Treatment for perpetrators of sexual violence in close relationships: a systematic review
Systematic review
|Published
The objective of this systematic review was to examine the effect of treatments for persons who commit sexual violence in close relationships.
Key message
Sexual violence in close relationships is a serious social problem and we lack knowledge about the effect of treatment measures for perpetrators. The objective of this systematic review was to examine the effect of treatments for persons who commit sexual violence in close relationships.
We searched for randomized and non-randomized controlled trials in research databases. We screened the references, extracted data and analysed studies that met our inclusion criteria. We included three randomised trials and five non-randomized studies, which examined eight different interventions.
The main findings when it comes to new acts of sexual violence are:
- Multisystemic therapy may result in a slight reduction compared to cognitive behavioural group therapy.
- There is probably little to no difference between group cognitive behavioural therapy and mindfulness-based stress reduction
- There is probably little to no difference between pretrial diversion and no treatment.
- It is uncertain whether the Duluth model, specialist community-based treatment, Stichting Ambulante Prevention and group-based outpatient treatment have an effect compared to no treatment.
We assessed the certainty of the evidence for these results as low or very low. The current evidence base is too limited to determine the effect of treatments for perpetrators of sexual violence in close relationships.
Summary
Background
Many people experience sexual violence during their lifetime. The perpetrator is often a member of their family or someone else they know. The relationship between the victim and the perpetrator can be of great importance for how the violence is experienced and for how the support systems should intervene.
The Norwegian Centre for Violence and Traumatic Stress (NKVTS) has previously mapped out the treatment offers for adult perpetrators in Norway. They found that sexual violence is the form of violence that is treated by the fewest treatment units, and that treatment options for adults with specific problems related to sexual violence are not available in large parts of the country. Many of the treatment methods used today do not have a documented effect and there is a need for more knowledge about the effect of different treatments. The objective of this systematic review was to examine the effect of treatment measures for persons who commit sexual violence in close relationships.
Objective
What are the effects of treatments to prevent new acts of violence in persons who commit sexual violence in close relationships?
Method
We conducted a systematic review of primary studies. We conducted searches in relevant databases in December 2020. We included randomised controlled trials and non-randomised studies. Two researchers reviewed all titles and abstracts and relevant articles in full text and assessed the included studies for risk of bias. We described the results from each study separately, which also corresponded to unique treatments and comparison, as it was not possible preform a meta-analysis across studies and treatments. We assessed the certainty of the evidence for our primary outcome, sexual violence, using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) method.
Results
We included three randomised and five non-randomised trials. The randomised trials were from the United States and Norway and had a total of 656 participants (99.5% men). Two of the studies involved the treatment of adults and one the treatment of adolescents. The non-randomised trials were from the United States, Australia and the Netherlands and had a total of 712 participants (98.6% boys/men). Four studies involved the treatment of adult men and one child/adolescent of both sexes, aged 10-17 years. The participants in the studies had committed violence primarily against family members or acquaintances.
The randomized studies examined the effect of: The Duluth model compared to no treatment for men who had committed intimate partner violence (IPV), measured with the Conflict Tactics Scale (CTS2); multisystemic therapy compared to group-based cognitive behavioural therapy for adolescents with harmful sexual behaviour, measured with the Adolescent Sexual Behaviour Inventory (ASBI); and group cognitive behavioural therapy compared with mindfulness-based stress reduction for men who had committed IPV, measured with the CTS2. The non-randomised studies examined the effect of: pretrial diversion compared to no treatment for men who committed sexual violence against children within the family, measured with criminal records, specialist community-based treatment compared to no treatment for non-convicted adolescents with sexual harmful behaviour, measured with criminal records; group-based cognitive behavioural therapy compared to group-based supportive therapy for men who had committed IPV, measured with the CTS2; and the Stichting Ambulante Preventive Project compared to no treatment for men who had committed sexual violence, measured with criminal records; and group-based outpatient treatment compared to no treatment for men who had committed sexual violence, measured with criminal records.
The results for the primary outcome, sexual violence, and our certainty in the results are summarized below.
| Summary table of effects of treatments for persons who commit sexual violence in close relationships | ||||||
|
Population: perpetrators of sexual violence in close relationships. Setting: Norway (1), USA (2) |
||||||
|
Treatment/control
|
Outcome: sexual violence Number of participants (studies) |
Relative effect |
Anticipated absolute effects* (95% CI) |
Certainty |
||
|
Risk for control group |
Risk with treatment |
Difference |
||||
|
Group cognitive behavioural therapy/ mindfulness-based stress reduction |
Assessed with CTS2 sexual coercion scale |
RR 0.87 |
6.9% |
6.0% |
0.9% fewer |
⨁⨁◯◯ |
|
Multisystemic therapy/cognitive behavioural group therapy
|
Assessed with ASBI sexual risk/misuse scale |
RR 0.62 |
48.3% |
30.0% |
18.4% fewer |
⨁⨁◯◯ a,b,d |
|
Pretrial diversion/ no treatment
|
Assessed with criminal records n = 208 (1 non-randomised study) |
RR 0.73 |
10.8% |
7.9% |
2.9% fewer |
⨁⨁◯◯ a,d |
|
*The risk in the treatment group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). a. Downgraded due to imprecision (only one study with few participants) b. Downgraded for risk of bias due to a lack of explanation separate randomization. c. Downgraded for indirectness (treatment was compulsory, which would not happen in Norway) d. Downgraded for risk of bias (groups were dissimilar at baseline and selected based on different criteria) |
||||||
Our main findings show that when it comes to new acts of sexual violence, Multisystemic therapy may result in a slight reduction compared to cognitive behavioural group therapy. There is probably little to no difference between group cognitive behavioural therapy and mindfulness. There is probably little to no difference between pretrial diversion and no treatment. We have low confidence in these results.
Concerning the other treatments, it is uncertain whether the Duluth model, specialist community-based treatment, Stichting Ambulante Prevention Project and group-based outpatient treatment have effect on sexual violence. We assessed the certainty of the evidence for these results as very low.
Discussion
We identified few studies on the treatment of perpetrators of sexual violence in close relationships, and most of the studies we identified were over 10 years old. Only one of the eight included studies was conducted in Norway, and the remaining in the USA, the Netherlands and Australia. It is unclear how comparable treatments and contexts of the different countries are in terms of different demographics, legislations and treatment systems. In Norway, treatment of perpetrators of violence is mainly voluntary. Several of the included studies examined compulsory treatments and is uncertain to what extent such treatments (and their effects) are transferable to the Norwegian context.
The included studies have methodological limitations, and the reported results are very uncertain. This makes it difficult to draw conclusions about the effect of the interventions. Therefore, it is relevant to develop and conduct studies that can reduce the risk of biases, have enough participants, and have satisfactory reporting of methods and results. Standardised measurement methods should also be developed so that outcomes from several studies can be analysed in meta-analyses. This can improve our confidence in the results, and we can to a greater degree trust what the real effects of the treatments are. The results of this review can stimulate further research to optimize the existing or new treatments offered to perpetrators sexual violence in close relationships.
Conclusion
The current scientific evidence is too scarce to determine the effect of treatments for perpetrators of sexual violence in close relationships. We need more high-quality, adequately powered research studies, on effective treatment options for this group.
