Background: Borderline personality disorder (BPD) is a severe condition traditionally considered difficult to treat and is characterized by both self and relational pathology. Attachment and mentalization theory considers the poorly developed capacity for stable, close relationships and (epistemic) trust as essential aspects of BPD. In terms of the therapeutic relationship, the development and maintenance of the alliance constitutes a central issue in therapy and a curing mechanism for BPD patients. However, it is well known that the therapeutic alliance can be extremely complicated in the treatment of BPD. There are also indications that patients with poor alliance are more likely to drop out early in therapy. BPD patients typically have a dropout rate of around 29%, but the real number may be higher due to bias towards the publication of studies with high completion rates. Importantly, different evidence-based treatments for BPD seem to work through, and create different, alliances. Therefore, the therapeutic alliance seems a crucial factor both for keeping patients in therapy and for effective treatment. The impact of alliance for patients with personality disorders (PDs) has been shown to be six times higher than for other patient groups. Despite the challenges involved in psychotherapy with individuals with BPD, there exists little literature on the therapeutic alliance in this population. Therefore, as the alliance, the quintessential common factor (CF) between treatments, has been poorly investigated in BPD treatment, this thesis examined the role of the working alliance and the CFs in an evidence-based treatment for BPD. An overall question in the thesis is how psychotherapy research can build a bridge between the so-called CFs and specific factors (i.e., the current schism). As reliable fidelity measures are necessary to judge a specific treatment technique as being superior to another, this thesis also aimed to establish reliable integrity measures for mentalization-based treatment (MBT).
Objectives: Paper I examined the reliability of a measure of treatment fidelity for the group component of MBT. Paper II investigated whether differences in rated MBT quality can be investigated through the lens of the CFs; for example, can therapeutic alliance account for some of the differences between high and poor ratings of therapists’ MBT fidelity? Paper III studied how aspects of therapeutic alliance (goals, tasks, and bonds) have developed over time in MBT for patients with BPD with different outcomes. Methods: Paper I applied generalizability theory (G-theory) in a reliability study (G-study) where five raters rated eight MBT group (MBT-G) and eight psychodynamic group (PDG) sessions according to the newly developed adherence and quality scale for MBT-G (MBT-GAQS). Paper II applied purposeful sampling to a pool of 108 rated sessions in the Quality Lab for Psychotherapy at Oslo University Hospital. The four selected sessions were subject to interpretative phenomenological analysis. Paper III applied linear mixed models to investigate the longitudinal development of alliance for 155 BPD patients in MBT. Psychosocial functioning measured using the Global Assessment of Functioning (GAF) scale indicated clinical outcomes. Subscales (goals, tasks, and bonds) of the working alliance as measured by the Working Alliance Inventory-Short Revised (WAI‐SR) were the dependent variables in an analysis of alliance development in MBT therapies with good (end-GAF < 60) and poor outcomes (end-GAF ≥ 60).
Results: The results in Paper I showed high reliability for both adherence and quality (competence). The mean absolute G-coefficient for adherence was .86 (range .63–.97) and for quality was .88 (range .64–.96) with five raters. The reliability for overall adherence (.97) and quality (.96) ratings (five raters) were both excellent. The nine group-specific items displayed high reliability for both adherence (range .83–.95) and quality (range .78–.96). With one rater, the reliability was also high for overall MBT-G adherence (.86) and quality (.83). However, the results indicated low reliability for items connected to psychic equivalence and pretend mode, especially with few raters. Paper II identified four themes that seemed to characterize therapy processes with different ratings of MBT quality (competence): 1) alliance, 2) strategic competence, 3) quality, and 4) “battles of the comfort zone”. Therapeutic alliance seemed to be fostered by battles of the comfort zone, quality, and strategy. Given an existing, adequate alliance between patient and therapist, the alliance seemed to become further nurtured when the interventions targeted the patients’ maladaptive patterns. Highly rated therapists intervened according to an overarching strategy and challenged the patients’ comfort zone. They also manifested a steadfast focus on the agreed therapeutic project (tasks and goals). The bond part of the alliance appeared as an asset in this “battling” process. Poorly rated therapists abandoned the therapeutic project and seemed overwhelmed by countertransference reactions. The therapeutic strategy seemed random and the sessions had low levels of challenging maladaptive patterns. Paper III showed that MBT treatments with good outcomes were characterized by positive development in the working alliance. Differences between subgroups with good and poorer outcomes were most prominent for the tasks subscale. Initial ratings of goals, bonds, and tasks did not differ by subgroup; levels were within a satisfactory range, but change over time was significantly different by subgroup. Comorbid paranoid PD was more frequent in the subgroup with poor outcomes and associated with poorer alliance development in this subgroup. Mood disorder was associated with significantly lower initial alliance levels but not with change in the working alliance subscales.
Conclusions: The overall MBT-G adherence and quality can be rated by one rater. Two of the core components of MBT theory, psychic equivalence and pretend mode, had low reliability. Paper II found that MBT-I may foster a strong therapeutic alliance and that CFs, such as alliance, believing in one’s own method, staying steadfast to the therapeutic project, and challenging the patient’s problems according to an overarching strategy characterizes highly rated MBT. Paper III demonstrated satisfactory levels of initial working alliance among BPD patients in MBT irrespective of clinical outcomes and that a positive temporal development of alliance characterized treatments with good outcome. Focusing on tasks in therapy seems especially important among these patients. In terms of the overarching title, “Measuring MBT – a marriage of the common and specific psychotherapy factors”, the thesis as a whole discusses how treatment processes in specialized treatment tailored for poorly functioning patients with BPD highlight how the CF and the specific factor approaches interact and suggests that “embedded alliance measures” (the alliance fostered by the specific technique) should be developed and implemented