Assessing the likelihood of suicide is a skill often attributed to, and indeed claimed, by psychiatrists. But several recent meta-analyses (statistical ways of combining data from multiple studies) find that the practice of suicide risk assessment has no scientific basis and may even be harmful to patients. One study concluded that “use of suicide risk assessment was distracting clinicians from therapeutic involvement with patients . . . We [psychiatrists] may well be putting our own professional anxieties above the needs of service users and paradoxically, increasing the risks of suicide following self-harm”.
In 2016, another study of the last 40 years of suicide risk assessment research by Matthew Large and his group in Australia found that 95 per cent of patients judge to be at "high risk" of suicide never kill themselves and 50 per cent of patient suicides were people thought to be at lower risk. As several other recent meta-analyses produced similar results it is now clear that there is no technique, skill or warning sign to identify patients at high risk of suicide in a way that would help treatment. In addition, designating some patients as low risk is "false reassurance" and, therefore, potentially harmful.
The adverse effects of so called suicide risk assessment are not confined to clinical situations, but also impact our legislation including the Protection of Life During Pregnancy Act, 2013, (PLDPA) and the Mental Health Act, 2001, (MHA).
X case
The shambles that is now the PLDPA is, of course, derived from the case Attorney General v X, 1992. In that case, the Supreme Court interpreted the eighth amendment to the Constitution to include in the clause "real and substantial risk to life" a risk to the life of the mother through suicide. In neither the High Court or the Supreme Court phases of the case was any expert psychiatric evidence considered regarding the validity of assessment of suicide risk.
The ultimate decision was a pragmatic, compassionate, anxiety-reducing solution to a difficult human problem. But “hard cases make bad law” and our legislators were saddled with the consequences, which involved dragging reluctant (and some enthusiastic) psychiatrists into the proceedings.
As the impossibility of suicide risk assessment has become clearer in the meta-analyses published in the last year, the current legal status is now even more untenable and dysfunctional. The most recent example is the case where the PLDPA was not invoked but instead the adolescent girl interviewed for risk of suicide was detained under the MHA. Both pieces of legislation require the impossible: assessing the likelihood of suicide in an individual in a clinically relevant time frame.
It is now time for us to stop allowing professional and societal anxieties about suicide (expressed through scientifically-invalid risk assessment) to take precedence over proper engagement with patients in crisis and rationalising our legal framework.
Abandoning traditional risk assessments for patients who feel suicidal would have two benefits. Firstly, for people in crisis it would facilitate better therapeutic engagement, thus helping to address and manage psychosocial problems and maladaptive coping strategies. Interventions such as safety plans for emergencies, brief cognitive-behaviour therapy for outpatient follow-up and longer-term dialectical behaviour therapy can reduce self-harm and improve quality of care. Suicide prevention in clinical services should be based on universal measures, for example, contact with all patients (irrespective of perceived risk) discharged from inpatient treatment within 24 hours.
Societal issue
Secondly, it would help us to recognise that preventing suicide is more a societal than a medical issue. The most effective measures are those of means reduction; for example, banning the use of pesticides (a very effective measure in Asia), reducing access to guns, curtailing access to iconic locations such as bridges or cliffs and reducing access to medications commonly used in overdoses. We need also to reduce our consumption of alcohol and to provide adequate resources for the treatment of mental illnesses. There is some evidence in favour of school education programmes.
While each suicide is a tragedy, we should not let our anxiety blind us to the evidence when formulating treatment plans for patients and law for our citizens. The current interpretation of the eighth amendment and those sections of the PLDPA dealing with suicide are based on a false premise. They may be doing more harm than good by promoting the idea that suicidality can be used as a currency to obtain rights or benefits and by spreading the false reassurance among the public that psychiatrists can predict and prevent individual suicides.
By the same logic, compulsory admission under the MHA on the grounds of suicide risk is also questionable. Life is complicated and there are no easy answers. Difficult issues must be tackled on the basis of evidence (while simultaneously managing the emotions of all involved) and not misguided compassion or “a wink and a nod”. Suicide risk is a complicated, controversial issue. No one benefits when provision of services and legislation is based on fiction.
Patrick Devitt and Declan Murray are consultant psychiatrists who recently authored the Scientific American article "Suicide Assessment Doesn't Work"