Psychotherapies for Personality Disorders: A Systematic Review of RCTs with Focus on Psychosocial Functioning

Background. Psychotherapy is central to the care of personality disorders, yet trials often emphasize symptom change rather than psychosocial functioning, which is the conceptual core of contemporary AMPD and ICD-11 models. The field lacks an up-to-date review of how functioning is measured in RCTs for adult personality disorders. Objective. To systematically identify randomized controlled trials of psychotherapies for adult personality disorders that used inactive comparators and reported functioning at the end of treatment, and to classify functioning outcomes into global, interpersonal, and self domains. Methods. Reporting followed PRISMA 2020. Searches of PubMed, Embase, and APA PsycInfo to November 2024 informed a larger database from which eligible trials were selected. Inclusion criteria were adults with DSM or ICD personality disorder, a manualized or structured psychotherapy, an inactive comparator, and at least one functioning measure assessed at treatment end. Study characteristics were extracted with a standardized template. Risk of bias was appraised with RoB 2. Synthesis was qualitative. Results. The search yielded 4,914 records; 13 trials met the inclusion criteria. Studies were conducted in Canada, the Netherlands, Italy, the United Kingdom, and the United States, with a total randomized sample of 2,008 participants. Diagnoses included borderline personality disorder in six trials, any personality disorder in four, avoidant in one, antisocial in one, and mixed non-BPD in one. Interventions comprised dialectical behavior therapy, schema therapy, cognitive-behavioral and interpersonal approaches, short-term dynamic psychotherapy, psychoeducation/structured support, and a webbased program. Comparators were treatment as usual (n=9), wait-list (n=3), or monitoring-only (n=1). Functioning outcomes were reported as global in seven studies, interpersonal in eight, and self in one; no trial used AMPD Level of Personality Functioning measures. Overall risk of bias was high in 9/13 trials (69%) and raised some concerns in 4/13 (31%); none were low risk. Conclusions. Across randomized studies, functioning was included but operationalized heterogeneously; assessments of self-functioning were sparse, and no study employed AMPD-consistent measures. The overall certainty is weakened by methodological issues. Next trials should prospectively specify functioning outcomes, use validated AMPD or ICD-11 instruments, register protocols in advance, blind outcome assessment where feasible, and report functioning together with symptoms at post-treatment and follow-up.