Post on 16-Dec-2015
Clinician-judged attachment narrative style & the course and outcome of psychodynamic
therapy in people with intellectual disabilities
Allan Skelly Caitriona Collins Mandip Dosanjh
Psychodynamic Psychotherapy
• Efficacy reviewed by Jonathon Shedler in American Psychologist Feb-March 2010
• Effect sizes of psychodynamic therapy in meta-analytic reviews: median d=.69 to 1.46, depending on methodology [74 studies]
• Effect sizes of CBT median d=0.58 to 1.0 depending on methodology [95 studies]
• Effect sizes of antidepressants approved by the FDA: median d=.17 to .31 [83 studies].
Psychological therapy and people who have intellectual disability
Prout, H.T. and Nowak-Drabik, K.M. [2003] Am Jnl Men Rtdn, 108, 2:82-93.
“A moderate degree of change in outcome measures and moderate effectiveness in terms of benefit to clients”
Prout, H.T. and Browning, B.K. [2011] Adv Mntl Hlth & Int Dis, 5, 5, 53-59.
“psychotherapy, as broadly defined, appears to be at least moderately beneficial”.
What is meant by “psychodynamic therapy”?
1. Focus on affect and the expression of emotion2. Exploration of attempts to avoid distressing thoughts
and feelings3. Identification of recurring thoughts and patterns in the
patient’s ideas4. Discussion of past experience and development5. Focus on interpersonal relations6. Focus on the relationship between therapist and patient. 7. Exploration of fantasy life, e.g. in dreams, ideas and
comments.[From
Shedler, J. 2010]
Attachment measure:Quality of Early Relationships Rating Scale [QuERRS]
• Measuring what?• Narrative style ≠ attachment behaviour• QuERRS: Skelly, A. and Reay, R. [2013] Clin Psy &
P w LD, 11, 1-2.:– 20-item therapist rating scale based on at least 2
assessment sessions where childhood was discussed
– Internally consistent– IRR is good [95+% concordance on sec-insec, 80+
% on category A-D]– Seems to be predictive of HoNOS-LD case severity
Why is the person’s attachment narrative important?
• QuERRS predicts initial severity on HoNOS-LD in terms of identified factors [c.f. Skelly & D’Antonio, 2008]:
• externalising, • internalising, and • social & adaptive aspects of well-being• BUT NOT cognitive status [CCC factor and also IQ
control analysis].
Psychotherapy process measurement can be objective
• Personal Relatedness Profile [Hobson, R.P., Patrick, M.P. and Valentine, J.D.] BJ Psych, 173:172-177.
- A therapist-rated scale of session quality in terms of [a] paranoid-schizoid functioning, and [b] depressive functioning.
- inter-rater reliability established- 2 Factors emergent from Factor analysis that were negatively
correlated- Borderline Personality Disorder diagnosis associated with more
‘paranoid-schizoid’ functioning than in patients receiving a diagnosis of depression
Group process measure
• A short measure intended to test for increasing group ‘membership’
– Empathy– Comments to each other in support– Reality rather than fantasy
• Inter-rater reliability kappa=0.71 to 0.92• Remains to be validated
Patients included in our analysis
• Individual psychotherapy [n=35]• Group psychotherapy [n=6]
– Looked more closely at process on PrP and group measure
• Psychodynamic staff support [n=13]• Overall n=54
• Not randomised allocation to treatment but outcome independently measured; i.e. practice based or ‘service evaluation’ design – a limitation on the findings
Efficiency of psychotherapy: number of sessions ("dose")
5
19
6
3 3 34
01
0 0 0 0 01
0 to 4 5 to 9 10 to 14 15 to 19 20 to 24 25 to 29 30 to 34 35 to 39 40 to 44 45 to 49 50 to 54 55 to 59 60 to 64 65 to 69 70 to 74
Number of sessions
Median number of sessions = 8 Mean number of sessions = 13.7
Attachment and Outcome
CHANGE IN HoNOS-LD TOTAL SCORE BY ATTACHMENT CLASSIFICATION
0
5
10
15
20
25
T1 T2
SECURE
INSECURE
TIME F(1,52)=31.85, p<0.01, d=.85
Interaction F(1,52)=5.86, p<0.02
Attachment & Outcome
CHANGE IN HoNOS-LD COG /COMM COMP BY ATTACHMENT CLASSIFICATION
0
0.5
1
1.5
2
2.5
3
T1 T2
SECURE
INSECURE
TIME F(1,52) = 0.49, nsINTERACTION F(1,52) =0.85, ns
Attachment and Outcome
CHANGE IN HoNOS-LD BMR SCORES BY ATTACHMENT CLASSIFICATION
0
1
2
3
4
5
6
7
8
T1 T2
SECURE
INSECURE
TIME F(1,52) = 19.54, p<0.01, d=0.68INTERACTION F(1,52)=10.23, p<0.01
Attachment and Outcome
CHANGE IN HoNOS-LD LAF SCORES BY ATTACHMENT CLASSIFICATION
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
T1 T2
SECURE
INSECURETIME F(1,52)=15.65, p<0.01, d=0.60
INTERACTION F(1,52)=1.12, ns
Attachment and Outcome
CHANGE IN HoNOS-LD INTERNAL DYSREG SCORES BY ATTACHMENT CLASSIFICATION
0
0.5
1
1.5
2
2.5
3
3.5
T1 T2
SECURE
INSECURETIME F(1,52)=24.61, p<0.01 d=0.74INTERACTION F(1,52)=1.40, ns.
