DISCONTINUITY BETWEEN SUICIDE ATTEMPTS AND COMPLETED SUICIDE Implications for Prediction

1
DISCONTINUITY BETWEEN SUICIDE ATTEMPTS AND
COMPLETED SUICIDEImplications for Prediction

• Angus H Thompson
• Dept of Public Health, Flinders University

2
EVIDENCE FOR DISCONTINUITY

• Many Attempters do not wish to die

• Very few attempters take their own lives

• There are some differences between the two
  groups

3
SINGLE DIMENSION MODEL
Thoughts of Death
Suicidal Thoughts
Plans/Serious Ideation
Suicide Attempts
Completed Suicide
4
SINGLE DIMENSION MODEL

• Attempts signify higher risk

• Escalation in lethality if no resolution

• Observed differences between the two may be
  due to differences in escalated intensity

5
PREDICTION OF SUICIDE

• Too Rare to be Usefully Predicted?
• About 11 per 100,000 in Australia and Canada
  currently
• (lt 1 Ninetieth of one percent)

6
STUDY 1
Canadian University Students N 227 65
Female
7
STUDY 1
Q1 Has there ever been a period when you felt
like you wanted to die?
Q2 Have you ever felt so low you thought of
committing suicide?
Q3 Have you ever made definite plans to commit
suicide (even though you did not actually make
an attempt)?
Q4 Have you ever attempted suicide?
8
RESULTS
9
RESULTS
10
IMPROVING PREDICTION 1

• Annual Rate (then) 0.014

• Lifetime (40 Yr) rate 0.56

• Lifetime Ideator rate 1.18

11
STUDY 2 COMMUNICATION OF INTENT/ATTEMPTS

• Review Psychological Autopsy Studies
• Suicides with Prior Communication and Attempts

12
STUDY 2
Premise If an attempt is often a means of
communication, then attempts should be considered
along with other forms of expression of suicidal
intent for the purposes of prediction and
understanding.
13

• COMMUNICATION INCLUDES
• Direct Statements
• Dire Predictions, Allusions to Being Better
  off Dead, Expressing a Wish to be Dead

14
PSYCHOLOGICAL AUTOPSY STUDIES
15
Thus, on average 75 of suicides were preceded
by a communication of intent (Range 63-97)

• Conservative? All key informants?
• Conservative? Drop low scores?
• Liberal? Retrospective bias?

• Not a predictive value (how many
  communicators complete?

16
PROPORTION OF ATTEMPTERS WHO COMPLETE
17
ESTIMATING THE PROPORTION OF COMMUNICATORS WHO
COMPLETE
We Have Estimates of the Interrelationships
ofDeath Wishes, Ideation, Attempts Suicides
but Without Communication of Intent
Assume the Ratio of Communicators (Including
Attempters) to Attempters Only Among Completed
Suicides (1.69) is the Same AmongAll Showing
Parasuicidal Behaviour
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
MORTALITY DUE TO SUICIDE( of All Deaths)
24
ARE THERE RISK FACTORS?

• Most cases are captured in Death Wishes

• We need to know what prevents movement to
  the next level of intensity

• We need to identify precursors of the first
  death wish

25
HONING PREDICTION

• Distal prediction is relatively low because of
  many false positives. Distal nature and absence
  of negatives make it ideal for population and
  health promotion approaches

• Proximal prediction is much higher, but late in
  the game. Lower false positives. Suited for
  clinical and individually-based services

26
IMPROVING PREDICTIONMatching Activity
Predictability

• Primary Prevention
• Early Detection/Intervention
• Crisis Intervention / Rx

27
IMPROVING PREDICTIONMatching Activity
Predictability

• Primary Prevention Population Focus 1.7
• Early Detection/Intervention
• Crisis Intervention / Rx

28
IMPROVING PREDICTIONMatching Activity
Predictability

• Primary Prevention Population Focus 1.7
• Early Detection/InterventionDeath Wish/
  Ideation 1.7 - 2.5
• Crisis Intervention / Rx

29
IMPROVING PREDICTIONMatching Activity
Predictability

• Primary Prevention Population Focus 1.7
• Early Detection/InterventionDeath Wish/
  Ideation 1.7 - 2.5
• Crisis Intervention / RxPlans/Communication 6.5
  - 29

30
IMPROVING PREDICTIONProblem Synthesis

• Social problems are interrelated Common base?

• Preventing one will prevent others

• Common Underlying Factors

31
IMPROVING PREDICTIONAdding Risk/Protective
Factors

• Focus on Precursors
• Childhood Trauma Brain Development
• Secure Attachment Neglect / Abuse Punishment
  as Suppressant

• Focus on Trajectory Modifiers Coping
  Strategies Social Support Gene
  Action Treatment Random Acts Transitions

32
WHEN I GET TO BE DICTATOR
33

• For More Information
• Angus H Thompson
• Dept of Public Health, Flinders University
• angus.thompson_at_flinders.edu.au
• www.socialproblemindex.com

Many Thanks
34
Serotonin Transporter Gene Experience in Early
Life - Depression Age 26
Depression Risk
.70
SS
S Short Allele L Long Allele
.50
SL
LL
.30
No Abuse
Moderate Abuse
Severe Abuse
Founders Network
A. Caspi, Science, 18 July 2003, Vol 301.
35
RESULTS