Developments in Psychological Assessment

1
Developments in Psychological Assessment
Treatment and Management for Adult Offenders with
Developmental Disabilities4th International
Conference on the Care and Treatment of Offenders
with a Learning DisabilityUniversity of Central
Lancashire, PrestonWednesday, 6th April 2005

• Professor John L Taylor
• Northumbria University and
• Northgate Prudhoe NHS Trust
• john2.taylor_at_unn.ac.uk

2
Developments in Psychological Assessment
Treatment and Management for Adult Offenders with
Developmental Disabilities

• Definitions
• Prevalence
• Recidivism
• Risk
• psychometric
• actuarial
• clinical
• (Sex offending)
• (Fire-setting)
• Anger and aggression
• Process issues
• Conclusions

3
Acknowledgements

• Dr Bruce Gillmer, Northgate Prudhoe NHS Trust
• Professor Gregory OBrien, Northgate Prudhoe
  NHS Trust
• Alison Robertson, Northgate Prudhoe NHS Trust
• Ian Thorne, Northgate Prudhoe NHS Trust
• Professor Ray Novaco, University of California,
  Irvine, USA
• Professor Bill Lindsay, NHS Tayside and The State
  Hospital
• Dr Todd Hogue, Rampton Hospital
• Dr Sue Johnston, Rampton Hospital

4
Key References

• Frankish, P. (Ed.). (2001). Special issue on
  people with learning disabilities who offend.
  British Journal of Forensic Practice, 3.
• Fraser, W.I. Taylor, J.L. (Eds.). (2002).
  Forensic learning disabilities The evidence
  base. Supplement 1. Journal of Intellectual
  Disability Research, 46.
• Lindsay, W.R. (Ed.). (2002). Offenders with
  intellectual disability. Special issue. Journal
  of Applied Research in Intellectual Disabilities,
  15.
• Lindsay, W.R., Taylor, J.L. Sturmey, P. (Eds.).
  (2004). Offenders with developmental
  disabilities. Chichester Wiley.
• Lindsay, W.R. Taylor, J.L. (in press). A
  selective review of research on offenders with
  developmental disabilities Assessment and
  treatment. Clinical Psychology Psychotherapy.

5
Prevalence of Offending and People with DD

• Prevalence studies of offending amongst people
  with DD report
• large variations in rates depending on
• Inclusion criteria used - particularly whether
  those with borderline intelligence are included
  or not
• The type of assessment instrument used to detect
  DD (e.g. file review vs. standardised IQ
  assessment)
• Location of the sample - community vs. courts vs.
  prisons vs. secure hospital settings
• Study design and methodology (e.g. case note
  review vs. clinical evaluation vs. informant
  survey sampling method)
• Changes/differences in criminal justice, social
  health care policies affect apparent incidence,
  visibility and reporting (e.g. de-institutionalisa
  tion Lund, 1990).

6
Prevalence of Offending and People with DD
/contd.

• Prevalence of offending among people with DD is
  difficult to estimate with any degree of accuracy
  
  (Lindsay, Taylor Sturmey, 2004)
• The evidence base is poor with regard to
  epidemiological studies - in particular there is
  dearth of well-controlled studies including
  comparison reference groups.
• (Lindsay Taylor, in press)
  
• It is not clear, therefore, whether people with
  DD are over- or under-represented in the offender
  population, and whether offending is more
  prevalent among people with ID than the general
  population
  (Day, 1993 Holland et
  al., 2002 Simpson Hogg, 2001)

7
Recidivism Amongst Offenders with DD

• Studies of recidivism amongst offenders with DD
  have reported rates between 39 (Walker McCabe,
  1973) and 72 (Lund, 1990)
• There are major problems in interpreting the
  findings of recidivism studies (Lindsay Taylor,
  in press)
• Recidivism rates for DD offenders subject to
  probation orders are high, but no higher than for
  offenders in general (Linhorst et al., 2003)
• Voluntary clients have markedly higher recidivism
  rates than those mandated to community case
  management programmes by the courts (Linhorst et
  al., 2003)
• Longer-term treatment and supervision under
  probation orders reduces recidivism of SXO
  (Lindsay et al., 1998)
• Treated sex offenders followed up over 4 years
  showed re-offending rates of 4, 12, 13 and 21
  over this period (Lindsay et al., 2002)

