Prospective dynamic assessment of risk of sexual reoffending in

Document Sample
Prospective dynamic assessment of risk of sexual reoffending in Powered By Docstoc
					 Prospective dynamic assessment of risk of
    sexual reoffending in individuals with
intellectual disability and a history of sexual
             offending behaviour.

     Rachael Lofthouse, School of Psychology, Bangor University, UK
William R. Lindsay, Castlebeck, Darlington, UK and Bangor University, UK
      Vasiliki Totsika, School of Psychology, Bangor University, UK
    Richard P. Hastings, School of Psychology, Bangor University, UK
        Douglas P. Boer, The University of Waikato, New Zealand
                James L. Haavan, Oregon, Portland, USA
          Forensic risk assessment
1. Unstructured clinical opinion
Low levels of reliability & accuracy (Monahan, 1981)
2. Actuarial assessment
      Fixed variables – e.g. number of previous
      Effective in long term prediction (Hanson &
    Brussiere, 1998; Quinsey et al., 2004)
       Initial work in ID – apply mainstream tools to ID
    offenders (Quinsey et al., 2004; Lindsay et al., 2008)
    Static assessments with ID
• Actuarial assessments show promising
  results with ID (Lindsay et al., 2008), but effect
  sizes are moderate (Lindsay & Beail, 2004)
• Opinions divided regarding use of actuarial
  assessments with ID
• Some risk factors are less important for ID
  (Lindsay et al., 2004a)
• Criticism: Don’t recognise variability in risk.
  Fail to measure response to treatment
  (Hudson et al., 2002)
            Dynamic risk assessment
• Subdivided into Stable & Acute (Hanson & Harris, 2000)
 Stable: endure e.g. attitude
 Acute – rapidly changing e.g. access to intoxicants (Lindsay et
   al., 2004b)

• Violent behaviour: dynamic ID tools lead to good prediction
  (e.g. DRAMS, Lindsay et al., 2004b; Steptoe et al., 2008)
• Dynamic variables better at predicting risk of sexual reoffending
  (Lindsay et al., 2004a)
• ARMIDILO-S (Boer et al., 2004): predicts sexual recidivism for
  ID offenders more accurately than static measures (Courtney,
  2008; Blacker et al., 2010).
        Current research project
                   Research aims
Add to the evidence for predictive accuracy of

1.   Environmental & offender factors
2.   Prospective design
3.   All items
4.   IQ scores
64 males with ID & history of SO. Average age
  32.8 years (17-63), IQ 67 (54-75)
Community ID service, Scotland, UK
Measures – baseline 2003
ARMIDILO-S (Boer et al., 2004)
30 items: divided into stable & acute; environment
  & offender.
3 scores obtained: environment, offender & total
             Measures cont.....
Actuarial assessments from file review:
• Static -99 (Hanson & Thornton, 2000)
• VRAG (Quinsey et al., 1998)

2009 Outcome variable: ‘Sexual recidivism’
Any behaviour with a sexual motive including contact &
  none contact offences, official & unofficial reports.
Predictive accuracy: Receiver Operator
  Characteristic (ROC) Area Under Curve
   ARMIDILO-S sub-scale        ROC AUC
Stable environment            .79, p<.000
Acute environment             .77, p<.000
Environmental score           .81, p<.001
Stable offender               .88, p<.000
Acute offender                .67, p<.029
Offender score                .90, p<.001
ARMIDILO total score          .92, p<.001
Figure 1 Characteristics of the receiver operating characteristic
(ROC curve for ARMIDILO-S (total score, offender subscale score
and staff & environment subscale score)
Static-99 (AUC=.74, p=.002) VRAG (AUC=.58, p=.33)
 Figure 2 Characteristics of the receiver operating characteristic
(ROC) curve for Static-99 & VRAG
   Analysis for IQ < 70 (n=42)
       Risk Assessment       ROC AUC
ARMIDILO-S environment      .79, p = .003
ARMIDILO-S Offender         .85, p < .001
ARMIDILO-S Total subscale   .90, p < .001

Static-99                   .74, p = .010

VRAG                        .56, p = .555

• Dynamic nature of ARMIDILO-S accounts for
  efficacy based on Static - 99 performance and
  previous studies (Lindsay et al., 2008)
• Consistent with previous research: dynamic items
  outperformed static (Lindsay et al., 2004a)
• Large effect sizes support notion to consider
  individuals & environment separately (Boer et al.,
• Static measures useful with ID if relevant to
      Discussion & Limitations
• Dynamic assessments = responsive risk
• RRASOR not included (Boer et al., 2004)
• Dynamic tools of value with ID offenders
• Dynamic tools address some limitations of
  actuarial tools
• ARMIDILO-S worthy of further validation
Thank you for listening

