Predictors of Psychotherapy Dropout in Patients with Borderline Personality Disorder: a systematic review

Abstract

Introduction: Borderline Personality Disorder (BPD) is a highly debilitating psychiatric condition. Despite the expansion of new BPD specific forms of psychotherapy in the last few decades, high dropout rates in these treatments have been reported. Treatment discontinuation is associated with poor patient outcomes, inefficient utilization of resources and the demoralization of healthcare providers. Methods: In order to identify predictors of psychotherapy dropout among patients with BPD, a systematic search of Medline, the Cochrane Library, PsycInfo and PsycArticles was conducted. Studies included were randomized-controlled trials in which patients diagnosed with BPD were exposed to a therapeutic intervention consisted of an evidence-based psychotherapy. The quality of evidence in the studies was assessed through the use of Revised Cochrane risk of bias tool for randomized trials. Results: Six articles, incorporating four types of treatment (Dialectical Behavior Therapy, Schema Therapy, Transference Focused Psychotherapy and Systems Training for Emotional Predictability and Problem Solving), were included. The dropout rates range from 31% to 69%. Overall, the studies present a low risk of attrition and reporting bias, and an unclear risk of selection, performance and detection bias. High hostility and a weaker therapeutic alliance were consistently predictive of higher risk of dropout. In contrast, better mindfulness skills and greater performance in specific neuropsychological domains, such as memory and executive control, were identified as predictive of lower risk of dropout. Sociodemographic variables and comorbidity did not influence treatment retention. Conclusion: Factors that influence discontinuation should be taken into consideration in future treatment programs, in an effort to optimize retention. Qualitative assessments of patients' reasons for dropping out may also help guide adjustments.