Self-harm is costly to individuals and health services and has high associated risks of further self-harm and completed suicide. Self-harm presentations to hospitals offer an opportunity to engage patients in interventions to help reduce future episodes and associated costs. This presentation reviews clinical trials for self-harm interventions conducted over the past twenty years in hospital emergency departments (1996–2016) comparing successful vs. unsuccessful trials (defined by the whole or partial achievement of trial defined outcomes) in terms of methodology, type of intervention and type of outcome measure.

Databases were searched using defined keywords. Randomized trials of adult subjects presenting to emergency departments were selected.

Twenty-four studies are included in the review. There was no significant difference between the type of intervention and “success”, nor were there index/control differences by sample size and follow-up length. Most trials (79%) used re-admission to hospital after a further episode as the primary outcome; only 4 (16%) of the studies reported social outcomes. As an example of social interventions and outcomes, we discuss trial results of a new social intervention for adults (many of whom do not receive a (UK-mandated) psychosocial assessment), and who are usually provided with little/no support after leaving the emergency room.

The findings suggest that the use of repetition and re-presentation as outcome indicators may be missing the importance of social precipitants of self-harm and the need to assess social circumstances, interventions and outcomes. We discuss findings from a new social intervention trial, which addresses these limitations.

The authors have not supplied their declaration of competing interest.