|How to Treat Youths who have Committed Sexual Offenses|
Research into the effectiveness of different kinds of treatment for individuals who have committed sex offenses is surprisingly rare in recent years. Part of this is because it is hard to do. The low base rate of sexual recidivism makes it hard to have large enough samples to detect even relatively substantial treatment effects. And to interpret the recidivism rate observed for a treatment group you need some comparison group that they can be contrasted with, a group that either is not treated, or which is treated differently. This is hard to arrange. Withholding treatment from people who are assessed as presenting a “high risk” for sexual recidivism is hard to justify, but if you don’t do that, how do you ever find out whether the treatment actually helps. Ensuring the comparison group is genuinely comparable to the treatment group prior to treatment is not easy but if you don’t ensure comparability any result will be misleading.
Sometimes I get the impression that researchers have decided the topic is just too hard to study and moved on to doing something easier. The problem is that, in the absence of guidance from research, changes in treatment practice simply become a matter of what is fashionable. Following fashion can create the illusion of progress but an illusion is all it is.
While thinking such gloomy thoughts, I was pleasantly surprised to see a few recent papers that spoke to the effectiveness of different kinds of treatment. The most striking of these is by Aebi et al. (2022). It speaks to the effectiveness of different kinds of treatment for juvenile males who have sexually offended and is notable for both having a sound methodology and informative results.
Objectives of the Paper
The purpose of the paper was to compare the effectiveness of two different methods of treating juvenile males who have committed sexual offenses.
The first program, called Therapy Program for Adequate Sexual Behaviors - I (ThePaS-I), focused on understanding the individual’s sexual offense and developing offense-specific skills. These included accepting responsibility for the offense, developing the individual’s awareness of triggers, identifying and practicing internal and external behavioral controls, and developing a relapse prevention plan. This program is offense specific and draws heavily on traditional adult approaches to treating men who have committed sexual offenses.
The approach adopted by ThePaS-I is very consistent with what was originally advocated by Ryan & Lane (1997) when treatment for juveniles who had committed sex offenses was first being developed. More recently, Dopp et al. (2015) criticized this use of adult derived offense-specific treatment with juveniles. Current thinking has emphasized developmental issues common across adolescents with different kinds of problem behavior. Caldwell’s work (e.g., Caldwell, 2016) in particular has emphasized that juveniles who have committed a sexual offense rarely reoffend sexually and they are much more likely to engage in other kinds of criminal behavior. Caldwell states that for more recent releases the juvenile sexual recidivism rate is under 3%.
The second program, ThePaS-II, was designed to be consistent with this critique. It aimed to develop more general social, emotional, and psychological skills. It was adapted from the mindfulness training literature and social competence therapy for adolescents. Consequently, it taught many skills designed to help individuals achieve their goals in non-offending ways but did not require attention to the specific sexual offense the individual had committed.
Both programs were designed out of the University Hospital of Psychiatry Zurich and were of similar duration, involving about 19 or 20 modules, and about 30 sessions each lasting 60 minutes. They were both run by psychotherapists with at least master’s level training in psychology using treatment manuals and working under supervision. Importantly both treatments were administered by child and adolescent psychotherapists in a therapeutic, supportive, respectful, age-appropriate manner. Both programs could be run for individuals or for small groups of up to six participants.
Who participated in the study?
Participants were youth who had been referred by juvenile justice authorities to three forensic mental health institutions between 2011 and 2017. Inclusion criteria were that they had committed “crimes against sexual integrity” as defined by Swiss law (except for “pornography offenses”) and were aged 10 to 18 at the time of the first sex offense. Exclusion criteria were IQ below 70, insufficient knowledge of the German language, had engaged in severe non-sexual violence, had an acute psychotic disorder, a major depression, or a paraphilic disorder, or received previous treatment by ThePaS program.
Number of participants
75 youth were referred to treatment, 3 met exclusions criteria, 7 did not consent to participate in the study, 1 individual had to be later excluded because of a computer error, leaving a final sample of 64 participants. Mean age at time of the index offense was 14.75 and 15.15 at the time of pretreatment assessment.
How were comparable groups created?
All participants went through pre-treatment assessment and then were assigned between the two treatment groups using covariate adaptive randomization (see Suresh, 2011). This process uses randomization to assign cases. As assignment progresses future assignments take into account any differences between groups on pretreatment variables and makes the next assignment in a way designed to reduce these differences. This maximizes the similarity of the groups assigned to the two treatments.
