Study Selection

The aforementioned databases were screened with the use of the search terms leading to an identification of 1049 records. After removing 282 duplicates and identifying 12 articles through further methods, 779 articles were screened for abstract and title. This led to an inclusion of 18 articles into the database of this systematic review. An overview of the article selection process is provided in Fig. 1. Moreover, seven further web-based perpetration prevention initiatives without peer-reviewed published data were identified, which will be mentioned cursorily.

Fig. 1

PRISMA 2020 flow chart of included articles

Study Characteristics

Included studies are six quantitative randomized control trials (RCT; 33.33%) [31••, 32•, 33•, 34, 35••, 36•], three qualitative studies (16.66%) [37,38,39], two experimental-longitudinal studies (11.11%) [40, 41], two single-group intervention studies (11.11%) [27, 42], one between-participant experimental study (5.55%) [43], one quantitative non-randomized study (5.55%) [44], one mixed method study (5.55%) [45], one intervention development study (5.55%) [46], and one RCT study protocol (5.55%) [47]. Attrition rates varied from low [36•] to high [31••, 35••] across studies. Most studies used a mainly Caucasian sample; only one study was conducted in Vietnam [36•], one in Brazil [46], and one with marginalized youths [45]. Moreover, in 14 of the articles included, the authors reported demographics on gender. In 10 of the 14 studies, over 90% of the sample was male [27, 31••, 32•, 33•, 34, 35••, 36•, 39, 43, 44]. The sample characteristics and the study design as well as the main findings can be found in Table 1. As this systematic review aims for a comprehensive picture of all web-based prevention efforts taken, including both the prevention of initial sexual offense perpetration and also of recidivism, it discusses initiatives with somewhat different target groups. Among those articles, four (22.22%) cover web-based perpetration prevention initiatives for individuals who have sexually offended [31••, 32•, 44, 47]. These programs vary from online interventions based on cognitive behavioral therapy (CBT) to emotion regulation interventions. Three of those interventions target individuals who have sexually molested a child [31••, 44, 47], whereas one targets men with a sexual offense perpetration history, not necessarily with regard to children [32•]. The 14 remaining articles (77.77%) discuss perpetration prevention programs for non-offending individuals who are sexually attracted to children [37, 38], for college students [33•, 34, 35••, 36•, 40,41,42,43, 45, 46], for youths [43, 45], or for individuals fulfilling the criteria for hypersexuality disorder [27]. It is worth noting that two perpetration prevention programs for college students (RealConsent and KisS) are respectively discussed in two articles [33•, 35••, 40, 41].

Table 1 Design, sample characteristics, program information, and main findings of included studiesWeb-Based Initiatives for Individuals Who Have Sexually Offended—Target Group, Content, and Efficacy

As mentioned above, four studies discuss web-based perpetration prevention initiatives for individuals who have sexually offended. For only two of the four initiatives, an effectiveness evaluation is published. Lätth et al. [31••] conducted an RCT (single-blinded participants, parallel-group, superiority, psychological placebo-controlled) with active CSAM users (18 +) who took part in an anonymous online intervention called Prevent It, developed by the Karolinska Institutet in Stockholm. Notably, this RCT (pilot) was the first of its kind in this field. Prevent It is a CBT-based program in English consisting of eight modules related to the use of CSAM. The primary outcome was the self-reported time spent watching CSAM within the last week, assessed by the SChiMRA + measure [48]. Compared to the placebo control group, participants in the Prevent It group showed a significant decrease in CSAM viewing time. The program was well-appreciated by participants who reported hope and a heightened awareness of their thoughts and behaviors.

Troubled Desire is an anonymous app-based perpetration prevention program for individuals who are sexually attracted to children [44]. The app consists of self-assessment and self-management training modules and is available in German, English, Spanish, Portuguese, French, Hindi, and Marathi. There is no published data on its effectiveness. Descriptive analysis of the self-assessment data revealed that 78.9% of the sample had a sexual interest in children. It is however unclear how this percentage came about, as the target groups of Troubled Desire are individuals with a sexual interest in children. Individuals who offended and who reported sexual interest in children were more likely to experience distress as well as report CSAM use vs. individuals who sexually offended with no interest in children.