8
Risk Assessment in Offenders with DD

• 2 recent reviews focusing on risk assessment in
  offenders with ID
• Johnston (2002) Quinsey (2004)
• A special issue of JARID on risk assessment in
  forensic DD (Eds. Lindsay Beail, 2004)
• Psychometric Risk Assessment
• Staff-rated risk using the Short Dynamic Risk
  Scale (SDRS) is predictive of community
  offending-type behaviour, including violence
  (Quinsey, 2004)
• Fire-setting is closely associated with
  antecedents of anger, low social attention and
  low mood as measured on the FSAS (Murphy Clare,
  1996 Taylor et al., 2002)
• Self- and staff-rated anger using the NAS and
  WARS is predictive of inpatient violence (Novaco
  Taylor, 2004)

9
Hierarchical Regression of Violence Risk and
Anger Predictors of Patient Assaultiveness in
HospitalRef Novaco Taylor (2004) in
Psychological Assessment
Predictors beta t R2 R2 Change
F change p Step 1 Age
-.148 1.47 WAIS-R (Full Scale) -.214 2.12 Violen
ce Offence .143 1.45 .081 .081 2.77 (3,95)
.046 Step 2 NAS Total .369 3.95 .211
.131 15.59 (1,91) .000 Step
3 Extraversion (EPQ) .224 2.43 .258 .047
5.90 (1,93) .017 Note The dependent
measure is the number of assaults since hospital
admission (square root transformed). At Step 3,
STAXI Trait Anger and Anger Expression, and the
EPQ-Lie scale were statistically excluded in the
stepwise procedure. For the final model
including the covariates, NAS Total, and EPQ-E, R
.508, F (5,93) 6.48, p .000
10
Risk Assessment in Offenders with DD
Actuarial Assessment

• The Applicability of Personality Disorder and
  Risk
• Assessment (DSPD) Measures in a Sample of
• Intellectual Disability Offenders
• Study funded by the Home Office
• Grant No. RDS/01/247
• Dr Todd Hogue Rampton Hospital
• Dr Sue Johnston Rampton Hospital
• Professor John Taylor Northgate Prudhoe NHS
  Trust
• Professor Greg OBrien Northgate Prudhoe NHS
  Trust
• Professor Bill Lindsay NHS Tayside
• Dr Anne Smith NHS Tayside

11
Study Participants and Sites

• 212 men with ID/developmental disabilities and
  offending and offending-type histories from 3
  sites
• Rampton Hospital (High Security) N 73
• Northgate Prudhoe Hospitals (Medium Low
  Security) N 70
• Tayside (Community Forensic Service) N 69
• Mean Age 37.5 years (SD 11.4 Range 18 69
  years)
• Mean Full Scale IQ 65.8 (SD 8.6 Range 43
  89)
• Mean Length of Stay 8.1 years (SD 7.5 Range
  1 26 years)

12
Risk Instruments

• Violence Risk Appraisal Guide (VRAG Harris, Rice
  Quinsey, 1993) - actuarial
• Static 99 (Hanson Thornton, 2000) - actuarial
• Risk Matrix 2000 (Thornton, 2000) - actuarial
• Individual Psychological Risk Factor (IPRF
  Hogue, 2003) - clinical
• Short Dynamic Risk Scale (SDRS Quinsey, 2003)
  clinical
• HCR-20 (Webster, Eaves, Douglas Wintrup, 1995)
  actuarial and clinical

13
Results HCR-20 by Site

• Rampton Northgate mean scores were
  significantly higher than Tayside on the
  Historical scale (F 42.95, df 3, p lt .001)
• Rampton mean scores were significantly higher
  than Northgate Tayside on the Risk Management
  scale (F 3.19, df 3, p lt .05)
• There were no differences in mean scores on the
  Clinical scale across sites (F 1.56, df 3,
  p .20)

