can assist in tailoring the modality and strategies offered in the delivery of psychotherapy to this population.