Prediction of suicide and attempted suicide

Abstract

A limiting factor in our ability to prevent suicide is that we are unable to predict suicidal behaviour in those individuals among the many who are suicidal. This has been known for over 40 years, and repeated attempts to refine prediction to the extent that it would be of clinical value have failed. This chapter focuses mainly on those studies designed to predict suicide, although it also addresses the prediction of repeated attempted suicide. In doing so it emphasizes the statistical limitations, in terms of experimental design, that are imposed upon those pursuing research examining the effectiveness of programmes designed to prevent suicide or repeated attempted suicide. It is suggested that the ideal research design of randomized controlled trials may not be feasible in confirming the efficacy of such programmes, and that alternative studies utilizing innovative methodologies may be more appropriate. Such studies may provide persuasive evidence for the clinician that, despite our inability to predict suicidal behaviour in any individual, there are treatments with demonstrated effectiveness.