Article Text
PDF
Statistics from Altmetric.com
Question
Question: Are psychological or pharmacological interventions effective for preventing child sexual abuse in known abusers and individuals at risk of abusing?
Outcomes: Conviction on charges of sexual offences against children (including possession of child pornography), arrest by police on suspicion of the same offences, breach of conditions while serving a sentence for sexual offending and self-reported child sexual abuse.
Methods
Design: Systematic review.
Data sources: PubMed (NLM), PsycInfo (EBSCO), National Criminal Justice Reference Service Abstracts (EBSCO), Cochrane Library and the Campbell Library were searched up to June 2013, supplemented by manual search of reference lists, books and websites.
Study selection and analysis: Randomised controlled trials (RCT) and prospective observation studies (cohorts or case controls) of adult or adolescent perpetrators or potential perpetrators of child sexual abuse, or of children with sexual behaviour problems. Psychologists and psychiatrists with expertise in sexual abuse and offending reviewed eligible studies for inclusion, and evidence was combined using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. Eligible studies included at least 20 participants, at least 1 year of follow-up, included groups balanced for potential confounders and studied any pharmacological, psychological or psychoeducational intervention (excluding outdated treatment approaches). Control groups included treatment as usual or no treatment. Only studies with low or moderate risk of bias were included. Effect sizes (risk ratios) were calculated.
Main results
Eight studies were eligible, including one RCT and four observational studies in convicted adults with low to high risk of reoffending; one RCT and one observational study in convicted adolescents with moderate risk of reoffending; and one RCT in children with sexual behavioural problems. Sample sizes ranged from 48 to 484, and study follow-up ranged from 3–5 to 16 years. The majority of studies included manual-based group therapy based on cognitive behavioural therapy (CBT) principles. The five studies in convicted adults did not provide sufficient evidence that the intervention had any effect on the risk of reoffending. The one RCT in 484 moderate-risk adults in a prison setting found no effect of CBT compared with no treatment over 5–14 years of follow-up (RR 1.10, 95% CI 0.78 to 1.56). The one RCT in 48 convicted adolescents with moderate risk of reoffending found weak evidence that multisystemic therapy reduced the risk of reoffending compared with usual community services over 9 years (RR 0.18, 95% CI 0.04 to 0.73). The observational study in adolescents found non-significant effects. The one RCT in 135 children with child-directed sexual behavioural problems found that CBT had no effect on risk of sexual offences compared with group play combined with parental support over 10 years (RR 0.16, 95% CI 0.02 to 1.25). No studies were identified in adults considered at risk of child sexual offences but with no previous convictions. No studies of pharmacological interventions (eg, anti-androgens) were of sufficient quality for inclusion.
Conclusions
There is a lack of available evidence to inform whether psychological or pharmacological interventions are effective at preventing child sexual abuse in individuals at risk of abusing. High-quality study is needed to inform realistic clinical strategies to reduce risk factors in those at risk of reoffending.
Abstracted from
Commentary
Långström and colleagues present a meta-analysis of the research on psychological and medical intervention effectiveness for preventing reoffending by child sexual abusers. This systematic review breaks no new ground and reviews a subset of studies from recent meta-analyses.1 As with previous studies, the authors conclude that they ‘identified remarkably little research of acceptable quality’ (p.3). This conclusion is similar to a 1989 narrative review,2 that concluded ‘there is as yet no evidence that clinical treatment reduces rates of sex reoffenses in general and no appropriate data for assessing whether it may be differentially effective for different types of offenders’ (p.27). Clearly, the message to take from the recent meta-analyses, including this one, is that there is a desperate need for rigorous research on sex offender treatment effects, especially in adult populations. Clinicians working with adult sexual offenders are in much the same place where they were in 1989. As Karl Hanson suggested in his comment on the Dennis et al1 meta-analysis, they should look to the general offender literature and probably should also be guided by the literature on psychotherapy, considering the applicability of Positive Psychology interventions, which have shown robust effects across problem areas.3
In Långström et al meta-analysis the findings with respect to adolescent treatment are probably more convincing than that for adults, although only two studies were included in this review, one random clinical trial. The practitioner working with adolescents can be guided by the work on multisystemic therapy to use interventions that are ecological, focus on parental training and address the social systems in which youngsters reside.
Footnotes
-
Sources of funding The Swedish Government, Department of Social Affairs.
Footnotes
-
Competing interests None.