RESUMO
Objective: The objective of this study was to test whether the therapeutic alliance mediated the relationship between previously identified predictors of premature termination and dropout during the first three sessions of treatment. Method: In this naturalistic study, 994 cases receiving individual, couple and family, or high-conflict coparenting therapy provided demographic information and completed assessments prior to treatment. Following the first session, clients completed a measure of the therapeutic alliance. Two hundred and five (20.6%) discontinued therapy prior to the fourth session. Logistic and ordinary least squares regression was used across m = 20 imputed datasets to examine the effect of pressure to attend therapy, age, gender, education, distress, therapy format, and therapist experience on whether clients continued in therapy and whether the alliance mediated this relationship. Results: After controlling for age, therapist experience, education, and pressure to attend therapy; general distress and participating in high-conflict coparenting were associated with higher rates of early termination. The effect of both distress and therapy format on dropout, however, was mediated by the therapeutic alliance. Conclusions: By focusing on improving the therapeutic alliance with high-conflict coparenting cases as well as clients with higher levels of distress, therapists may be able to increase client retention.