Intended for healthcare professionals

Education And Debate

Controversies in Management: Better treatment of mental illness is more appropriate aim

BMJ 1994; 309 doi: (Published 01 October 1994) Cite this as: BMJ 1994;309:860
  1. G Wilkinson
  1. London Hospital Medical College, London E1 2AD.

    Whensoever any affliction assails me, mee thinks I have the keyes of my prison in mine owne hand, and no remedy presents it selfe so soone to my heart, as mine own sword

    It seems unlikely that suicide can be prevented easily by medical means. Non-medical considerations are much more important, including questions of moral philosophy2,3 and sociopolitical matters.

    Although ancient moral arguments exist against suicide, it is difficult to resist the conclusion that suicide is not morally wrong. Indeed, suicide may be rational. So if someone is suicidal to what extent are doctors obliged to do something about it? The law is clear on one issue: that we should not assist people to commit suicide. Other concerns are less clear. If we provide the means for suicide, having fully informed the patient of the likely effects, we face a potential legal hazard. While we have a duty of care for people who are suicidal such people cannot expect unlimited professional advice and help. How long should doctors have the power to keep suicidal people alive against their will?

    Suicide cannot be predicted or prevented reliably

    The most important risk factor for suicide seems to be mental illness, but this knowledge has had little impact on our ability to predict suicide. Suicidal intent does not seem to be constant - “it waxes, wanes, and disappears, and it may surface abruptly.”4 Moreover, predictors significantly associated with suicide identify few suicides and many non- suicides. When an event is as rare as suicide even a predictive factor with high specificity and sensitivity includes too many false positives for practical purposes.

    For example, statistical modelling was used to predict suicide in a group of 1906 residents of Iowa with affective disorders who had been admitted to a tertiary care hospital.5 Risk factors identified included the number of previous suicide attempts, suicidal ideation on admission, bipolar affective disorder, sex, outcome at discharge, and unipolar depressive disorder in people with a family history of mania. The model identified none of the 46 patients who committed suicide.

    The apparent reduction in suicide after an educational programme for general practitioners in Gotland (a small island of 56 000 inhabitants off the Swedish mainland) has been much vaunted.5 But the results have been interpreted overoptimistically and cannot be generalised to other settings.*RF 6-9* During 1983-4, 16 of the 18 permanently employed general practitioners in Gotland took part in a programme comprising 20 hours of lectures, discussions, and videotape presentations covering aspects of depression. There were fewer than expected suicides during the period studied. However, suicide rates in Gotland were falling before the study began, and recorded numbers of suicides were small (the suicide rate was lowest in 1985 and rose again in 1986). In addition, suicide data were presented differently and inconsistently in the relevant publications, making independent assessment impossible. To evaluate the effect of teaching general practitioners about preventing suicide requires large sample sizes over a long time, and studies would be unlikely to show any statistical difference unless there was an enormous difference in the efficacy of different methods.10 As MacDonald suggests, it is time for “a moratorium on this idea that practitioners can prevent suicide.”10

    Political initiatives are incoherent and implausible

    Internationally, there are large, poorly explained variations in moral and societal attitudes to suicide, as well as different definitions, putative causes, methods, rates, and policies for treatment, research, and, inevitably, prevention.11 The result is strategic confusion and incoherence.

    The World Health Organisation's European strategy is to reverse rising trends in suicides (and attempted suicides) in the region by the year 2000. The WHO states: “This target could be achieved if improvements were made with regard to societal factors that put a strain on the individual, such as unemployment and social isolation; if the individual's ability to cope with life events were strengthened by education and social support; and if the health and social service personnel were better trained to deal with people at high risk.”12 But the reality is that there is no convincing evidence that education, improved social conditions and support, or better training play a substantive part in preventing suicide.

    In the United Kingdom politically motivated initiatives designed to urge others by the year 2000 to reduce the overall suicide rate by 15% and the suicide rate of severely mentally ill people by at least 33% seem, at the least, implausible.13 Optimistic cynics speculate that such targets were carefully chosen to match projections based on current trends. The latest data, however, suggest the targets may not be met. In England the three year average suicide rate for 1990-2 rose by 0.9%,14 and recorded suicides in England and Wales were 3950, 3893, and 3952 in 1990, 1991, and 1992, respectively (Office of Population Censuses and Surveys, personal communication). Moreover, at this late stage, even measuring the suicide rate in people with serious mental illness is a conundrum.

    Treating mental illness is better than preventing suicide

    There is too little established knowledge on medical aspects of suicide for a programme of primary preventive measures. The emphasis should be on secondary prevention - concentrating scarce resources on identifying and treating people with mental illness properly.

    Purchasers of health care should be wary of well meaning exhortations to develop and implement policies for preventing suicide. There is no good evidence of benefit from such commonly cited measures as seeking to influence the means available; targeting general hospital services for people who have attempted suicide; encouraging responsible reporting of suicide in the media; education in schools and of the public on mental health; and addressing problems such as family and marital breakdown, unemployment, and poverty.

    In the end Sir Richard Doll's pithy verdict on prospects for preventing mental illness remains particularly apposite: “it will, I suspect, be many years before we can design a programme for the prevention of mental illness, at either the individual or the social level, that will be more cost effective than a programme that provides for the treatment and support of affected individuals.”15

    I thank Professor Michael Shepherd and Dr Brian Barraclough for their comments. I also thank Ms S Floate, librarian at the Royal College of Psychiatrists, and Dr Jim Rodgers, Medical and Dental Defence Union of Scotland, for their advice.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 12.
    12. 13.
    13. 14.
    14. 15.