In-session interactive dynamics of the psychotherapy process between therapeutic alliance, therapist expertise, therapist technical intervention, patient metacognition and functioning

Empirical research has explored different dimensions of the therapy process and their associations, often showing bidirectional links: for example, metacognition may be favoured by a positive alliance with the clinician; on the other hand, metacognitive difficulties may be an obstacle for the alliance. However, little is still known about the overall relationship between multiple dimensions during the psychotherapy process. The aim of this study is to further explore the in-session interaction of therapeutic process variables, focusing on patient metacognition, therapeutic alliance, technical intervention, therapist expertise, and patient functioning. Participants included 45 patients involved in a psychodynamic weekly treatment in two clinical centres. Therapists were both in-training and experienced clinicians. Four instruments were applied on four psychotherapy sessions (178 verbatim transcripts): Metacognition Assessment Scale-Revised (MAS-R) assessing metacognition, Collaborative Interaction Scale (CIS) assessing therapeutic alliance, Psychodynamic Intervention Rating Scale (PIRS) assessing therapist technical interventions, and Shedler–Westen Assessment Procedure (SWAP)-200 assessing patient functioning. Sequential analyses revealed that specific therapist interventions co-occurred with three different levels of therapeutic alliance: a first one characterized by positive collaboration, the second one by neutral collaboration, and the third one by ruptures. Moreover, and critically, the patient metacognition, patient functioning, and therapist expertise were found to exert different effects in the three alliance levels. These findings suggest the existence of a specific interdependence between the variables involved in the research. These results further indicate that the therapist expertise is a key element in the therapeutic process, as it can drastically affect the in-session interactive dynamic.