Forensic Psychology
Page 98
Using a research design in which the control participants were placed on a waiting list, Taylor, Novaco, Gillmer, and Thorne (2002b) evaluated the effectiveness of cognitive behavioural treatment for anger. Twenty adult men with mild ID were employed as participants. All had been detained under sections of the England and Wales Mental Health Act (1983) for reasons of violence and aggression. Ten were allocated to the AMT condition while 10 were placed on the waiting list. The programme involved a preparatory phase of six sessions designed to present information on the nature and purpose of anger treatment, encourage more division to change and develop basic skills including self-disclosure, emotional awareness and self-monitoring. Participants then proceeded to the 12-session treatment phase that included core components of cognitive restructuring, arousal reduction and behavioural skills training. All participants were evaluated using a version of the Novaco Provocation Inventory (Novaco, 1988) and the WARS. Participants’ reported anger intensity was significantly lower following the anger treatment, compared to the routine care waiting list control condition. Key nurse ratings were taken following the completion of treatment and at one-month follow-up. The means for the post-treatment ratings suggested modest improvement and these improvements were maintained to the follow-up assessment. None of the participants in the AMT condition became worse on any of the dimensions after treatment whereas the self-report ratings for the control group participants were slightly poorer at the post-treatment assessment point.
In a study employing participants referred to a community forensic ID service, Lindsay et al. (2004) compared 33 court-referred participants who received AMT with 14 who made up a waiting list control condition. Treatment lasted for 40 sessions and control participants were on the waiting list for around six months. For the AMT group, there were significant within group improvements on all measures. In post-treatment comparisons between the two groups, there were no differences on the Provocation Inventory while the self-report diaries and the anger provocation role-plays showed significant differences between groups. These improvements were maintained at the 15-month follow-up. The study also reported on the number of aggressive incidents and re-offences recorded for both groups. At the post-waiting-list assessment point (six-months follow-up), 45% of the control group had committed a further assault while at the post-treatment assessment point (nine months) 14% of the treatment group had committed further incidents of assault. Therefore, although the follow-up period was relatively short, there was some evidence that AMT had a significant impact on the number of aggressive incidents recorded in these participants.
In an extension of the Taylor et al. (2002) study, Taylor et al. (2005) reported on a larger study with 40 men who had mild to borderline ID and histories of offending. All participants were detained in a specialist forensic ID service under sections of the Mental Health Act (1983). The intervention employed the same detailed protocol (Taylor & Novaco 2005) with 20 participants allocated to the AMT condition and 20 allocated to routine care waiting list control condition. Scores on self-reported anger disposition and reactivity indices were significantly reduced following intervention in the treatment group compared with scores for the control group and these differences were maintained to four-month follow up. Staff ratings of study participants’ anger disposition converged with self-report ratings but did not reach statistical significance. Willner et al. (2013) conducted a randomised controlled trial of AMT with 179 people with IDD identified as having problems with anger. Treatment lasted 12 weeks, was delivered by care staff who were guided by a treatment manual and assessments were conducted at pre-treatment, post-treatment and 10-months follow-up. There were no significant improvements on the primary provocation self-report measure and any treatment gains were modest. However, keyworkers reported that treated individuals had significantly lower provocation scores and treated individuals reported significantly greater use of coping skills following treatment. These studies are the most extensive and well controlled pieces of research in the field of AMT evaluation.
A number of treatment studies have now been conducted on the anger and violence in offenders with ID. The studies by Taylor and colleagues (Taylor et al. 2002, 2004, 2005) evaluated AMT on detained and violent offenders in a secure setting demonstrating significant improvements on self-reported measures of anger and behavioural reaction indices with improvements maintained for up to four months following treatment. Supporting this body of work are a number of controlled research studies from different treatment centres, notably Willner and colleagues. These not only demonstrate the superiority of anger treatment over control conditions but also suggest that treatment produces a reduction in the number of incidents perpetrated by participants. This body of evidence suggests that AMT interventions are effective for offenders with ID.
CASE STUDY 21.1 ANGER/VIOLENCE PROBLEMS
Bill is a 28-year-old man with an IQ of 71, referred from supported accommodation because of extreme violence. One evening he had been playing on his computer when staff asked him to help prepare dinner. He complained was being treated like a slave, the incident escalated and attacked two members of staff, barricaded himself in his room and the police were called. Next day he appeared in court for breach of the peace. Bill has been arrested on three previous occasions for altercations/assaults. After appearing in court, Bill began anger management treatment (AMT) as a condition of a probation order. Treatment followed the standard AMT procedures introducing the concept of anger is an emotion, beginning some anxiety reduction exercises, and developing an understanding of how to record incidents daily and weekly. The next phase of treatment involved analysing the components and functions of anger. There was also an analysis of the way in which threat was perceived in a range of situations both general and individual. For the final phase, a number of techniques were involved including the construction of a range of anger provoking situations, an analysis of the situations that are actually threatening, and stress inoculation in imagination, and through role-play so that Bill could develop coping strategies for individual situations.