How was treatment integrity assessed?
After each session was completed, it was coded by the therapist for fidelity (did the youth turn up and were the session contents consistent with the manual) and success (therapists’ impression of how well the youth understood and benefited from the session).
How was recidivism measured?
Criminal recidivism after treatment over an average follow up of about 4 years was derived from two sources. First, they looked at official records of reoffenses. Second, the researchers had access to juvenile justice case files which included reports from supervisors, institutions, and others who had post-treatment contact with the youths. These case files were reviewed to identify any behavior that could have been charged under Swiss law, regardless of whether it had been charged. Two categories of criminal recidivism were considered: sexual recidivism (excluding illegal pornography) and general recidivism. Importantly, the raters reviewing the files did not know which treatment program the youths had participated in.
Which other outcomes were assessed?
Internalizing and externalizing mental health problems were assessed using the Youth Self-Report questionnaire. Two scales from the Multiphasic Sex Inventory Adolescent Male form were used to assess sexual knowledge and beliefs as well as problems with sexual confidence. Beliefs about victims’ experiences were assessed through the Victim Empathy Questionnaire for Adolescents. These measures were taken prior to treatment, immediately after treatment, and 12 months after treatment.
The most important findings are as follows. Note that I have only stated that there was a difference in outcomes between groups when this difference was statistically significant. I have avoided going into the intricacies of statistical analysis, but, for example, the recidivism data were analyzed with Cox Regression which allows covariates to be controlled and allows for time at risk.
- Therapists’ ratings of fidelity and success were high for both programs but a little higher for the offense-specific program (ThePaS-I).
- Youths’ ratings of satisfaction with treatment success were generally positive but materially lower for the offense-specific program (ThePaS-I).
- Self-reported internalizing problems declined during both treatment programs, but self-reported externalizing problems declined only during ThePaS-II. Externalizing problems did not decline during the offense-specific treatment.
- About half of both groups had some form of criminal recidivism during the follow up period.
- Sexual recidivism was much lower after offense-specific treatment (ThePaS-I). Rates of sexual recidivism were 8.6% after ThePaS-I and 31.0% after ThePaS-II.
At least some groups of juveniles are more likely to sexually recidivate than might be supposed based on Caldwell’s publication (Caldwell, 2016). Sexual recidivism from official sources was 8% overall in the present sample, and when sexual offenses identified from case files were added in, it was 19%. It seems that Caldwell’s reliance on official charges may have underestimated about half of identifiable sexual reoffending. Nevertheless, the fact that these youth were referred by juvenile justice authorities for specialized treatment by a forensic mental health agency likely made them a “high-risk high-need” sample which would be expected to have a higher sexual recidivism rate.
The main practice implication of the study is that how you focus treatment effort determines which outcomes you affect. The offense specific program was better at reducing sexual recidivism but worse at reducing general externalizing behavior. The more general treatment program was better at reducing externalizing behavior but worse at reducing sexual recidivism. Critically, if you want to reduce sexual recidivism, your intervention should include work on the specific offense the person has committed and the development of attitudes, knowledge, and skills specifically relevant to avoiding offenses of that kind.
Aebi, M., Krause, C., Barra, S., Gunnar, V., Vertone, L., Manetsch, M., Imbach, D., Endrass, J., Rossegger, A., Schmeck, K., & Bessier, C. (2022). What kind of therapy works with juveniles who have sexually offended? A randomized-controlled trial of two versions of a specialized cognitive behavioral outpatient treatment program. Sexual Abuse. Online First: March 2022. https://doi.org/10.1177/10790632211070804
Caldwell, M. F. (2016). Quantifying the decline in juvenile sexual recidivism rates. Psychology, Public Policy, and Law, 22(4), 414. https://doi.org/10.1037/law0000094
Dopp, A. R., Borduin, C. M., & Brown, C. E. (2015). Evidence-based treatment for juvenile sexual offenders: review and recommendations. Journal of Aggression, Conflict and Peace Research, 7, 223-236.
Ryan, G. D., & Lane, S. L. (1997). Integrating theory and method. In G. D. Ryan, & S. L. Lane (Eds.), Juvenile Sexual Offending: Causes, Consequences, and Correction (pp. 267–321). Wiley.
Suresh, K. (2011). An overview of randomization techniques: An unbiased assessment of outcome in clinical research. Journal of Human Reproductive Sciences, 4(1), 8. https://doi.org/10.4103/0974-1208.82352