The Alliance-Outcome Relationship in Individual Psychotherapy for Early Psychosis: A Secondary Analysis of a Randomized Clinical Trial.

Background: Studies of the alliance-outcome relationship in psychosocial treatments for early psychosis are limited both in number and design, with few utilizing subjective measures as outcomes, multiple alliance timepoints, or existing benchmarks to distinguish high or low levels of the alliance. We addressed these gaps in context of a pilot randomized controlled psychotherapy trial for early psychosis. Methods: Twenty-eight clients with early psychosis completed alliance ratings at mid-treatment (at 4.5 months) and post-treatment (at 9 months). We tested the alliance-outcome relationship through change scores between mid- and post-treatment and use of a pre-defined benchmark to distinguish high versus low mid-treatment alliance. We also examined baseline differences between levels of the alliance. Outcomes included objective and subjective targets. Results: Change in the alliance did not predict any outcomes at post-treatment. Clients with high mid-treatment alliance had lower perceived stress and higher psychological well-being at post-treatment. At baseline, clients with high alliance had lower negative symptoms and loneliness as well as higher community functioning and well-being. Discussion: Use of an empirical alliance benchmark may help identify early psychosis clients who need additional support in alliance formation as well as for facilitating positive psychotherapy outcomes.