Prediction of Treatment Discontinuation in Adolescents with Borderline Personality Disorder: Pilot Data from the Model Project of the Adolescent Center at the Central Institute for Mental Health Mannheim

Theoretical background: Borderline personality disorder (BPD) can be reliably diagnosed and effectively treated already in adolescence. Treating BPD as early as possible with disorder-specific therapies (e. g., dialectical behavioral therapy, DBT) seems to make sense because it can prevent self-harming behaviors and chronic courses with frequent suicidal crises as well as improve long-term psychosocial development. Several treatment efficacy studies in adolescence showed very high dropout rates (up to 75 %), but there are few consistent results on the prediction of treatment dropout. Objective: The main objective of the present study was to investigate, within the framework of an observational study, a) the dropout rate and b) the possible prediction (clinical and demographic variables) of dropout from inpatient treatment with DBT for adolescents in N = 172 patients aged 15 - 18 years. Method: We used logistic regression to examine 18 predictors (including age, number of diagnoses, symptom severity) in SPSS, which were collected during 3 months of inpatient treatment in routine care. Results: Of the 18 predictor variables examined in logistic regression, only younger patient age was statistically significantly associated with an increased risk of treatment discontinuation. We found no statistically significant associations with treatment discontinuation for the examined psychiatric comorbidities (e. g., for the presence of an eating disorder), for the severity of BPD-specific psychopathology, and for BPD-typical problematic behaviors (e. g., self-harming behavior or high-risk behavior). Discussion and conclusion: Our study finds no evidence of an increased risk of dropout from treatment with DBT for severe symptom distress. To replicate the increased dropout rate in younger adolescents in further studies, researchers would need to conduct systematic investigations on variables of particular importance in younger adolescents which could be related to treatment dropout (e. g., the relationship to important attachment figures). This information in turn could be used to derive targeted therapy modifications to counteract treatment discontinuation.