The target group of a third web-based intervention, called @myTabu, are individuals (18 + ; males and females) under community supervision for CSA or CSAM use. The program is currently being evaluated, and so far, only the study protocol for the RCT is available, while the results are still pending [47]. The program consists of six predefined modules that target diverse dynamic risk factors with coach-guided feedback. Both conditions (intervention vs. placebo) include 24 sessions to be completed in 24 weeks. The program is based on the risk-need-responsivity (RNR) approach [49] and CBT techniques. A monetary reward system is included to motivate participants. To assess efficacy of this intervention, recidivism (recorded re-offenses) and self-report questionnaires on empirically supported dynamic risk factors will be assessed for both groups.

While the previously described initiatives (Prevent It, Troubled Desire, and @myTabu) target individuals who are sexually attracted to children, Davis et al. [32•] tested a program addressing men with a history of sexual perpetration against adults who reported at least one incident of sexual coercion in their lifetime. The study used an RCT design to test how alcohol intoxication influences two different five-to-10-min emotion regulation interventions (cognitive restructuring (CR) and mindfulness) that aim to reduce sexual aggression compared to an attentional control condition. Participants were blinded to all conditions and were told that they would practice coping strategies for negative emotions. Results showed that the CR intervention as well as the mindfulness intervention resulted in less non-consensual sex intentions compared to control. This effect was particularly pronounced for the CR group. Notably, for intoxicated men with a more severe perpetration history, the mindfulness condition led to higher coercive tactic intentions compared to individuals in the control and CR condition.

Web-Based Initiatives for Non-offending Individuals—Target Group, Content, and Efficacy

These next studies included in this review target primary perpetration prevention. As previously mentioned, two prevention programs—RealConsent and KisS (competence in sexual situations)—are respectively discussed by two articles [33•, 35••, 40, 41]. RealConsent targets perpetration prevention and is an intervention based on social cognitive theory, social norms theory, and the bystander educational model. The intervention focuses on contents such as sexual consent, rape myth, norms regarding gender roles, effective communication, alcohol and rape, victim empathy, and bystander behavior. It comprises six modules which last 30 min each and can be completed within three weeks. A third article discusses a further expansion (from three to 12 weeks) and cultural adaptation of RealConsent, called GlobalConsent [36•]. Both programs helped to reduce SV perpetration [33•, 35••, 36•]. The effect of RealConsent seems to be mediated by four factors [33•]: knowledge of effective consent for sex, hostility towards women, date rape attitudes, and hypergender male ideology. Even though it was reported in both studies that the effect sustained over a follow-up period, Yount et al. [36•] summed the results of posttests 1 (after six months) and 2 (after 12 months), making it impossible to determine the actual effect at the respective time points. Besides this, there was no posttest immediately after the intervention. KisS aims to reduce risk and vulnerability factors of sexual aggression perpetration and victimization. Based on sexual scripts and social learning theory, it is designed as a six-week prevention program with weekly modules. Compared to the control group, risky sexual behavior was lower in the intervention group nine and 12 months after the intervention, indicating a long-lasting impact. The impact on risky sexual behavior at nine and 12 months was mediated via less risky sexual scripts one week after the end of the program [40]. Tomaszewska et al. [41] additionally reported indirect effects of the intervention on sexual aggression perpetration: the impact at nine and 12 months was mediated via less risky sexual scripts one week after the program and risky sexual behavior nine months after the intervention [41].

Further web-based initiatives, Dating SOS, introduced by Murta et al. [46], and DOT (Dream, Own & Tell), discussed by Segura et al. [45], aim at preventing dating violence (DV) in youths. Dating SOS is a guided web-based program for Brazilian youths consisting of individual sessions where participants learn about different aspects linked to romantic relationships, such as victimization and perpetration of DV. Subsequently, they receive personalized feedback via text message. An effectiveness and usability evaluation (RCT) of the program is ongoing. Project DOT consists of three core components, an “Educational and Leadership Curriculum” in group sessions on Zoom, development of a social media campaign, and a community mobilization practicum [45]. Pre-COVID-19, the program was administered in-person but was adapted to an online format due to the pandemic. This online version was found to be valuable and was perceived as a safe space by participants. In a between-subject experimental study with youths, Nicolla et al. [43] used existing personal narrative videos on TikTok (a social media platform for viewing and sharing short videoclips) discussing SV to prevent SV perpetration through conveying personal experience. Viewing led to more information on consequences of SV and higher perceived severity of SV compared to control.

