There has long been an association between mental illness and dangerousness in the public mind, even in advance of any systematic studies of the matter. Research over recent decades, however, has shown that individuals with mental illnesses such as schizophrenia are just slightly more likely to engage in acts of violence compared to individuals without such illnesses. The vast majority of people with mental illness are not violent in the slightest.
People with mental illness are, however, far more likely to be the victims of crime (rather than perpetrators), compared to those without mental illness. At population level, the proportion of violent crime attributable to mental illnesses, such as schizophrenia, is extremely low and much is attributable to co-occurring drug misuse, which increases the risk of violence in individuals with and without mental illness.
Because violence is such a rare event in mental illness, it is extremely difficult to predict. The most detailed predictive models, which include almost all known risk factors for violence in schizophrenia, can only explain approximately one-quarter of the variation in violence between individuals with schizophrenia. It is mostly unexplained and unpredictable.
Even if there was a predictive model that was 90 per cent sensitive and 90 per cent specific (both of which are unrealistically high levels of prediction in any field of medicine), the rarity of homicide by individuals with severe mental illness means that such a predictive model would generate at least 2,000 false positives for every true positive, ie, the model would predict that 2,001 mentally ill individuals were at high risk of committing homicide but in fact, only one would have gone on to do so.
So, if prediction is impossible, is prevention impossible too? Possibly not. Given that many people who engage in, for example, murder-suicide have a history of depression, it is possible that better treatment of depression might help prevent murder-suicide. On the one hand, it is certainly not proven that better treatment of depression prevents murder-suicide, but, on the other hand, it is difficult to see how there could ever be such proof, because it is impossible to count events that have been prevented, especially very rare ones like murder-suicide. In any case, there are already many good reasons for better treatment of depression, including reducing suffering, improving quality-of-life, and so forth.
As a result, and especially given its slow rate of occurrence, violence in the context of mental illness remains statistically unpredictable at the individual level, notwithstanding ongoing efforts to understand it better and to build improved risk models that will hopefully help prevent such events in the future. In the meantime, good primary care, better mental healthcare, better treatment of depression, increased focus on the welfare of children, better communication, increased involvement of families in services, and enhanced responses to domestic violence will certainly deliver proven benefits, such as reduced suffering and better quality-of-life for all.
These issues are far from new, and media preoccupation with them is a constant, especially in the UK. In 1994, Dr Anthony Clare, Irish psychiatrist, author and presenter of In the Psychiatrist’s Chair (BBC Radio 4, 1982-2001), delivered a fascinating lecture on this theme, titled ‘Violence, Mental Illness and Society’ at the Royal Society of Medicine in London. Clare’s lecture was the 23rd in the series of the Stevens Lectures for the Laity. Previous lecturers included Prof Sir Richard Doll (the epidemiologist who linked smoking with health problems) and Sir Ludovic Kennedy (a journalist and campaigner whom Clare interviewed on In the Psychiatrist’s Chair).
In his 1994 lecture, Clare noted that “society yearns for simple solutions to complicated problems”, such as the apparent rise in societal violence: “… anxious voices have been raised concerning the possible relationship between psychiatric illness and violence, the anxiety nourished by a series of dramatic, well-publicised and highly disturbing events — of which the most recent and tragic was the murder of an innocent bystander, Jonathan Zeto, by Christopher Clunis, a patient suffering from schizophrenia.”
After a brisk run-through of the history of mental hospitals, Clare turned to the contemporary situation and noted that, owing to a lack of psychiatric services, “only when patients deteriorated to the point of violence were they being admitted. This is a truly dismal state of affairs and it is not new — I recall it well during my period working as a psychiatrist in Hackney in the early 1980s.” After considering the relevant research evidence, Clare concluded that the risk of violence associated with mental illness was “low” but, despite this, problems of perception remained: “… the combined effect of the deployment of persistently inadequate community resources and the growing proliferation of media accounts of violence and mental illness has been to increase public misperceptions and the stereotyping and stigmatisation of the psychiatrically ill. What is urgently needed now is corrective action [including] the appropriate and adequate provision of a quality and quantity of care for psychiatrically ill people that stands comparison with what we provide for the physically sick.”
It was also necessary “to ensure that the psychiatrically ill enjoy the same public treatment, respect, understanding and tolerance as have been gained by other minority groups”. Clare recommended greater engagement with the media in order to help ensure that this occurred.
Overall, Clare’s 1994 lecture was an impassioned plea for understanding, tolerance and action in order to dispel myths about mental illness and provide better care to those who needed it. Today, almost a quarter of a century later, Clare’s wise advice is still as important and urgent as ever.
Quotations from AW Clare’s ‘Violence, Mental Illness and Society: The Stevens Lectures for the Laity 1994’ are reproduced by kind permission of the Royal Society of Medicine.