14
Correlations between HCR-20, other Risk
Measures and Violent/Aggressive Incidents


HCR-20
Historical Clinical
Risk Managt. VRAG

.65 .23
.22 Short Dynamic Risk Scale
.39 .42
.27 Total Incidents
.28
.28 .20 Note. p lt .01. All
correlations are Spearman Rho, two-tailed tests.
N 151 203.
15
Correlations between HCR-20 and EPS Scales


HCR-20
Historical Clinical
Risk Managt. EPS Scales Physical
Aggression
.32 .35
.23 Verbal Aggression
.30 .37
.19 Anxiety
.12
.18 .15 Withdrawal

.17 .13
.03 Note. p lt .01. All correlations are
Spearman Rho, two-tailed tests. N 154 171.
16
Mean HCR-20 Scores, Grouped According to
Conviction for Violent Offences
17
Mean HCR-20 Scores, Grouped According to
Violent/Aggressive Incidents During Previous 12
Months
18
Conclusions HCR-20 for Offenders with DD

• HCR-20 scales show good concurrent validity with
  conceptually relevant clinical and actuarial risk
  measures
• The scales have good levels of discriminant
  validity, as measured against EPS clinical and
  aggressive behaviour indices
• HCR-20 scales correlate significantly with
  proximal violent incident data
• HCR-20 differentiates between clients with
  convictions for violence (and those who have been
  violent recently) and non-violent clients in a
  logically consistent manner

19
Risk Assessment in Offenders with DD Treatment
PlanningRef Taylor, J.L. Halstead, S. (2001).
Br. J. of Forensic Practice

• Case Study - Mr L.
• 27 years old
• Admitted from prison under s.47 of the MHA 1983
• Convicted of Indecent Assault reduced from
  Attempted Rape
• Victim sister-in-law, 13 years old
• Sentence 6 months imprisonment ) reduced on
  appeal from 18 months)
• Married for 3 years
• Wife, 19 years old
• 1 son, 2 years old
• 1 daughter, 1 year old
• Full Scale IQ 67 Level 2 Social Reasoning

20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
Some Conclusions about Forensic-Clinical Risk
AssessmentRef Taylor, J.L. Halstead, S.
(2001). Br. J. of Forensic Practice

• Clinicians tend to avoid systematic risk
  assessment
• There is a gap between practice and the science
  of risk assessment
• This results in clients being denied effective
  assessments and targeted interventions to reduce
  their risks
• Use of shared clinical models that allow for risk
  analysis is one way forward
• This approach could enable the development of
  clinically defensible judgements concerning
  patients risks
• More research and evaluation is required
  concerning the implementation of this approach
  and its predictive validity

24
Sex Offenders and Fire-Setters with DD - Key
References

• Sex Offenders
• Lindsay, W.R. (2002). Research and literature on
  sex offenders with intellectual and developmental
  disabilities. J. of Intell. Disability Research,
  46 (Suppl. 1), 74-85.
• Lindsay, W.R. (2004). Sex offenders
  conceptualisation of the issues, services,
  treatment and management. In Lindsay, Taylor
  Sturmey (Eds.), Offenders with developmental
  disabilities. Chichester Wiley.
• Courtney, J. Rose, J. (2004). The effectiveness
  of treatment for male sex offenders with learning
  disabilities A review of the literature. Journal
  of Sexual Aggression, 10, 215-236.
• Fire-Setters
• Taylor, J.L., Thorne, I., Robertson, A. Avery,
  G. (2002). Evaluation of a group intervention for
  convicted arsonists with mild and borderline
  intellectual disabilities. Crim. Behaviour and
  Mental Health, 12, 282-293.
• Taylor, J.L., Thorne, I. Slavkin, M.L (2004).
  Treatment of fire-setting behaviour. In Lindsay,
  Taylor Sturmey (Eds.), Offenders with
  developmental disabilities. Chichester Wiley.

