Flückiger, C., Rubel, J., Del Re, A. C., Horvath, A. O., Wampold, B. E., Crits-Christoph, P., Atzil-Slonim, D., . . . Barber, J. P. (2020). The reciprocal relationship between alliance and early treatment symptoms: A two-stage individual participant data meta-analysis. Journal of Consulting and Clinical Psychology, 88(9), 829–843.
The therapeutic alliance (patient and therapist agreement on tasks and goals of therapy and their emotional bond) is the most researched concept in psychotherapy. The research clearly indicates that a positive alliance reliably predicts patient outcomes in terms of reduced symptoms. However, researchers still debate whether the alliance is at all necessary. That is, some argue that the alliance is the result of patients feeling better early in therapy, and so the alliance is only an outcome of early symptom reduction. If that is the case, then the alliance is an artifact of symptom reduction, and clinicians need not pay much attention to it. In this meta-analysis, Fluckiger and colleagues collected 17 studies representing over 5000 patients that evaluated whether alliance in a previous session predicted outcomes in a subsequent therapy session, and vice versa. In other words, they looked at all studies that evaluated if change in alliance preceded change in symptoms and if change in symptoms preceded change in the alliance. What is unique about this meta-analysis is that they gathered patient-level data from the original studies. That allowed them to test the therapeutic alliance theory for each individual patient on a session by session basis for the first 7 sessions of therapy. (For the stats geeks out there, the authors analysed within-person [between-session] effects using multilevel time-lagged models). Their analyses found that high alliance at a preceding session was related to lower symptoms at the subsequent session (B adjusted = -.065 (95% CI [-.092, -.038]; p < .0001)), and higher symptoms at the start of a session was related to lower post session alliance (B adjusted = -.148 (95% CI [-.215, -.081]; p < .0001). They also found that patients who generally reported high alliance scores showed a stronger alliance – outcome relationship, and those with greater symptoms had a weaker alliance - outcome relationship.
This meta-analysis is another indication of the importance of therapists and patients coming to a collaborative agreement on the tasks of therapy (what is done during sessions) and the goals of therapy (what issues to work on), and of their relational bond. The alliance is not always easy to establish – especially with regard to agreeing on goals. Also, the alliance should not be forgotten once established – alliance ruptures or tensions occur frequently and can have a negative effect on patients’ mental health outcomes. Patients of psychotherapists who repair alliance tensions generally have better mental health outcomes.