Children & disaster trauma, evidence & best practice.
Brett McDermott Mater Child and Youth Mental Health Service
Acknowledgements (& credibility!)
Sutherland Bushfire Trauma Project (n=4000)
– Victor Storm, Don Finlayson, Sue Pammet, school counselors
Canberra Bushfires (n=400)
– Erica Lee, Marianne Judd, Peter Gibbon
Cyclone Larry Children Project (n=2700)
– Vanessa Cobham, Helen Berry, Kevin Freele, EdQ and Cathed counselors and teachers
Brisbane Storms 08-09 (n=2000)
– Judi Krause, NW CYMHS.
Explanatory models
Take to the event
Personality Temperament Genes Past history
outcomes
PTSD Anxiety depression
Event related factors
Exposure Threat perception
Distal
Proximal
Outcome1
2
Large scale public health approach is possible
Cyclone Larry 2978 school children screened
EdQ pre-3
4-7
Cath 201
236
Total 703
819
502
583
8-12
913
543
1456
27 schools: - 18 Primary schools and 2 high schools from Ed QLD - 6 primary schools and 1 high school from Catholic Ed
Screening Instruments
Child report PTSD-RI exposure questions community connectedness questions Parent report SDQ-Em exposure questions community connectedness questions Family Adjustment Device or Family Resiliency Scale Screening 3 months post-disaster
}
Delivered through schools + local resources
CLCP: PTSD by Age
Mean PTSD by Gender
Grade 4-7 30 25 Grade 8-12
Mean PTSD-RI
20 15 10 5 0 female Gender male
Logistic regression
Number of obs =
2252
-----------------------------------------------------------------------------ptsdcat1 | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------age | 0.78 0.03 -7.00 0.000 0.72 0.83 ------------------------------------------------------------------------------
CLCP: PTSD by Age
Severity of PTSD by School Grade
Gr 8-12
60 50
Gr 4-12
Percentage
40 30 20 10 0
none mild moderate severe very severe
PTSD Category
Logistic regression ptsdcat1 Odds Ratio Std. Err. z
Number of obs = 2250 P>z [95% Conf. Interval]
age sex
0.78 2.72
0.028 0.526
-6.84 5.16
0.000 0.000
0.73 1.86
0.84 3.97
Modelled relationship of emotional distress symptom score to class at school.
2.58 2.56
2.54
2.52
2.5
L og e IE S S U M
2.48
2.46
2.44
2.42
2.4
2.38
2.36 4 5 6 7 8 9 10 11 12
C lass at sch ool
McDermott & Palmer, 2001
CLCP: Past emotional problems
23 22
Total PTSD-RI 21 score
20
19 No Yes
Past emotional problem
Logistic regression ptsdcat1 Odds Ratio age sex prevdiff 0.68 3.29 1.16 Std. Err. z 0.08 1.03 0.42 -3.29 3.81 0.41
Number of obs = 588 P>z 0.001 0.000 0.679 [95% Conf. Interval] 0.54 1.78 0.57 0.86 6.07 2.37
CLCP: Threat perception
28 24
Total PTSD-RI score 20
16
12 No Yes
Perception of threat (self)
Logistic regression
Number of obs = 2176
ptsdcat1 Odds Ratio
age sex selfdie 0.84 2.32 8.24
Std. Err. z
0.03 0.48 1.59 -4.64 4.03 10.91
P>z
0.000 0.000 0.000
[95% Conf. Interval]
0.78 1.54 5.64 0.90 3.50 12.03
Explanatory models
Take to the Event:
•Younger age •Female •Thought would die •NOT past emotional problem
outcomes
PTSD
Event related factors
Exposure
CULTURE - SOCIETY
Sutherland Bushfire Disaster
Measure Emotional distress (IESSUM) Mean (SD) Mean (SD) (2 tailed) p _______________________________________________ Gender Male Female 12.03 (14.05) 15.54 (14.81) .000 _______________________________________________ Experience: YES NO p Evacuation 21.90 (17.10) 11.65 (13.04) 12.28 (13.08)
12.99 (14.54) 14.07 (14.54)
.000 .000
.000 .009
Child separation 16.58 (16.51)
Home damage 23.49 (18.94)
Home destroyed 22.65 (14.07)
Self may die Parent may die
21.70 (15.19) 21.29 (15.98)
15.64 (14.85) 15.25 (14.85)
.000 .000
Canberra Bushfire Disaster
Prevalence PTSD: Mild 15.8% Moderate 3% Severe 1% Increased risk PTSD: OR Thought would die 2.3 Parent die 2.1 Separation 1.6 Within 50 meters 1.6 How frightened vs PTSD p .000 .000 .02 .01 .000
Cyclone Larry: Disaster-related event
Mean Odds Ratios for PTSD by event-related variable OR X2 p 95% CI Saw debris 1.54 3.15 .074 .95 – 2.51 Home damage 2.23 14.88 .000 1.47 – 3.39 Lose part roof 2.18 20.46 .000 1.51 – 3.09 Lose whole 3.62 22.01 .000 2.03 – 6.43 Live else 3.15 14.02 .000 1.74 – 7.09