The Novaco Anger Scale (NAS) was used throughout treatment to review Bill’s feelings of anger and reaction provocation. The repeated measures can be seen in Table 21.1. The NAS contains three anger expression scales and three anger regulation scales: cognitive, behavioural and arousal respectively. The lowest possible score on any of the expression scales is 16 while the poorest score, indicating no anger regulation strategies, on each regulations scale, is 4. As can be seen, Bill reported extremely high levels of anger expression on all three scales at both baseline assessments (one taken prior to attending court and the second prior to commencement of treatment). As treatment progressed, and as his understanding of the emotion of anger developed, his anger expression scores reduced fairly regularly until they were to relatively manageable level. It is interesting that his reports reflecting cognitions related to anger maintained at a higher level than the behavioural and arousal domains. He continued to maintain that he felt justified if he was going to be angry and that there were a number of people it would be impossible for him to trust. These cognitions contributed to this higher score on the cognitive expression domain. These improvements maintained to follow-up at one month and six months.
Table 21.1 Use of the Novaco Anger Scale (NAS)
NAS scale
Base- line1
Base- line 2
3 months
6 months
Post- treatment
Follow-up 1
Follow-up 2
Anger expression
Cognitive 42
44
35
32
30
34
33
Behavioural 40
45
38
25
23
24
25
Arousal 38
45
30
25
24
22
&
nbsp; 22
Anger regulation
Cognitive 4
4
4
6
9
9
10
Behavioural 4
4
5
6
7
9
10
Arousal 4
4
5
7
7
9
8
At first, Bill found it completely impossible to generate any regulations skills or coping skills that he might use when he was angry. As treatment developed he began to recognise and report a number of coping strategies and these are reflected in his course on the anger regulation domains. Once again, these improvements maintained until one- and six-month follow-up. During this period, nine months of treatment and six-months follow-up, there were no reported aggressive incidents that require the intervention of the police. Indeed, staff reported that he maintained generally good relationships with those around him. There were periods when they observed that he was pent-up with rage and commented that he was able to control his anger without having an outburst of verbal or physical aggression. This was a huge improvement on his normal way of dealing with what they perceived as provocation and imitations.
21.5.3 Treatment of Sexual Offenders with ID
Early reports of treatment focussed on skills training and sex education. Griffiths et al. (1989) described a series of 30 cases to illustrate their methods and they also reported no reoffending after one-year follow-up. Haaven, Little, & Petre-Miller (1990) described a wide-ranging series of treatments including social-skills training, sex education, and the promotion of self-control in a comprehensive programme that addressed sexual offending under a behavioural management regime. The report was related to a secure inpatient treatment unit and participants were supervised constantly throughout the follow-up period and had little opportunity to reoffend.
However, the most significant development in the field of sex offender treatment has been based on problem-solving and cognitive behavioural therapy (CBT) techniques. The application of CBT to sexual offenders with ID came somewhat later than in mainstream sex offender research. One of the first studies was by O’Conner (1996), who developed a problem-solving intervention for 13 adult male sex offenders with ID. This involved the consideration of a range of risky situations in which offenders had to develop safe solutions for both themselves and potential victims. She reported positive results from the intervention with most participants having achieved increased community access.
In a series of case studies, Lindsay and colleagues (Lindsay, Marshall, Neilson, Quinn, & Smith, 1998a; Lindsay, Neilson, Morrison, & Smith, 1998b; Lindsay, Olley, Jack, Morrison, & Smith, 1998c; Lindsay, Olley, Baillie, & Smith, 1999) reported the development of treatment based on cognitive principles in which various forms of denial and mitigation of the offence were challenged over treatment periods of up to three years. They noted several aspects of treatment that involved the adaptation of basic principles from mainstream work to that of offenders with ID, and these principles and adaptations remain to the present day (Lindsay, 2009; Rose, Rose, Hawkins, & Anderson, in press). In the case studies reported by Lindsay and colleagues, all participants showed reductions in aberrant cognitions, and when followed up for a minimum period of four years, they found that only one individual had reoffended and one individual was suspected of reoffending.
One of the difficulties in evaluating outcomes in several treatment studies is illustrated by the report on six sex offenders with ID by Craig, Stringer, and Moss (2006). They conducted a seven-month treatment programme incorporating sex education, addressing cognitive distortions, reviewing the offence cycle and promoting relapse prevention. They found no significant improvement on any of the measures but in a follow-up period of 12 months, they reported no further sexual offending. However, they also noted that all participants receive 24-hour supervision, when individuals are continually supervised in the community or elsewhere; presumably, they have little opportunity to engage in any inappropriate behaviour including sexual behaviour. Therefore, the value of outcome data in these studies is limited.
