Effectiveness of systemic family therapy versus treatment as usual for young people after self-harm: a pragmatic, phase 3, multicentre, randomised controlled trial

Abstract

Background: Self-harm in adolescents is common and repetition occurs in a high proportion of these cases. Scarce evidence exists for effectiveness of interventions to reduce self-harm.

Methods: This pragmatic, multicentre, randomised, controlled trial of family therapy versus treatment as usual was done at 40 UK Child and Adolescent Mental Health Services (CAMHS) centres. We recruited young people aged 11-17 years who had self-harmed at least twice and presented to CAMHS after self-harm. Participants were randomly assigned (1:1) to receive manualised family therapy delivered by trained and supervised family therapists or treatment as usual by local CAMHS. Participants and therapists were aware of treatment allocation; researchers were masked. The primary outcome was hospital attendance for repetition of self-harm in the 18 months after group assignment. Primary and safety analyses were done in the intention-to-treat population. The trial is registered at the ISRCTN registry, number ISRCTN59793150.

Findings: Between Nov 23, 2009, and Dec 31, 2013, 3554 young people were screened and 832 eligible young people consented to participation and were randomly assigned to receive family therapy (n=415) or treatment as usual (n=417). Primary outcome data were available for 795 (96%) participants. Numbers of hospital attendances for repeat self-harm events were not significantly different between the groups (118 [28%] in the family therapy group vs 103 [25%] in the treatment as usual group; hazard ratio 1·14 [95% CI 0·87-1·49] p=0·33). Similar numbers of adverse events occurred in both groups (787 in the family therapy group vs 847 in the treatment as usual group).

Interpretation: For adolescents referred to CAMHS after self-harm, having self-harmed at least once before, our family therapy intervention conferred no benefits over treatment as usual in reducing subsequent hospital attendance for self-harm. Clinicians are therefore still unable to recommend a clear, evidence-based intervention to reduce repeated self-harm in adolescents.

Funding: National Institute for Health Research Health Technology Assessment programme.

Figure 1

Trial profile A full list of reasons for dropouts at each stage of the study is provided in the appendix. SHIFT=Self-Harm Intervention: Family Therapy. *Data was not obtained for these people because clinical services did not fill in forms or contact was lost. †Reasons for loss to follow-up were unable to contact, contacted but unable to arrange visit, withdrawal from researcher visits, and visit arranged but cancelled or no one home.

Figure 2

Kaplan-Meier plot of time to self-harm Bars show 95% CI.

Figure 3

Moderator analysis: hazard ratio for risk of hospital attendance due to repeat self-harm (A) Baseline young-person ICU unemotional subscale score (range 0–15). Higher scores indicate more unemotional traits. (B) Baseline caregiver FAD affective involvement subscale score (range 1–4). Higher scores indicate poorer family functioning. FAD=Family Assessment Device. ICU=Inventory of Callous Unemotional Traits.