Ten controlled studies (n at least 1,664) were included: 3 RCTs (n at least 225), 4 clinical trials (n=645) and 3 quasi-experimental studies (n at least 794).
In terms of study quality, methodological flaws included selection bias, lack of blinding, low rates of participation, treatment groups not comparable at baseline, and ascertainment bias.
Group therapy (1 RCT): this was the only intervention that significantly reduced repetition of self-harm. The RCT found significantly fewer adolescents had two or more episodes of self-harm at 7 months with the intervention than with standard care. The intervention group was found to attend fewer psychotherapy appointments than the control group, but there was no significant difference between treatments in suicidal ideation.
Problem-solving therapy (1 controlled clinical trial): there were no statistically significant differences between treatments for any outcomes at the 3-month follow-up.
Intense management with outreach: 1 RCT, 2 quasi-experimental studies with historical controls, and one 3-arm quasi-experimental study with historical and contemporary controls were identified. The RCT found that the intervention reduced self-harm over 1 year, but found no difference between treatments for depression, hopelessness and suicidal ideation. Flaws included a lack of raw data, treatment groups dissimilar at baseline, and limited generalisability of the results. The other 3 studies found mixed results: two found that patients receiving the intervention had reduced rates of hospitalisation at 3 years (1 study) and reduced readmission rates for suicide attempt at 12 months (1 study), while the third found no significant difference between the interventions for repeat suicide attempts.
Readmission on demand (1 controlled trial): this study found no significant difference in readmission or suicidal behaviour at 12 months.
Enhanced family therapy: 1 RCT, 1 quasi-experimental study with medical and community control groups, and 1 quasi-experimental study with historical control were identified. The RCT found no significant differences between treatments for repetition of self-harm, adherence to psychotherapy, or in suicidal ideation at 2 and 6 months. Neither of the other studies reported suicidal behaviour or repeat self-harm. One of the studies found the intervention was not associated with greater attendance at follow-up appointments, but it was associated with reduced suicidal ideation when compared with the historical control.