25
Anger Aggression in People with DD

• Aggression is a common feature of populations of
  people with DD (Deb et al., 2001 Hill
  Bruininks, 1984 Sigafoos et al., 1994 Smith et
  al., 1996 Taylor et al., 2004)
• Prevalence in the UK DD population is 12-22
  (est. 144,000 240,000 individuals)
• Physical violence is a significant
  clinical/management problem in people with DD and
  forensic histories in institutional settings
  (Novaco Taylor, 2004)
• Aggressive behaviour presents significant
  problems for staff in DD services (Bromley
  Emerson, 1995 Jenkins et al., 1997 Kiely
  Pankhurst, 1998).
• Anger is a significant activator of, and is
  predictive of violence in psychiatric, forensic
  and DD populations (Novaco, 1994 Novaco
  Renwick, 2002 Novaco Taylor, 2004)

26
Northgate Anger Treatment Project

• Stage 1
• diagnostic/screening assessment of a group of 129
  detained DD men with offending histories to
  examine the nature and scope of anger problems in
  this population and to investigate the
  psychometric properties of several criterion
  measures of anger and aggression
• Stage 2
• development of an anger treatment protocol
  designed specifically for people with DD and
  histories of aggression and offending behaviour
• Stage 3
• evaluation of a cognitive-behavioural anger
  treatment by comparison of post-treatment
  measures in the treatment group with
  pre-treatment measures in the waiting list
  control group

27
Cognitive-Behavioural Treatment of Anger for
People with DD

• Research on anger treatment for people with DD is
  limited, but there is some evidence of successful
  CBT-based interventions (see Taylor, 2002
  Whitaker, 2001 for reviews)
• There are 7 small anger CBT outcome studies with
  DD clients that involved comparison groups
  (Benson et al., 1986 Lindsay et al., 2004 Rose
  et al., 2000, Taylor et al., 2002, 2004, in
  press Willner et al., 2002)
• There are some reports in the literature of CBT
  for anger in offenders with DD (Allen et al.,
  2001 Lindsay et al., 2003, 2004 Taylor et al.,
  2002, 2004a, in press)

28
Three Concatenated Anger Treatment Outcome
Studies - Research Design Analysis

• Wait-list controlled design (as considered
  unethical to withhold a potentially effective
  treatment from those who might benefit from it)
• Both groups continued to receive treatment as
  usual
• Patients meeting inclusion criteria allocated to
  the Anger Treatment (AT) group or Routine Care
  (RC) conditions

29
Anger Treatment for ID Offenders I

• Modification of Novacos (1975, 1993) treatment
  protocol
• Treatment is delivered individually by a
  qualified psychologist over 18 sessions (twice
  weekly)
• 6 session preparatory phase (psycho-educational)
• 12 sessions of treatment proper (cognitive
  re-structuring, arousal reduction skills
  training)
• Emphasises collaboration, personal
  responsibility, self-control the legitimacy of
  anger
• Utilises a range of assessment, educational
  training materials adapted to help patients with
  LD engage in the treatment process

30
Anger Treatment for ID Offenders II

• Key components of the treatment
• Analysis and formulation of individual patients
  particular anger problems
• Cognitive re-structuring
• Self-monitoring of anger frequency, intensity and
  triggers
• Construction of a personal provocation hierarchy
• Arousal reduction techniques
• Training behavioural coping skills
• Development of personalised self-instructions to
  prompt coping
• Stress inoculation to practice coping in
  imagination

31
Mean Novaco Anger Scale (NAS) Total scores over
Time ANCOVA (WAIS-R IQ as covariate) F(1,33)
4.74. p lt .05, r .35Ref Taylor et al. (in
press). Brit. J. of Clinical Psychology

32
Mean Provocation Inventory Total Scores over
TimeANOVA, F (1,17), 13.56, p lt .005, r
.66Ref Taylor et al. (2002). J. of Applied
Research in Intell. Dis., Vol. 15
33
Mean IPT Anger Composite Scores over Time
ANCOVA (Time 1 score as covariate) F(1,14)
11.20. p lt .01, r .67 Ref Taylor et al.
(2004). Clinical Psychol. Psychotherapy, Vol.
11
34
Mean WARS Anger Index scores over
TimeANCOVA (WAIS-R IQ as covariate) F(1, 33)
1.49, p lt .23 Ref Taylor et al. (in press).
Brit. J. of Clinical Psychology
35
Process Issues Related to Psychotherapy for
People with DD

• Can people with DD reliably engage in the
  collaborative, shared formulation approach of
  CBT?

