Abstract
1 There have been few controlled prospective investigations into the prevention of suicidal behaviour and by and large they have failed to demonstrate the efficacy of social work, psychotherapy or psychiatric treatment. 2 A group of 58 high-risk patients with multiple episodes of suicidal behaviour was treated with mianserin 30 mg at night or placebo in a six month double-blind trial of the efficacy of an antidepressant in reducing suicidal behaviour. 3 Patients were screened for depression, schizophrenia and organic disease. Patients were diagnosed as suffering from personality disorders according to DSM-III criteria mainly borderline or histrionic. 4 There was no significant difference in outcome between the mianserin and placebo treated group at any point in the six month study. 5 An item analysis of the MADRS showed that at entry the item 'reduced appetite' predicted subsequent suicidal attempt. The total MADRS score did not predict further suicidal acts at entry but was highly significant at four weeks. At four weeks the items 'reduced sleep' and 'reduced appetite' were highly significant predictors of further suicidal acts and the items 'lassitude', 'suicidal thoughts', 'inability to feel' and 'pessimistic thoughts' were significant predictors.