Alliance rupture and repair processes and therapeutic change in youth with borderline personality disorder

OBJECTIVES: This study aimed to investigate alliance rupture and repair processes in psychotherapy for youth with borderline personality disorder. It sought to examine whether alliance processes differ between treatments, across the phases of therapy, and what associations these processes might have with therapeutic outcomes. DESIGN: The study involves repeated measurement of both process and outcome measures. Hypotheses were addressed using within- and between-subjects analyses. METHODS: Forty-four people, aged 15-24, with a diagnosis of BPD were randomized to receive either 16 sessions of Cognitive Analytic Therapy (CAT) or a supportive treatment known as Befriending. In addition to pre-post outcome assessments, alliance processes were rated using the observer-based Rupture Resolution Rating Scale. RESULTS: Results indicated that CAT and Befriending did not differ in terms of number of ruptures, although CAT was associated with more stages of rupture resolution. Examination of alliance processes across time pointed to increasing ruptures, more frequent confrontation ruptures and increasing rupture resolution, suggesting increased volatility, directness and productivity in the therapeutic process across time. Contrary to hypotheses, there was no consistent link between alliance processes and outcome. However, two specific phases were significant. Early treatment ruptures were associated with poor outcome whereas greater late treatment resolution was associated with better outcomes. CONCLUSIONS: This study suggests that alliance processes can differ across treatments and the phases of therapy in psychotherapy for youth with BPD. Alliance ruptures are more likely to be problematic early in therapy but later in therapy, they appear to be opportunities for therapeutic growth. PRACTITIONER POINTS: Alliance ruptures are more likely than not to occur in any given session with a young person with Borderline Personality Disorder. Early in therapy, withdrawal type ruptures are more frequent, whereas late in therapy, confrontation ruptures are more frequent. Late in therapy, alliance ruptures should be viewed as opportunities for therapeutic change, rather than barriers to good outcomes.