36
Treatment Completers (n 18) Evaluations of
Preparatory and Treatment Phases of Anger
Treatment (PEAT Questionnaire Responses)
37
Patient G - Background Information

• Age 27 years
• Full Scale IQ 68
• Psychiatric Diagnosis Mild Learning Disability
• MHA Section 37 Hospital Order
• Length of stay in hospital 8.5 years
• Index Offence(s) Indecent assaults against
  young boys
• Rehabilitation Status Longer-stay low secure
  (slow-track rehabilitation)
• Previous Psychological Intervention 1) Positive
  response to an individual behavioral programme to
  reduce interpersonal conflict 2) Completion of
  group-based sex offender treatment programme with
  mixed outcomes

38
(No Transcript)
39
Process Issues Related to Psychotherapy for
People with DD

• Can people with DD reliably engage in the
  collaborative, shared formulation approach of
  CBT?
• Can people with DD do the cognitive component of
  CBT?

40
Cognitive-Behavioural Model of Emotion Simple
Linear
A
B
C
Events
Thoughts
Feelings
Behaviour
41
(No Transcript)
42
Process Issues Related to Psychotherapy for
People with DD

• Can people with DD reliably engage in the
  collaborative, shared formulation approach of
  CBT?
• Can people with DD do the cognitive component of
  CBT?
• What is the effect of IQ on treatment outcome?
• Can treatment effects be achieved and sustained
  in routine practice settings?

43
Mean NAS Total scores for AT (n 16) and RC (n
20) groups over time.
44
Do you think you have learned anything about
anger treatment from your involvement in the
project?(Nurses Responses n 14, Mean 3.7)
45
Has your involvement (in the treatment) had an
effect on the way you deal with other patients
anger problems?(Nurses Responses n 14, Mean
3.1 )
46
Process Issues Related to Psychotherapy for
People with DD

• Can people with DD reliably engage in the
  collaborative, shared formulation approach of
  CBT?
• Can people with DD do the cognitive component of
  CBT?
• What is the effect of IQ on treatment outcome?
• Can treatment effects be achieved and sustained
  in routine practice settings?
• Can treatment effects be maintained over time and
  across settings?

47
Summary

• It is not clear whether people with DD are over-
  or under-represented in the offender population,
  and whether offending is more prevalent among
  people with ID than the general population
• There are major problems in interpreting the
  findings of recidivism studies but mandated and
  longer-term treatments produce better outcomes
• Good progress is being made in developing
  psychometric, actuarial and clinical assessments
  of forensic risk
• The evidence for the effectiveness of
  interventions for sexually aggressive and
  fire-setting behaviour is building but lacks
  methodological rigour
• The evidence for anger treatment is best
  developed and continues to build the
  sustainability and impact on aggression is not
  clear
• Offenders with DD appear to be able to engage in
  and benefit from CBT approaches

48
Future Research Directions for Offenders with DD

• More research concerning reliable and valid
  assessments is required to facilitate risk
  assessment and management, treatment planning,
  and evaluation of effectiveness
• Larger, more powerful and better designed
  controlled trials are required to show if the
  effects obtained to date are stable phenomena
• The mechanisms underlying treatment gains are
  not clear - component analysis and longer-term
  follow-up is required to evaluate these issues
• Process issues relating to optimum length of
  treatment, mode of delivery of treatment (group
  vs. individual), systematic involvement of carers
  and relative costs need more enquiry
• The sustainability and generalisability of
  treatment gains are not proven research into
  optimum maintenance schedules is indicated

49
Contact Details

• Professor John L Taylor
• Head of Psychological Therapies Research
• Northgate Prudhoe NHS Trust
• Northgate Hospital
• Morpeth
• Northumberland
• NE61 3BP
• Tel [phone]
• john2.taylor_at_unn.ac.uk