Psychiatry for General Practitioners Study Days 2017 December 13th 2017
Suicide Risk Assessment
Keith Hawton
Psychiatry for General Practitioners Study Days 2017 December 13th 2017
Suicide Risk Assessment: Moving from the Impossible
to Suicide Prevention
Keith Hawton
The problem of trying to predict risk of suicidal behaviour
Risk prediction does not work
Prediction of Suicide
Assessment of risk prior to suicide
Assessment of risk prior to suicide
Empirical Data Regarding Those Who Die by Suicide
The majority of patients who die by suicide actually deny having suicidal
thoughts when last asked prior to their death, OR communicate their risk in
more behavioural ways vs. verbal messaging
– Appleby L, et al. BMJ. 1999;318(7193):1235-1239.
– Barraclough B, et al. Br J Psychiatry. 1974;125(0):355-373
– Busch KA, et al. J Clin Psychiatry. 2003;64(1):14-19.
– Chavan BS, et al. Indian J Psychiatry. 2008;50(1):34-38.
– DeLong WB, et al. Am J Psychiatry. 1961;117(8):695-701.
– Hall RC, et al. Psychosomatics. 1999;40(1):18-27.
– Hjemeland H, et al. Soc Psychiatry Psychiatr Epidemiol. 1996;31(5):272-283.
– Isometsä ET, et al. Am J Psychiatry. 1995;152(6):919-922.
– McKelvey RS, et al. Aust N Z J Psychiatry. 1998;32(3):344-348.
Prediction of Repetition of Self-harm
Assessment of risk following self-harm
Risk (N) n (%) repeating self-harm
Low (1721) 165 (9.6)
Moderate (1738) 288 (16.6)
High (369) 95 (25.7)
(Kapur et al BMJ 2005)
Assessment of risk following self-harm
Risk (N) n (%) repeating self-harm
Low (1721) 165 (9.6)
Moderate(1738) 288 (16.6)
High (369) 95 (25.7)
(Kapur et al BMJ 2005)
11
126 consecutive patients admitted to the John Radcliffe Hospital following self-harm
A scale developed in 1983 by Patterson et al in Canada for teaching medical students about assessment of suicide risk
Based on the 10 major risk factors for suicide:
Sex (Male)
Age (<19 or >45)
Depression
Previous attempts
Ethanol abuse
Rational thinking loss
Social supports lacking
Organised plan
No spouse
Sickness
Scoring:
1 point for each factor
0 = very low risk 10 = very high risk
0-2 – send home with follow up
3-4 – close follow up; consider hospitalisation
5-6 – strongly consider hospitalisation
7-10 – hospitalise
SADPERSONS score < 7
SADPERSONS score ≥7
Referral to
secondary care
(N=69)
65 (94.2%) 4 (5.8%)
Psychiatric inpatient
care (N=5) 4 (80%) 1 (20%)
Repetition of self-
harm within 6
months (N=30)
28 (93.3%) 2 (6.7%)
SADPERSONS missed:
65/70 referrals to 2o care
4/5 admissions to psychiatric hospital
28/31 who repeated SH at 6/12
Risk tools and scales
Risk tools and scales
Risk tools and scales to predict
suicide after self-harm:
• Positive Predictive Value about
5%
• So they are wrong 95% of the
time
• And they miss suicide deaths in
the large ‘low risk’ group
Risk tools and scales
UK NICE Guidelines (2011)
Do not use risk assessment tools and
scales to predict future suicide or
repetition of self-harm.
Do not use risk assessment tools and
scales to determine who should be
offered treatment or who should be
discharged.
Risk assessment tools may be
considered to help structure, prompt,
or add detail to assessment.
Limitations of Solely Relying on Risk Factors
The problem with using risk factors to assess the probability of suicide is that they all produce high false negatives.
We have never been very good at predicting the future – especially for low-base rate behaviours.
What is the best alternative to risk prediction?
• Recognise that risk prediction is a fallacy
• Address patient needs
• Focus on the therapeutic aspects of the assessment
• Use clinical guidelines and make evidence based treatments available
• Individualised assessment which informs management
• Adopt population approaches to prevention – ‘something for everyone’
Population approach
Assessment, Not Prediction
• The goal is to assess the patient and use information to
plan for safety and treatment
• The clinician cannot
– Assign weights to risk and protective factors
– Subtract the protective factors from the risk factor, thereby
getting the net suicide risk
• The clinician can
– Use a process of progressive questioning to gather relevant
information to inform a risk formulation that leads to an
individualized intervention
Suicide Risk Formulation
A Typical Risk Formulation Approach (Prediction Model)
A Typical Risk Formulation Approach
Variable\Rating Nominal Low Moderate High
Ideation None reported Passive
wishes,
Fleeting
thoughts
Some
thoughts
Active and/or
pervasive
thoughts
Plan None reported
No plan Vague plan Specific plan
Lethality None reported
Low Potentially
lethal
Lethal
Access None reported
Limited Accessible Possesses
Intent None reported
Vague Some Desires to die
Pisani AR, et al. Reformulating Risk Formulation: From Prediction to Prevention. Acad Psych, 2015.
Managing risk
Focus on individualised risk reduction in all patients
e.g. - safety planning
- evidenced-based treatments
- therapeutic relationship
- involvement of family and others
- reduced access to means
- clear and communicated plans
This presentation discusses independent research funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research scheme (RP-PG-0606-1247, RP-PG-0610-10026) and Policy Research Programme. The views expressed in this presentation are those of the author and not necessarily those of the NHS, the NIHR or the Department of Health.
Acknowledgements
Professor Nav Kapur (Manchester)
Dr Morton Silverman (USA)
Psychiatry for General Practitioners Study Days 2017 December 13th 2017
Suicide Risk Assessment: Moving from the Impossible
to Suicide Prevention
Keith Hawton