Logistic regression Number of obs = 2113
ptsdcat1 Odds Ratio Age Sex Selfdie losewhol 0.84 2.19 7.65 1.95
Std. Err. 0.03 0.47 1.53 0.66
z -4.48 3.70 10.19 1.97
P>z 0.000 0.000 0.000 0.049
[95% Conf. Interval] 0.78 1.45 5.17 1.00 0.91 3.33 11.32 3.78
Explanatory model
Take to the Event:
•Younger age •Female •Thought would die •NOT past emotional problems
outcomes
PTSD
Event related factors
Exposure •Home damage •Roof damage •Evacuation experience CULTURE - SOCIETY
CLCP: Social connectedness
28 24
Total PTSD-RI score 20
16
12 Low normal Degree of Social Connectedness high
ANOVA F 38.37, p = .0000
Number of obs = 269
Logistic regression
ptsdcat1 Odds Ratio age sex selfdie evacda schigh 0.71 4.88 7.09 4.12 4.25
Std. Err. z 0.13 2.62 3.03 2.29 2.52 -1.88 2.95 4.58 2.54 2.44
P>z 0.060 0.003 0.000 0.011 0.015
[95% Conf. Interval] 0.49 1.70 3.07 1.38 1.33 1.01 13.99 16.39 12.29 13.62
e9 e2 e1 e3 e5 e6
home damage (39%)
saw flying debris (4%)
lose part of roof (35%)
windows broken (37%)
had to live elsewhere (14%)
how frightening (34%)
think family member would die (45%)
e8
.19 .62
.59 .61 .38
z2
.56 .66 think you would die (85%)
e7
CYCLONE TRAUMA EXPOSURE
.15
CHILD THREAT PERCEPTION (23%)
.92
z1
.30 .14 -.44
.33
school activities (0%)
e13
people can talk to (11%)
e12
GIRLS' PTSD (70%)
-.69
CHILD CONNECTEDNESS
.37
friends to play with (14%)
e10
.47 trust others (22%)
e11
z4
AGE
e14 .25
z3
CLCP: PTSD by School
No difference in PTSD symptom rates were found when compared between: School systems Χ2 = .704, n.s.
Or
School sizes Χ2 = .010, n.s.
(<100 = small, >100 = large)
CLCP: Family resilience
28 24
Total FRM score
20
16
12 normal abnormal
Family resilience
Logistic regression ptsdcat1 Odds Ratio Std. Err.
Z -1.88 3.43 4.84 1.46 2.21 -0.81
Number of obs = 344 P>z [95% Conf. Interval]
age sex selfdie evacuate schigh frescat
0.71 8.87 7.30 2.03 3.68 0.49
0.13 5.64 3.00 0.98 2.17 0.43
0.060 0.001 0.000 0.144 0.027 0.420
0.49 2.55 3.26 0.79 1.16 0.09
1.01 30.86 16.35 5.24 11.67 2.72
PTSD prediction model (final)
Personal factors
•Younger age •Female •Thought would die •NOT past emot. problem •NOT fam. resilience
Exposure Event related •Home damage •Roof damage factors •Evacuation experience
PTSD (T1)
PTSD (T2)
Societal factors •Social connectedness
•NOT school size or system
SBTP therapy for primary age children: a guided therapy workbook.
Which mainly contained nothing!
As well as Therapy tasks… Article: theme random nature of event (even out of the control of the family cat) Also task of re-experiencing
Themes: Preparation for next season, again controlled re-experiencing, dealing with anniversary reactions & empowerment (internal locus of control)
A certificate from the Commissioner
FACTBOX: Changing stories like typing on a computer
If you’re typing something on the computer, you can underline certain words or sentences or put them in bold to make them stand out. And you can delete words and sentences as well – just as though they were never there. How we deal with memories is a little like this.
HINT
You may not even realise it yourself – but your friends and family may think of something you have been avoiding.
One common avoidance example is YOUR STORY So lets go over it again
Intruder thoughts: What are they?
Do you remember Peter from last session? This is what Peter had to say about one of the problems he was having after the fires:
Conclusion
From a symptomatic perspective: PTSD is common (12% severe/very severe) Individual, event-related & social independent significant associations ‘Targets’ for early identification Maybe a ‘target’ for promotion/prevention From a longitudinal & impairment perspective: More work to be done On symptom chronicity On impairment (school, social, relationships) Treatment Trauma-focused CBT adapted for children!!