In an RCT, Wong et al. [34] tested a 45-min brief intervention aimed at reducing sexual aggression perpetration of heterosexual male college students. The study revealed that participants in the self-persuasion intervention condition, which addressed sexual aggression, reported decreases in self-perceived likelihood of committing sexual aggression after the intervention. However, this effect was not maintained until the six-month follow-up assessment.

In a pilot study, Gilmore et al. [42] tested the + Change initiative, another brief intervention. The program targets alcohol misuse, sexual assault perpetration, bystander behavior, and sexual assault risk reduction. On average, participants took 17.52 min to complete the program. The findings revealed that after the program, participants were significantly more conscious of the problem of sexual perpetration and showed greater willingness to make changes to prevent sexual assault perpetration.

In the last study discussing a primary prevention initiative—a pilot study by Hallberg et al. [27]—participants were not college students or youths, but individuals who fulfill the criteria for hypersexuality disorder (HD). HD is associated with recurrent sexual thoughts and behaviors, which can be a risk factor for recidivism in sexual offenses [39]. A quarter of participants fulfilled at least the criteria for one paraphilic disorder or reported paraphilic thoughts. In this uncontrolled study, participants underwent an intervention to reduce hypersexual symptoms. The intervention consisted of 10 modules distributed over a 12-week period. The intervention led to a significant pre-post decrease in hypersexual symptoms which was sustained three months after completion.

Peer-to-Peer Support Forums for Non-offending Individuals Who Are Sexually Attracted to Children

Among the 18 included articles, three are qualitative studies that cover topics discussed by non-offending individuals who are sexually attracted to children in the peer support forum Virtuous Pedophiles [37, 38] and a Dutch peer support forum [39]. Virtuous Pedophiles has more than 4700 members and offers help for those with a sexual interest in children who report that they are living offense-free. The forum has strict rules about topics of discussion and can be monitored by the Police. Jones et al. [38] identified three predominant themes of discussion by applying thematic analysis to 326 forum posts: accepting pedohebephilic interest, coping strategies in risk situations, and substitutions to satisfy sexual needs. In an ethnographic study about the same online forum, Pedersen [37] concludes that peer-to-peer support seems to be a relevant resource for a fulfilling life by reducing fear to seek help. He states that non-offending pedophiles should not define themselves as monsters and individuals are not responsible for their thoughts and are able to control their actions. Thus, one might assume an increased willingness to seek help from professionals and thereby prevent sexual offenses. The Dutch forum is an online community of individuals with a sexual interest in children. It consists of three chat rooms and had over 600 members in the last five years. Bekkers et al. [39] analyzed 15 interviews with members, moderators, and mental health professionals involved in the forum and identified that peer support can lead to increased self-acceptance and coping with stigma as well as less loneliness and psychological distress of members. The authors conclude that online communities can have a positive impact on behavioral change and well-being.

Additional Perpetration Prevention Initiatives

At present, diverse web-based programs are being developed and evaluated for individuals at risk of committing CSA. Among those that are already implemented but still under evaluation are @myTabu, Troubled Desire, and Project Bridge [50], an adapted version of Prevent It [51] as well as HelpWanted [52]. Project Bridge consists of the five-week CBT-based self-help program ReDirection as well as motivational interviewing (Mi Bridge), which both aim to deliver strategies to stop CSAM use. It is available in English, Finnish, Spanish, Czech, Slovak, and German. Prevent It [51] has been revised since the first version and now consists of nine modules and targets individuals with sexual urges towards children, regardless of whether they have offended in the past or not. The program is available in German, Swedish, and Portuguese as well as in English. HelpWanted [52] is a five-module online course for adolescents and young adults who are sexually attracted to younger children, aiming to support participants in living a safe and non-offending life. Further available resources comprise the self-help website of StopItNow! UK & Ireland [53] (including Get Support [54] and Get Help [55]) from the Lucy Faithful Foundation, which offers educational material as well as a live chat for individuals worried about their sexual thoughts about children or their online behavior. StopItNow! offers another program which specifically targets juveniles, called What’s OK [56], which provides information on sexually safe behaviors and a live chat.