Forensic Psychology

by Graham M Davies

  19.3.2.6 Treatment Content of Sex Offender Programmes

  A survey of 1,379 sexual offender treatment programmes across North America (McGrath et al., 2010) revealed, somewhat surprisingly, that the majority of programmes focus on issues that have not been shown to have a strong relationship with recidivism, such as taking responsibility for offending and victim empathy. It is likely that the focus on these matters stems from influential earlier texts on sexual offender treatment (e.g. Salter, 1988), written before criminogenic need research became so well established. Mann, Hanson and Thornton (2010) and Thornton (2013) have attempted to encourage a change in treatment programme design by reviewing the risk factor literature to create lists of those risk factors (and hence those areas that should be targeted in treatment) with the greatest empirical support. Table 19.3 summarises the outcome of this review – further description of each risk factor can be found in the source paper.

  Table 19.3 Empirically-based risk factors for sexual recidivism

  Empirically supported risk factors Promising risk factors Unsupported but with interesting exceptions/Worth exploring Not risk factors

  Sexual preoccupation Hostility towards women Denial Depression

  Sexual preference for children Machiavellianism View of self as inadequate Poor victim empathy

  Sexualised violence Callousness Major mental illness Lack of motivation for treatment at treatment intake

  Multiple paraphilias Sexualised coping Loneliness Poor social skills

  Offence supportive attitudes Externalising Adversarial sexual attitudes

  Emotional congruence with children Fragile narcissism

  Lack of emotionally intimate relationships with adults Sexual entitlement

  Lifestyle impulsivity

  Self-regulation problems

  Poor problem-solving

  Resistance to rules

  Grievance thinking

  Negative social influences

  These empirically supported risk factors should be the main focus of treatment programmes. However, as McGrath et al. (2010) revealed, this is not always the case in practice. For example, deviant sexual interest is the risk factor with the strongest relationship with recidivism, yet only about two thirds of programmes in the US reported addressing this issue. McGrath et al. (2010) concluded that treatment targets of many sexual offending programmes “are often at odds with” the research into the factors that predict sexual recidivism.

  The discrepancy between practice and the evidence base probably exists because it takes some time for treatment programmes to change – for instance, manuals may have to be rewritten, staff may have to be retrained, and so on. In some cases, where research contradicts strongly held beliefs (such as the very widespread belief that offenders must take responsibility for their offending in order to reduce their risk), staff may actively resist change. Programme designers may therefore be likely to wait some time after publication of research findings before introducing major changes to treatment programmes.

  CASE STUDY 19.2 TREATMENT FOR A SEXUAL OFFENDER

  Joe is a 23-year-old offender who has been convicted of sexual activity with a female child under 16. He has no previous offences. When asked about the offence during an assessment, Joe describes having a “relationship” with the victim, who was 12 at the time of the offence. He says this relationship lasted three weeks. He claims he was in love with the victim. He reports his belief that his sexual activity with the victim was not an offence because she enjoyed it. He describes it as “just a little fun”. He also describes three other “relationships” with age appropriate females, lasting several days at the longest.

  Joe is very close to his parents and will move back in with them after his release. He worries that he will struggle to find a job because his previous experience has been working with young people and he’s not allowed to do that now. He spends most of his spare time with people who are generally 13–21 years of age because he shares similar interests with them such as skateboarding and video games. He reports that he feels people his own age are intimidating. Joe has an older sister who is supportive of him. She has two children. Joe’s sister believes that he is innocent and occasionally asks Joe to babysit her kids in a pinch.

  TREATMENT TARGETS/CRIMINOGENIC NEEDS

  Offence supportive attitudes (e.g. “It was not an offence because she enjoyed it”); emotional congruence with children (e.g. having “relationships” with kids, having interests similar to children, and finding adults intimidating); and lack of emotionally intimate relationships with adults.

  RELAPSE PREVENTION APPROACH

  Treatment would involve developing a relapse prevention plan, which would include determining potential risky situations Joe might encounter and how to avoid these. For instance, Joe would likely identify that socialising with children would put him at an increased risk of reoffending and therefore would need to develop methods to avoid being in such social situations, such as perhaps leaving a gathering if someone underage joined them. He would also develop a number of potential responses for how to cope with the situation of his sister asking him to babysit for her, even if she is desperate for help.

  GOOD LIVES APPROACH

  Treatment would involve identifying which “goods” Joe was trying to meet through his offending. It is likely that Joe was attempting to meet the “good” of (intimate and romantic) friendship. A GLM approach would assist Joe to develop the skills needed to develop the confidence to pursue intimate relationships with age-appropriate partners.

  19.4 THE EVIDENCE BASE FOR THE TREATMENT OF DANGEROUS OFFENDERS

  19.4.1 Violent Offenders

  There is a surprising lack of empirical evidence from which to draw conclusions as to the effectiveness of violent offender treatment. This probably reflects the fact that most jurisdictions have focused their resources on the treatment of other offenders – most notably, sexual offenders (Howells, Watt, Hall, & Baldwin, 1997; Polaschek, 2006). That said, most criminal justice systems recognise the importance of providing treatment to these serious violent offenders. Therefore, they either provide general criminogenic programmes (as outlined above), or have more recently developed specific intensive treatment programmes for this group (Serin, Gobeil, & Preston, 2008). This explains why there have been comparatively few attempts to thoroughly evaluate specific violent offender treatment programmes.