   Contact details:
Blacker J., Beech A.R., Wilcox D.T & Boer D.P. (2010) The assessment of
   dynamic risk and recidivism in a sample of special needs sexual
   offenders. Psychology, Crime and Law.
Boer G.P., Tough S. & Haaven J. (2004) Assessment of risk manageability
   of intellectual disabled sex offenders. Journal of Applied Research in
   Intellectual Disabilities, 17, 275-283.
Boer D.P., McVilly K.R. & Lambick F. (2007) Contextualising risk in the
   Assessment of Intellectually Disabled Individuals. Sexual Offender
   Treatment 2(2)
Courtney J. (2008) The identification of Recidivism Indicators in
   Intellectually Disabled Violent Individuals. Unpublished MSc thesis, The
   University of Waikato, New Zealand
Hanson R.K. & Bussiѐre M.T. (1998) Predicting relapse: A meta-analysis
   of sexual offender recidivism studies. Journal of Consulting and Clinical
   Psychology, 66, 348-362.
Hanson R.K & Harris A.J.R. (2000) Where should we intervene? Dynamic
   predictors of sexual offense recidivism. Criminal Justice and Behaviour,
   27, 6-35
Hanson R.K. & Thornton D. (2000). Improving risk assessments for sex
   offenders: A comparison of three actuarial scales. Law and human
   behaviour, 24, 119-136
Lindsay W.R. & Beail N. (2004) Risk assessment: actuarial prediction and
   clinical judgement of offending incidents and behaviour for intellectual
   disability services. Journal of Applied Research in Intellectual
   Disabilities, 17, 229-234.
Lindsay W.R., Elliot S.F. & Astell A. (2004a) Predictors of sexual offence
   recidivism in offenders with learning disabilities. Journal of Applied
   Research in Intellectual Disabilities, 17, 299-305.
Lindsay W. R., Hogue T. E., Taylor J. L., Steptoe L., Mooney P., O'Brien
   G. et al. (2008) Risk assessment in offenders with intellectual disability:
   A comparison across three levels of security. International Journal of
   Offender Therapy & Comparative Criminology, 52, 90-111.
Lindsay W.R., Murphy L., Smith G., Murphy D., Edwards Z., Chittock C. et
   al. (2004b) The Dynamic Risk Assessment and Management System:
   An Assessment of Immediate Risk of Violence for Individuals with
   Offending and Challenging Behaviour. Journal of Applied Research in
   Intellectual Disabilities, 17, 267-274
Quinsey V.L., Book A.B. & Skilling T.A. (2004) A Follow-up of
   Deinstitutionalized Men with Intellectual Disabilities and Histories of
   Antisocial Behaviour. Journal of Applied Research in Intellectual
   Disabilities 17, 243-253
Quinsey V.L., Harris G.T., Rice M.E. & Cormier C.A. (1998) Violent
   Offenders: Appraising and Managing Risk. Washington DC: America
   Psychological Association
                   ARMIDILO-S items

Stable dynamic environmental: attitudes towards ID
  sex offenders, communication amongst supervisory
  staff to ensure information sharing, client specific
  knowledge by staff, consistency of supervision,
  environmental consistency including the physical
  setting and relationships.

Acute dynamic environmental: changes in social
  relationships, supervisory staff, monitoring by staff,
  victim access, and environmental changes such as
  relocation. Changes in victim access, access to
Stable dynamic offender items: attitudes and
  compliance with supervision and treatment,
  knowledge of cognitive distortions, crime cycle, risk
  factors and relapse prevention plan, sexual
  knowledge and self management of sexuality, mental
  health, monitoring ability and self management, time
  management skills and planning ability, substance
  abuse, victim selection and acquisition/grooming
  behaviour, general coping ability and self efficacy,
  relationship and ‘relating to others’ skills, threat or
  use of violence against self or others, impulsiveness
  and offender specific stable dynamic factors, which
  are potentially numerous but can include family
  related problems for example.
Acute dynamic offender: changes in social support or
  significant relationships, changes to substance abuse
  pattern, changes in sexual preoccupation, changes in
  victim access or preoccupation with victim selection
  and acquisition or grooming of victim, changes in
  attitude or behaviour towards supervision or
  treatment, changes in ability to use coping strategies
  or recognise risky situations or failure to use problem
  solving strategies, changes to routine and offender
  specific acute dynamic variables; again there are
  numerous possibilities specific to the individual.
Lindsay et al. (2004b) - found a relationship between mood, anti
social behaviour and total DRAMS score the day before an
aggressive incident.
Steptoe et al. (2008) - mood (AUC=.74), anti-social behaviour
(AUC=.70) and intolerance/agreeableness (AUC=.71) 1- 2 days
before incidents of violent behaviour were predictive of a violent
Blacker 2010
IQ<75 (n=10) analysed separately

For this group the ARMIDILLO - Stable scale produced a
significant predictive effect (AUC=.86) as did the ARMIDILLO
acute (AUC=.75).

The actuarial measures performed little better than chance level
at distinguishing sexual recidivists from non recidivists.

Shared By: