Background The Norwegian Directorate for Health and Social Affairs made an inquiry to the Norwegian Knowledge Centre for the Health Services (NOKC) to do a health technology assessment (HTA) of the effects of interventions by the psychiatric health services for the prevention of suicide. The HTA will be part of the basis for national guidelines for suicide prevention.

Psychotherapy, drug intervention and electroconvulsive therapy have been discussed in part 1 of the report (report 24-2006). The present part 2 focuses on other interventions.

Methods The work was carried out by a review team of external professionals. A literature search was performed in January 2006 in order to find systematic reviews. The Cochrane databases and the Health Technology Assessment database, in Medline, Embase, Psychinfo and Cinahl were searched using terms for suicide and self harm combined with a search filter developed by SIGN (appendix 1).

Of the 852 hits 29 reviews were deemed relevant by at least two group members after reading titles and abstracts (appendix 2).

These 29 articles made it clear what interventions have been described in the scientific literature. The systematic reviews were the starting point for the current review but results from the primary articles have been included wherever relevant.

Results and discussion Literature was found about interventions to improve continuity of care and follow-up (chain of care, letter and telephone contact, home visits by psychiatric community nurse), interventions to make care more accessible (green card, community mental health teams, hospitalization) and protective interventions (no-suicide contracts, securing the physical environment, special observation).

In a number of studies “usual care” given the control group was poorly described in the article.

Several studies described care given patients treated in somatic hospitals after suicide attempts, inclusion criteria were often poorly described, and the duration of the interventions differed considerably.

Results from studies of interventions given psychiatric patients in general may not apply to seriously suicidal persons.

These and other factors complicate the interpretation of both positive and negative results and reduces the external validity of the studies.

Conclusions

• A number of interventions aim to make psychiatric health care more accessible to persons with mental health problems
• Methodological issues make it hard to draw conclusions from studies of chain-of-care interventions
• Follow-up with letters to patients with little motivation for treatment had preventive effect in one study
• Short-term telephone contact did not have any effect
• Long-term telephone contact with elderly women may have primary preventive effect
• Follow-up by psychiatric nurse after hospitalization has not been shown to have effect
• In one case-control study reduction of care was a risk factor for suicide
• None of the three studies of green cards showed significant effect of the intervention. Two small studies that also offered possibility of personal contact had trends toward effect, a third study of telephone contact alone found no effect
• Metaanalyses of studies of community mental health teams found a near statistically significant reduction of total mortality and death from suicide
• No studies of the effect of hospitalization allow conclusions as to what patients should be admitted
• No studies evaluating the effect of protective interventions were found

There is a considerable need for research into the effects of the different kinds of interventions offered suicidal patients.