  In the first extensive review of violent offender treatment, Polaschek and Collie (2004) summarised the outcomes of nine studies that they considered to be of sufficient methodological rigour to warrant inclusion. Two of these were cognitive skills programmes, three were anger management programmes, and the remaining three were classed as multi-modal programmes. Each of these studies reported promising outcomes. However, Polaschek and Collie (2004) considered all of these studies to have methodological weaknesses, or a lack of information, which prevented any firm conclusions as to the effectiveness of violent offender treatment being drawn. More recently, Jolliffe and Farrington (2007) systematically reviewed the effectiveness of violent offender programmes and could find only 11 outcome studies that met their methodological criteria for the identification of good treatment programmes.

  Since the initial Polaschek and Collie (2004) review, there have been a number of evaluations of multi-modal (intensive) violent offender programmes. These have also produced inconsistent results. Polaschek (2011) reported on the New Zealand prison-based intensive Violence Prevention Unit programme. This is an intensive group-based programme that ran for four sessions per week over a 28-week period (approximately 330 hours). Polaschek (2011) matched 112 medium and high-risk violent offenders who completed the program with 112 untreated offenders over a period averaged 3.5 years. Polaschek found that 12% fewer offenders from the treated sample had reoffended and those treated offenders who did reoffend took twice as long to commit a further offence than the matched controls.

  Cortoni, Nunes, and Latendresse (2006) compared 500 violent offenders who complet
ed the 94-session prison-based Violence Prevention Programme (VPP) in Canada with 466 matched untreated controls. They found that offenders who completed the VPP had significantly fewer major institutional misconduct charges in the six-month and one-year period following completion of the programme. More importantly, untreated offenders were more than twice as likely to be reconvicted for a violent offence over the 12-month period.

  Serin, Gobeil, and Preston (2009) evaluated the Canadian Persistently Violent Offender program with less positive results. They found that violent offenders who had completed this 144-hour programme were as likely to reoffend as offenders who completed an AM programme or no programme at all. Similarly there were little differences between offenders with respect to institutional misconducts or measures of treatment change.

  Results regarding the efficacy of anger management programmes have also produced mixed results. Dowden, Blanchette, and Serin (1999) reported an 86% reduction in violent reoffending for 110 anger management programme participants over a three-year follow up. Research has also demonstrated that anger management programmes can reduce rule violations, re-arrest rates, angry patient behaviours as observed by staff, disciplinary incidents, physical assaults in hospitals, and verbal and physical aggression incidents (Novaco, 2013). In contrast, anger management programmes evaluated in Australia appear to have produced only small effects (Howells et al., 2002). Of note, these programmes appear to have been shorter, and less intense, than those reported by Dowden et al. (1999).

  Evaluations of cognitive skills programmes have also produced mixed results. In a meta-analysis of 16 studies across four countries, Tong and Farrington (2006) reported a significantly positive effect of cognitive skills programmes. In England and Wales, a large-scale study of cognitive skills programmes (e.g. Reasoning and Rehabilitation, Enhanced Thinking Skills) reported significant reductions in reoffending after a two-year follow up period (Friendship, Blud, Erikson, Travers, & Thornton, 2003). In a large scale Canadian study, Robinson (1995) reported reductions in recidivism of up to 36%. Offenders with a variety of convictions completed these 36-session prison-based Reasoning and Rehabilitation cognitive skills programmes; here, violent offenders were more likely to benefit from the programme compared to offenders convicted of theft offences. A similarly large evaluation in England and Wales (Falshaw, Friendship, Travers, & Nugent, 2004) found no differences between the two-year recidivism rates of offenders who completed cognitive skills programmes and a matched control group.

  Babcock, Green, and Robie (2004) conducted a large meta-analysis of IPV programmes based on 22 studies. They concluded that IPV programmes had, at best, a small positive impact on reoffending, but for the most part these programmes were not effective.

  19.4.2 Treatment Effectiveness of Sex Offender Therapy

  The effectiveness of sex offender treatment has been studied and reviewed extensively (Dennis, Khan, Ferriter, Huband, Powney, & Duggan, 2012; Gallagher, Wilson, Hirschfield, Coggeshall, & MacKenzie, 1999; Hall, 1995; Hanson et al., 2002; Kenworthy, Adams, Bilby, Brooks-Gordon, & Fenton, 2004; Långström, Enebrink, Laurén, Lindblom, Werkö, & Hanson, 2013; Rice & Harris, 2003). Numerous factors should be considered when determining the effectiveness of treatment (e.g. Harkins & Beech, 2007; Levenson & Prescott, 2014). These include the type of treatment (e.g. insight oriented, CBT), the study methodology (e.g. incidental cohort, randomised control trial) and how effectiveness is measured (e.g. recidivism, change within treatment).