The most important information to be reported in terms of social policy is the extent and seriousness of any further incidents following treatment (Lindsay & Beail, 2004). Ideally, the effects of treatment should be evaluated in the absence of other major variables such as supervision and escort. This criticism could be levelled at a number of reports (O’Conner, 1996, Haaven et al., 1990). By contrast, in the reports by Lindsay and colleagues (Lindsay et al. 1998a, 1998b, 1998c, 1999) it is noted that all individuals had free access to the community and lived in a range of community settings.
Rose, Jenkins, O’Conner, Jones, and Felce (2002) reported on a 16-week group treatment for five men with ID who had perpetrated sexual abuse. The group treatment included self-control procedures, consideration of the effects of offences on victims, emotional recognition and strategies for avoiding risky situations. Participants were assessed using the QACSO attitudes scale, a measure of locus of control, a sexual behaviour and the Law measure and the victim empathy scale. Significant differences from pre-to post-treatment were found only on the locus of control scale. However, they reported that participants had not reoffended at a one-year follow-up. In an extension of their work, Rose et al. (2012) reported on a six-month treatment group for sex offenders living in the community. Basing part of their programme on the theoretical writing of Lindsay (2005, 2009), they made efforts to involve aspects of the offender’s broader social life into treatment by inviting carers to accompany the participants. They found significant improvements on the QACSO scale, changes in the Locus of Control measure, unfortunately, towards more external locus of control, and no reoffending at one-year follow-up. Craig et al. (2012) evaluated a programme for 14 sex offenders with IDD living in the community. They completed assessments of sexual knowledge, empathy and the QACSO. Sessions were conducted weekly for 14 months and they found significant improvements in attitudes (QACSO), victim empathy and sexual knowledge. They followed up all participants for six months and six participants for 12 months. All had access to the community and so were able to commit further incidents. No further incidents of ISB were reported.
Although there are number of treatment comparison studies evaluating the effect of sex offender treatment, they all tend to fall well short of adequate experimental standards and it is important to consider any results in the light of the methodological shortcomings. The strongest comparison is by Lindsay and Smith (1998) when they compared seven individuals who had been in treatment for two or more years with another group of seven who had been in treatment for less than one year. There were no significant differences between the groups in terms of severity or type of offence. The group who had been in treatment for less than one year showed significantly poorer progress and those in this group were more likely to reoffend, than those treated for at least two years. Therefore, it seemed that shorter treatment periods might be of limited value for this client group.
Keeling, Rose, and Beech (2007b) compared 11 “special needs” sexual offenders, with 11 mainstream offenders, matched on level of risk, victim choice, offence type and age. The authors noted a number of limitations including the fact that: special needs was not synonymous with ID and as a result they were unable to verify intellectual differences between the mainstream and special needs populations; the treatments were not directly comparable because of adaptations and the assessments for the special needs populations had to be modified. At post-treatment, there were a few differences between groups but follow-up (post-release) data identified that none of the offenders in either group committed further sexual offences. Lindsay, Michie, Haut, Steptoe, and Moore (2011b) compared the treatment progress of 15 offenders against women and 15 offenders against children. They found that both groups improved significantly over a three-year period of treatment and that prog
ress was fairly even (linear trend) over the treatment period. There were no differences between groups in reoffending rates (around 23% for both groups).
CASE STUDY 21.2 SEXUAL OFFENDING
Andrew is 36-year-old man with mild learning disabilities (measured IQ of 66). He had been charged following an offence of lewd and libidinous behaviour with an eight-year-old girl. While babysitting, he had asked her to sit on his knee while they were watching television and had felt between her legs and outside clothing, and rubbed against her to ejaculation. The girl had commented to her mother the next day on what she thought was unusual and strange behaviour. She also mentioned that it had happened before when Andrew had been babysitting. Andrew was arrested and admitted the behaviour of the previous evening during police interview. He denied that it had ever happened before and police charged him with the incident in question. He received a three-year probation order with court ordered treatment. Treatment following the usual format and is outlined in detail elsewhere (Lindsey, 2009). It covered modules on disclosure of the relevant incident, understanding pathways into offending, challenging cognitive distortions related to offending, victim empathy, the use of pornography, developing relationships, and relapse prevention, and developing a future positive quality of life using the Good Lives Model (see Chapters 19 and 23). Measures were taken using the QACSO on two occasions prior to the commencement of treatment and regularly thereafter until treatment finished. Data from the QACSO are shown in Table 21.2.
Table 21.2 QACSO scores
QACSO scale
Pre 1
Pre 2
6 months
12 months
18 months