Official mortality statistics show that each year in England and Wales approximately 4,000 people take their own lives.1 By universal assent this figure understates the true incidence of suicide although to what extent remains unclear. Hospitals provide care for nearly 2,000 cases of deliberate non-fatal self-poisoning every week. In addition an unknown percentage of the remaining 321,500 (1977 data) annual hospital admissions among those aged 15 years and over for other injuries and reactions involves some element of self-damaging intent as does an even more elusive proportion of the poisoning and wounds seen only by general practitioners or not brought to medical attention at all. Until very recently the number of completed suicides has tended to stabilise or even fall whilst the annual incidence of non-fatal deliberate self-harm has increased at an alarming rate.
The division of intentionally self-harmful behaviour into two discrete groups according to outcome as implied by the above statistics is, however, a potentially misleading representation of what many commentators consider to be one of the most serious of contemporary medical/social problems. Nevertheless, appreciation of the complexities involved is now substantially greater than 25 years ago. In the mid-1950s little distinction was drawn between suicide and 'attempted suicide': the latter was simply regarded as a bungled undertaking of the former. A more perceptive interpretation emerged in 1958 with the publication of Stengel and Cook's classic paper entitled Attempted Suicide. They identified important epidemiological differences between 'attempted suicide' and completed suicide and drew attention to the critical distinction between the two groups, namely that 'the person who has attempted suicide lives on as a rule and the attempt becomes a significant event in his life and calls forth actions from the human environment'.
Subsequently Kessel (1965), concerned with the emphasis on self-destruction in Stengel's approach, pointed out that in the majority of instances of so-called attempted suicide patients 'performed their acts in the belief that they were comparatively safe - aware, even in the heat of the moment, that they would survive their over-dosage and be able to disclose what they had done in good time to ensure their rescue'. Consequently he proposed that the terms 'deliberate self-poisoning' and 'deliberate self-injury' should be substituted as more appropriate descriptions of the behaviour in question. Both terms distinguish actions which are clearly not the result of an accident but at the same time avoid any suggestion that death is the desired outcome. With these advantages they became widely accepted by the end of the 1960s. In this paper nonfatal deliberate self-harm (DSH)2 will be employed to encompass both self-inflicted injury and poisoning, although, as it will become clear, the latter is the 'preferred' means in most cases coming to medical attention.
The last two decades have witnessed dramatic and divergent trends in the incidence of suicide and DSH. This paper examines these developments with a view to identifying the most plausible explanations for their occurrence. Superficially it would appear that there are legitimate grounds for regarding the two types of behaviour as quite distinct phenomena. But the paper then draws on more recently derived evidence to show that from epidemiological, psychiatric, motivational and other viewpoints there are many different subgroups which, together, constitute a wide spectrum of Suicidal/DSH behaviour. Against this background the final main section evaluates the success of various strategies designed to reduce the number of individuals who deliberately harm or kill themselves each year.