Psychotropic medication use among adolescents participating in three randomized trials of DBT.

Background: Frequently presenting with symptoms of mood or anxiety disorders, substance abuse or borderline personality disorder, suicidal and self-harming adolescents often are prescribed psychotropic medication. Though such treatment may be warranted, recurrent suicidal and self-harming behaviour is often linked to emotion dysregulation where pharmacological treatment has weak empirical support. There is a need for more clinical research into the frequency, type and rationale for pharmacological treatment in this group. In this secondary analysis of three randomized clinical trials of dialectical behaviour therapy for adolescents, we report on psychotropic medication use in the respective samples at the time of recruitment, compare use of psychotropic medication across trials and describe sample characteristics that may be associated with possible differences in psychotropic medication.

Findings: Trials were conducted in Norway, the US and Spain (labelled the Oslo, US and Barcelona samples). At baseline, 86% of the Barcelona sample, 67% of the US sample and 12% of the Oslo sample were taking at least one psychotropic medication with antidepressants as the most frequent, followed by antipsychotics (72%, 22% and 1.3% respectively) and mood stabilizers (14.2%, 16.2% and 0%). In the Oslo sample there was a significant association between receiving a diagnosis of major depression and the likelihood of receiving antidepressants, but no such association was found in the Barcelona and US samples. The overall 7-8 times higher proportion of participants in the US and Barcelona samples treated with psychotropic medication could only partially be explained by differences between the samples in diagnostic profiles, symptom severity or level of dysfunction.

Conclusions: Highly prevalent in use among suicidal and self-harming adolescents with borderline features, psychotropic medication was still very unevenly prescribed across trials, differences not explained by differences in sample characteristics suggesting that current treatment practices are not fully empirically supported. We call for continued medical education and increased availability of evidence-based psychosocial interventions.