Systematic review
Effects of secondary preventive interventions against self-harm
Systematic review
|Updated
We have found research on many interventions and studied whether people harm themselves less frequently and have fewer suicide attempts when they have received one of these interventions compared to people in a control group who have not received the intervention. We have also looked at whether the interventions can reduce psychiatric symptoms.
Download
Key message
There may be as many as 1 in 10 young people in Norway who have self-harmed, and many of them attempt suicide, but the exact occurrence is uncertain. When people contact health services after self-harm or suicide attempts, it is important to have effective interventions for preventing that it happens again. We have found research on many interventions and studied whether people harm themselves less frequently and have fewer suicide attempts when they have received one of these interventions compared to people in a control group who have not received the intervention. We have also looked at whether the interventions can reduce psychiatric symptoms.
We found:
- active contact and follow-up in emergency wards probably reduces new suicide attempts
- problem solving therapy possibly reduces repeat self-harm and psychiatric symptoms
- psychodynamic interpersonal therapy possibly reduces psychiatric symptoms
- intensive follow-up and outreach possibly reduces repeat self-harm, suicide attempts and suicides
- the effect of other secondary prevention interventions like e.g. cognitive therapy, cognitive behaviour therapy, telephone contact, and the school-based programs C-CARE (Counselors Care: Assess, Respond, Empower) and CAST (Coping and Support Training) are uncertain because the evidence is of very low quality
The results must be interpreted with caution as there are wide confidence intervals around the estimates.
Summary
Background
Intentional self-harm is a conscious and wanted injury that a person inflicts on his- or herself with or without intention to die, and the concept includes both suicide attempt (SA) and non-suicidal self-injury (NSSI). When attempting suicide, the person has a death wish, even though the wish can be more or less strong. A person with NSSI, on the other hand, does NOT want to die. NSSI is more connected to inflicting physical pain on oneself in order to change an intense negative thought, feeling or a difficult relationship. The prevalence of self-harm is uncertain, but according to the Norwegian Institute of Public Health, as many as one in ten youth may have harmed themselves.
It is important to prevent new episodes of self-harm in persons who contact the health services. In this report we provide an overview of available systematic reviews about the effect of psychosocial interventions to prevent repeated self-harm and suicide attempts and to reduce psychiatric symptoms. This overview is meant to be used for the assessment of the need for a guideline about diagnostics, follow-up and treatment of persons with suicide attempts and self-harm who contact the health services. The overview is also meant to guide the planning of interventions in the form of illustrative material on a national level. It is also meant to provide the best possible professional advice to politicians and service providers about how to work with the current audiences and problem areas.
Objective
This systematic overview summarizes the evidence for effects of psychosocial interventions to prevent repeated self-harm and suicide attempts and reduce psychiatric symptoms in people with a history of these behaviors.
Method
A research librarian performed a systematic literature search in September 2015. We searched for systematic reviews in the databases Epistemonikos, Medline, PubMed, Embase, PsycINFO, Cochrane Library (DARE, HTA), Campbell Library, CRD (DARE, HTA), Sociological Abstracts and, and Cinahl. We also searched the websites of the Swedish Agency for Health Technology Assessment (SBU) and the Danish National Centre for Social Research. We used a filter developed for finding references to systematic reviews. We did not search for primary studies. We searched for systematic reviews with literature search performed in 2010 or later. Two researchers independently scanned the records. If at least one researcher decided that the record might be relevant, it was acquired in full text. The same two researchers read the articles independently of each other and assessed them against the inclusion criteria. We assessed the methodological quality of the systematic reviews using a check list developed by the Norwegian Knowledge Centre for the Health Services. If there was disagreement about the quality assessment, the project manager decided. If we found more than one review with the same objective and same quality, we only included the one with the newest literature search. We have not assessed risk of bias in primary studies but reproduce the assessments of the authors of the included systematic reviews. One researcher (GS) extracted data from the included reviews, and another researcher (TKD) checked the numbers. Wherever the systematic reviews had performed meta-analyses, these are reproduced in our review. If meta-analyses had not been performed, we have only reproduced the results as they were presented in the systematic reviews. We have used Grading of Recommendations Assessment, Development and Evaluation (GRADE) for assessing the quality of the evidence.
Results
We have included four systematic reviews from 2015 about several types of secondary preventive interventions. We present the findings according to how strong confidence we have in the effect estimates. We use the word «probably» for moderate quality evidence, «possibly» for low quality evidence, and «uncertain» for very low quality evidence.
Moderate quality evidence:
- Active contact and follow-up probably leads to fewer repeated suicide attempts at 12 months follow-up (RR: 0.83, 95% CI: 0.71 to 0.97).
Low quality:
- Problem solving therapy (OR: 0.71, 95% CI: 0.45 to 1.11) possibly reduces repeated self-harm, but the confidence interval includes zero effect . There is possibly a positive effect of problem solving therapy on depression (Standardized mean difference [SMD]: -0.36, 95% CI: -0.61 to -0.11) and hopelessness (weighted mean difference [WMD]: -2.97, 95% CI: -4.81 to -1.13).
- Psychodynamic interpersonal therapy possibly reduces psychiatric symptoms (mean difference [MD]: -5.0, 95% CI: -9.7 to -0.3).
- Intensive follow-up and outreach shows somewhat better results than usual follow-up (OR: 0.84, 95% CI: 0.62 to 1.15) on repeated self-harm, but the confidence interval includes zero effect. The intervention has possibly no or uncertain effect on new suicide attempts after 18 months (OR: 1.02, 95% CI: 0.73 to 1.43), but it can possibly reduce the incidence of suicide after 18 months (OR: 0.11, 95% CI: 0.02 to 0.45).
Very low quality:
- The effects on repeated self-harm is uncertain for cognitive therapy, cognitive behaviour therapy, group-based psychotherapy, psychodynamic interpersonal therapy, dialectical behaviour therapy, mentalization, continuous follow-up with same therapist (versus change of therapist), emergency cards, admission to hospital (versus immediate discharge), therapeutic assessment, interventions to enhance compliance, home-based family intervention, and remote contact interventions.
- The effect on incidence of new suicide attempts are uncertain for psychotherapy and for the school programs C-CARE and CAST.
- The effect on underlying psychiatric symptoms is uncertain for cognitive therapy and psychodynamic interpersonal therapy.
- The effect on death from all causes is uncertain for active contact and follow-up.
- The effect on new hospital admissions is uncertain for nurse-led case management.
Discussion
We have not found evidence of high quality for any secondary preventive intervention. Even though the included systematic reviews are of high quality, the primary studies have significant methodological weaknesses. They are also mostly small in the sense that they have few events. There is a need for randomized controlled trials that are large enough to accumulate a sufficient number of instances of self-harm or suicide attempts, and with improved methodological quality. We have not assessed the quality of the primary studies, but it is important when systematic reviews are produced that the interventions provided are described in sufficient detail. It is not enough to only provide labels, e. g. «psychodynamic interpersonal therapy». There is a need for a detailed description of what the intervention contained, who were the therapists, how many treatment sessions were given, how often the intervention was provided, and for how long.
Conclusion
Many different interventions have been targeted for research, and some have shown a positive effect. Active contact and follow-up after treatment in emergency rooms probably reduces the number of new suicide attempts. But there is uncertainty associated with these effects, and we cannot tell whether any interventions are more effective than others.