Efficacy of Sex Offender Treatment Still Up in the Air
High-quality studies lacking on whether treatment lowers risk of reoffending.
Posted Sep 25, 2013
"Did he complete treatment?"
But does formal sex offender treatment really lower risk?
A systematic review found no scientifically rigorous studies that establish a link between treatment completion and a reduced risk of reoffending among men who have sexually abused children.
This isn't altogether fresh news. We knew from earlier research reviews that:
- Any effect of treatment was modest, at best
- Treatment works best for the tiny minority of very high-risk offenders, while possibly aggravating risk for the broad majority of men at lower risk of recidivism
- Older offenders, due mainly to their very low risk, derive no tangible benefits from treatment
But considering both the prevalence and the harm of child sexual abuse, there is surprisingly little high-quality research on effective interventions. Partly, this is because of the lock-'em-up-and-throw-away-the-key mentality of policy makers. And partly it is because of the ethical difficulties in implementing random-design procedures, a hallmark of the scientific method, because men assigned to a control group would be denied treatment that could reduce their risk and in some cases shorten their prison terms.
Scouring research databases, a six-member, international research team was able to locate only three well-designed experimental studies. These included one with adults, one with adolescents and one with children. In only the study with adolescents was treatment shown to reduce recidivism. That project used multisystemic therapy, a very promising approach that integrates the family and larger community in the treatment.
Even broadening the search to include observational studies that lacked experimental designs, the research team found only five studies with a low enough risk of research bias to be deemed reliable. None of the five observational studies demonstrated that formal treatment—primarily cognitive behavioral therapy with relapse prevention—impacts sexual reoffending.
High-bias studies, in which the study design introduced a high probability of unreliable findings, were excluded. An example of such research bias would be a study in which treated and untreated offenders differed on a variable known to affect risk. When subjects are not randomly assigned to treatment or control groups, any observed differences between groups may be due to factors outside of the treatment itself.
Treatment in most formal sex offender programs is cognitive behavioral, and relies primarily on manual-based group therapy. For example, group exercises challenge distorted thinking, denial and minimization.
The research team found no minimally adequate studies whatsoever on the efficacy of pharmacological treatment with antiandrogen drugs, more popularly known as "chemical castration." They found this omission "particularly striking," in light of the prominence of this method in public debates.
Can treatment cause harm?
Under certain circumstances, with some people and some interventions, treatment could increase the risk of sexual reoffending. For instance, prolonged or intense interventions for offenders at low risk of relapse, or grouping low risk offenders with those at high risk for reoffending, could result in adverse outcomes."
They especially cautioned against unnecessary treatment of children. With recidivism risk very low among untreated children, treatment may lead to "unjustified stigmatization and could negatively affect the child’s development…If these children are subjected to excessively intense or inappropriate therapy, this could increase the risk for future antisocial behavior."
The team was headed up by prominent researcher and professor Niklas Långström and included Canadian researcher R. Karl Hanson, psychologist Pia Enebrink, forensic psychiatrist Eva-Marie Laurén and researchers Jonas Lindblom and Sophie Werkö. The research was commissioned and partially funded by the Swedish government.
The Convention on the Protection of Children against Sexual Exploitation and Sexual Abuse, ratified by 27 countries so far, mandates effective treatment to sexual abusers of children, individuals at higher risk of committing such offences, and children with sexual behavior problems.
This mandate is a bit of a problem, given the inconclusive evidence that the dominant treatment approach works.
Manualized, one-size-fits-all approach
"They want to hear that I always have fantasies and that I have more bad ones than good ones. But I don't have bad ones that often. I make up the bad ones. I make them really bad because they won’t leave me alone."
Genuine treatment, as we all should remember from our graduate school training, is all about the empathic relationship—not the technique. Indeed, although more and more psychologists have internalized the insurance industry's mantra that cognitive-behavioral therapy (CBT) is the "evidence-based" treatment of choice for a variety of conditions, this is not actually true. For example, in a new randomized clinical trial published in the American Journal of Psychiatry, psychodynamic therapy performed just as well in the treatment of depression.
The research team cautioned that their failure to find significant effects of treatment should not be interpreted to mean that treatment as currently implemented is ineffective. The low base rates of recidivism among sex offenders make it difficult to find treatment effects without very large sample sizes and long follow-up periods, they point out.
Additionally, an early study out of California provided some evidence that it was not the formal completion of treatment per se that reduced risk but, rather, the internalization of treatment messages and a desire to change -- something that is harder to measure.
The research team issued a call for large-scale, multinational randomized controlled trials. In the meantime, in the absence of solid proof that manualized cognitive-behavioral group therapy works as intended, they recommend a shift to more individualized assessment and treatment.
That's a solid, and very welcome, recommendation.