Therapy process factors in early psychosis: The effect of working alliance on clinical symptoms and cognitive biases in Ultra High Risk young people.

Approximately 25 % of individuals at Ultra High Risk (UHR) transition to a full psychotic disorder. Cognitive biases are thought to play a role in this risk, and improvements in cognitive biases have been associated with better clinical outcomes. Early intervention is crucial, yet no therapeutic approach has proven superior, and treatment adherence remains a challenge. Therapeutic alliance, shown to enhance adherence and clinical outcomes, might also influence depression and cognitive biases, but its role in these domains among UHR patients remains unclear. A total of 202 UHR participants (59.4 % females, mean age 17.4), taking part in the sequential multiple-assignment randomized trial (SMART), completed questionnaires assessing cognitive biases, attenuated psychotic symptoms, quality of life, general and social functioning, depression, and overall psychopathology at baseline, and at 6- and 12-month follow-ups. Additionally, both patients and their corresponding therapists rated working alliance following the 6- and 12-month treatment periods. The results revealed that patient-rated therapeutic alliance significantly predicted improvements in attenuated psychotic symptoms, cognitive biases, general psychopathology, social functioning, and quality of life, with the strongest effects observed when measured 6 months into treatment. The relationship between working alliance and decreases in attenuated psychotic symptoms, as well as between working alliance and reduction in general psychopathology, turned out to be mediated by improvements in cognitive biases. Therapist-rated working alliance did not significantly predict clinical improvements, showing only a minor association with quality of life. Strengthening the therapeutic alliance and prioritizing cognitive bias modification early in interventions for UHR patients may improve treatment outcomes. Further research is needed to validate these findings.