  A useful method for evaluating various treatment approaches has been through the use of meta-analysis. This combines results from a number of studies to determine if there is an overall effect. It allows for small effect sizes to be detected in the large sample sizes that typically result from amalgamating studies.

  Hanson et al. (2002) conducted a meta-analysis examining treatment evaluation studies identified prior to May 2000. The studies analysed all had a comparison group, including those who had received no treatment, as well as those who attended programmes that were determined to be inadequate or inappropriate. This search yielded 43 studies (N = 9,534) from 23 published and 20 unpublished community and institutional treatment programmes, with an average length of follow-up time being 46 months. Hanson et al. reported a significant effect of treatment (12.3% for treated vs. 16.8% for untreated samples). Averaged across all types of treatment there was a significant effect of treatment. Breaking down treatment, by type of approach, Hanson et al. found that “older treatment” options (i.e. non-behavioural/non-CBT) appeared to have little effect in reducing and CBT had a positive treatment effect.

  Lösel and Schmucker (2005) reported similar results, analysing 69 studies (N = 22,181) that were completed prior to June 2003. This meta-analysis also identified a positive effect of treatment with treated sexual offenders. They found that physical treatments (i.e. surgical castration and hormone treatments) had larger effects than psychosocial approaches. Both CBT and classical behaviour therapy were also shown to have a significant impact on sexual recidivism. In contrast, more psychotherapeutic approaches (i.e. insight oriented, therapeutic community, and other unclear psychosocial approaches) did not significantly influence recidivism.

  Beech, Freemantle, Power & Fisher (2015) examined 54 treatment studies (N = 14,694), which included a range of different designs all using a control group. Results indicated a positive effect of treatment for both sexual and general recidivism, with an advantage of systemic and CBT approaches, in reducing both sexual and general recidivism. Beech et al. suggest that these results lend support for the efficacy of sexual offender treatment, particularly when the strongest treatment designs (i.e. randomised control trials and incident cohort combined) are used, with systemic therapy and CBT appearing to hold the most promise for effective interventions. Two recent systematic reviews have reached conclusions less supportive of the effectiveness of treatment. Långström and colleagues (2013) examined 167 treatment outcome studies with adult or adolescent offenders and child victims. They included eight of sufficient quality in their review. They concluded that the evidence is insufficient to determine if CBT with RP is effective for reducing sexual recidivism. In terms of adolescent offenders, there was evidence from one study (Borduin, Schaeffer, & Heiblum, 2009) suggesting that multisystemic therapy could be effective in preventing sexual reoffending but otherwise they concluded that there was no evidence to draw conclusions about the effectiveness of other treatment approaches. Dennis et al. (2013) reviewed 10 relevant studies (N=944 offenders) and concluded that further randomised control trials are needed because the current evidence does not support the position that treatment reduces risk of reoffending.

  Although findings about the effectiveness of treatment are mixed, a number of important considerations have been noted in terms of their impact on outcome studies (Levenson & Prescott, 2014). Randomized control trials are heralded by some as the “gold standard” for evaluating effectiveness, but others argue that it is unethical to withhold treatment from willing participants who may be more likely to reoffend without treatment (see Marshall and Marshall (2007) for a discussion of the problems with using RCT with sex offenders and Seto et al. (2008) for a rebuttal). Others have noted the importance of considering prosocial treatment change instead (e.g. Wakeling et al., 2013). The importance of considering process (as opposed to content) related variables (e.g. the therapeutic relationship instead of the specific material covered in treatment) to determine whether treatment has been effective, rather than a strict focus on recidivism as the only outcome of interest has also been noted (Levenson & Prescott, 2014).

  19.5 CONSIDERATIONS IN WORKING WITH DANGEROUS OFFENDERS

  There are number of considerations when working with dangerous offenders. We will now briefly consider some of these.

  19.5.1 The Psychopathic Offender

  Psychopathy is a condition marked by: (1) self-serving interpersonal traits (e.g. grandiosity), pathological lying, manipulativeness, shallow affect (e.g. lack of emotional
depth), lack of empathy, guilt or remorse; and (2) a set of broadly antisocial traits (e.g. impulsivity, persistent violation of social norms) (Hare, 2003). There are a number of studies and reviews discussing the commonly stated position that men who score high in psychopathy1 tend to respond poorly, as a group, to traditional treatment programmes (Hare, Clarke, Grann, & Thornton, 2000; Hare & Neumann, 2009; Hobson, Shine, & Roberts, 2000). Some studies have even seemed to indicate that treatment may make highly psychopathic men worse (i.e. more likely to recidivate; Hare, et al., 2000; Looman, Abracen, Serin, & Marquis, 2005; Rice, Harris, & Cormier, 1992; Seto & Barbaree, 1999). However, there are some problems with studies indicating negative treatment outcome for psychopaths and more current work suggests that psychopaths do not invariably have high recidivism rates (Abracen, Looman, Ferguson, Harkins, & Mailloux, 2010; Barbaree, 2005; Langton, Barbaree, Harkins, & Peacock, 2006).