Process: improvement in emotional processing
Accumulated mean "paranoid-schizoid" scores from the Personal Relatedness Profile [n=6]
35
45
55
65
Session 1-5 Session 6-10 Session 11-15 Session 16-20
Session progression
Mean improvement 8.97 points
d =0.91 if session 16-20 mean seen as 'outcome'
Mean phased average ratings for the whole group for each rater [mean kappa = 0.88]
6
8
10
12
14
16
18
20
Aug 12 to Oct12 Nov 12 to Jan 13 Feb 13 to Apr 13 Apr 13 to Aug 13
mean r1
mean r2
Group process measure – development of group membership
“Dose effects”
• Those classed as insecurely attached attended more sessions [on average 5 more sessions] but not significantly more [t(43)=1.42, ns]
• There was a positive association between number of sessions and change on the HoNOS-LD Total Score [rho(45) =.44, p<0.01].
Narrative style suggested attachment categories
Our clinical sample
Unresolved31%
Hostile17%
Preoccupied15%
Autonomous37%
Approx. proportion of attachment style in literature [not ID: Steele, 2002]
Hostile
Unresolved
Preoccupied
Autonomous
Conclusions: outcome of psychodynamic therapy and people with ID
• Psychodynamic therapy has efficacy generally.
• Psychodynamic therapy worked for people with ID in our service effectiveness.
• Effect sizes were good [for total HoNOS-LD scores d=0.85].
• Therapy was efficient with a median of 8 sessions, though some people need much more than this.
Conclusions: therapist rated attachment style
• Attachment representations rated by clinician predicted initial severity
• On some aspects of outcome, the attachment representations rated by clinician indicate a different course [sharper improvement]
• The number of sessions has some relationship to the amount of change – i.e. there was a possible effect of ‘dose’
Conclusions – group in detail
For our group examined more detail
• Progress coincides with less paranoid-schizoid rating on the PRP
• Progress coincides with more depressive ratings on the PRP to be confirmed – data awaiting analysis
• The group measure showed increasing levels of involvement, empathic comments about each other, affection and cohesion towards each other.
• But not coherent attendance. Discussion point: “internal model of the group” may be more important than actual attendance.
Control analyses
Outcome & age – no relationship [r=0.09].
Attachment & gender – no relationship[chi-square (1,54) =0.5, ns].
Outcome & gender – no relationship[e.g for HoNOS-LD total score:
F(1,53)=0.06, ns].
Free use of the QuERRS for hcpc registered psychologists
allan.skelly@ntw.nhs.uk
Additional slide – HoNOS-LD factor scores with waiting list
[n=6] & follow up [n=8] data
0
1
2
3
4
5
6
7
Time 0 Time 1 Time 2 Follow up
BMR
ADAPT
INT DYS
Additional slide: Dose effect
Is there a relationship of number of sessions and amount of change?
0
10
20
30
40
50
60
70
80
-15 -10 -5 0 5 10 15 20 25 30
change on HoNOS-Ld total scores
no
of
se
ss
ion
s
rho=.44