SUICIDE IS NOT PREDICTABLE
SUICIDE IS NOT PREDICTABLE
Alan Eppel MB, FRCPC
Assessment of suicide risk is a core aspect of every psychiatric evaluation. The reality is however that the prediction of completed suicide is a mathematical impossibility. This is because completed suicide is a relatively rare event in comparison to thoughts of suicide and suicide attempts.
Mulder et al. (1) question the traditional view that risk factors can be validly identified and used to predict the likelihood of suicide. In studies of completed suicide, risk factors are not specific enough and are common to a large number of patients presenting with psychiatric problems. Use of risk factors leads to a high level of false positives. The majority of individuals who complete suicide are not identified as at high risk.
Conversely risk assessment tools that categorize individuals as low risk may lead to many false negatives.
Mulder concludes that “perhaps it is finally time to acknowledge that rare events such as suicide… …are impossible to predict with a degree of accuracy that is clinically meaningful”.
Risk assessment scales also lack adequate specificity and sensitivity to be of use in predicting suicide. False negatives may lead to false reassurance and false positives may result in unnecessary and intrusive interventions.
These authors repeat the theory of other researchers that risk assessment may be merely a procedure to reduce physician and hospital anxiety about potential suicide and does not lead to improvements in patient care.
Reliance on risk assessment tools can lead to the false conclusion that a poor outcome is the result of an inadequate risk assessment rather than inadequate patient care.
Mulder recommends that psychiatry should abandon these misguided attempts at risk prediction and instead encourage real engagement with patients and their problems and situations.
Pisani et al (3) are critical of the time-honoured practice of categorizing patients into high, moderate and low risk for suicide. They state that there is little evidence for the validity, reliability or utility of this approach. Instead these authors recommend a shift from suicide prediction to suicide prevention.
They make a useful distinction between 1. Risk Status and 2. Risk State.
Risk Status refers to the patient’s risk compared to a particular population. Risk State refers to the patient’s risk of suicide compared to his or her baseline.
They identify two other dimensions of their approach: available resources and foreseable changes.
They point out that risk formulation should be “anchored in the clinical context and patient population in which the assessment occurs”. Physicians working in different practice contexts e.g. inpatients, outpatients or emergency services deal with patients at different stages of their illness and at different levels of risk. A clinician’s judgment about risk will be dependent on that clinical experience. A patient may be considered at high risk in the family physician’s office but in a psychiatric emergency be viewed as relatively low risk. Different settings have different levels of resources to manage risk and this will also influence the assessed risk level.
Risk formulation should also take into account the patient’s baseline level of risk and identify factors that may have led to a change in this baseline level. It is also important to evaluate any known potential future events that might seriously impact on this level of risk e.g. upcoming divorce, anniversary of a parent’s death or loss of a job.
Pisani and his colleagues stress that the purpose of risk formulation is not about prediction but about intervention strategies and working with healthcare providers, the patient and their families.
Static or enduring risk factors such as age, gender, and marital status are not helpful when assessing any one individual. These demographic factors may be of relevance when looking at population groups. It is the dynamic of fluid factors that may be of more use in the clinical context. This involves understanding the patient’s “predicament” (4) and the multiple psychosocial factors that are impacting on his or her mental state.
Is There Any Hope ?
Psychiatric practice like other branches of medicine is shaped by environmental, political, cultural and economic factors.
This has led to a more impersonal and technical approach to psychiatric care. There is a greater reliance on checklists and assessment tools and less on human interaction between patient and clinician.
Reduced lengths of stay have added to the lack of engagement between medical and nursing staff and patients in hospital.
Changes in staffing patterns have decreased continuity of care and added to the patient’s feelings of uncertainty and vulnerability when hospitalized.
Similar forces are operating in outpatient services where there are organizational and economic pressures to limit interventions to short-term or consultative models. These forces run counter to the therapeutic engagement that is critical to reduce suicide attempts and completed suicide (2).
The critical component in medicine and in psychiatry is the therapeutic relationship between the physician and the patient. This relationship is the pivot around which all other therapeutic interventions revolve. Without it there can be no true healing.
The establishment of a humane and therapeutic engagement between clinical staff and patients must be a priority for all psychiatric services. This is the sine qua non without which our interventions will fail.
1.Mulder R, Newton-Howes G, Coid J. The futility of risk prediction in psychiatry.
Brit J Psychiatry 2016, 271-272.
- Pereira1S,Woollaston K. Therapeutic engagement in acute psychiatric
inpatient services. Journal of Psychiatric Intensive Care 2007,v.31:3-11
- Pisani A, Murrie D, Silverman M. Reformulating suicide risk formulation:
from prediction to prevention. Academic Psychiatry 2016, 623-629.l
- Sakinofsky I. Suicide and suicidality in Psychiatric clinical Skills. Ed.Goldbloom. Mosby 2006.