Suicide prevention: The role of the psychiatrist

It is refreshing to read Professor Snowdon’s article (2016) reporting the decrease in suicide. He canvasses reasons why there may have been this reduction and states that ‘It is hoped that this article provokes discussion concerning factors that may contribute to suicidal thinking’. Such an aim is laudable, as there is much that remains unknown. However, arguably, a more pressing aim is to ensure that hard won knowledge has been put into practice.

This appears to be particularly important, as in recent years, there has been a trend to focus on non-clinical issues and also to emphasise the long established fact that we are unable to reliably predict suicide in an individual. The latter is because for all it is drama and the clarity which is only possible with hindsight, suicide has a low base rate, a feature commonly encountered in medicine. This imposes considerable constraints on our ability to predict the course of a condition in an individual.

This has led to some practitioners, not only psychiatrists, developing a sense of clinical impotence … we cannot predict the individual who will die by suicide; therefore, why take these persons seriously? However, given the diversity of contributing factors, it is crucial to ensure that there is comprehensive assessment and management, and it is here that the psychiatrist’s role becomes paramount, as is evident from the literature.

Large national register studies have demonstrated a number of significant socio-demographic and clinical features. Such research has utilised the Population Attributable Risk (PAR) statistic, which is singularly appropriate for research assessing contributing factors to a condition, in this case suicide, as it has the potential to place risk factors in perspective at the population level.

The overwhelming importance of mental disorders has been demonstrated repeatedly. This is not to deny the significance of other issues, but such results should not only give policy-makers an indication of where resources should be best directed, but they also provide both confidence, as well as a sense of responsibility, to clinicians in treating mental disorders, the majority of which have been associated with an increased risk of dying by suicide.

Other factors such as the availability of means of suicide and the potential influence of the media have been investigated with innovative research designs. The importance of restricting access to, for example, coal gas, firearms, ligature points for hanging and bridge jumping sites has been demonstrated, as has the effectiveness of media restraint in being associated with fewer copy-cat suicides. These and a range of other suicide prevention measures have been reviewed elsewhere (Goldney, 2013).

But some other significant associations with suicide are more problematic to address. Nevertheless, the impact of unemployment, being indigenous, having sexual identity concerns, being in custody and experiencing childhood abuse, may be modifiable by lobbying leading to broad social change.

However, such lobbying should not be done at the expense of good clinical care. Indeed, by doing so, we as a profession run the risk of being seen to divert attention from our own shortcomings. Thus, there are reports from both Australia and the United Kingdom that up to 20% of suicides in those who have had psychiatric hospitalisation could have been preventable but for poor staff–patient relationships, inadequate assessment and management of depression and other conditions and poor continuity of care, particularly in the transition between clinical services (Burgess et al., 2000; National Confidential Inquiry, 2001).

There is now evidence that the implementation of service-wide policies can have an impact on suicide. In the United Kingdom, when policies recommended by the National Confidential Inquiry were introduced, those health authorities who implemented the majority of recommendations had fewer subsequent suicides than those who had not. These included the provision of 24-hour crisis care, introducing specific policies for those with a dual diagnosis and having multidisciplinary reviews after suicide (While et al., 2012).

The fact that such policies are not in place, or frequently not acted upon, is painfully apparent to those who conduct Coronial inquiries, as are other concerns. Often encountered are inadequate assessment, particularly in those with co-morbidity when patients are seen to be hand-balled between services; poor continuity of care; infrequent use of second opinions; an ethos of hospitalisation being avoided at all costs (and, stretching credulity, often accompanied by administrative assertions that bed availability was not an issue); a reluctance to use compulsory treatment orders; less than optimal use of anti-depressants; and inadequate or no consideration for the use of alternative psychotropic medications, particularly lithium and clozapine. None of these points involves anything other than standard evidence-based clinical care.

Snowdon is correct. There has been a gratifying reduction in suicide, but I am confident he would agree that there is more to be done, not only in pursuing further research but also in ensuring that current research results are translated into good psychiatric clinical practice.

See Debate by Snowdon 50(1): 13–15.