From Research to Practice: Designing a Treatment Program for Individuals Convicted of Child Sexual Exploitation Material

The term evidence-based is increasingly found in treatment manuals and program titles designed for individuals convicted of a sexual offense. However, whether the presented evidence truly qualifies as “evidence-based” is questionable. I will share my experience as the clinical director for a private outpatient agency where we designed a program based on the existing peerreviewed literature on individuals convicted of Child Sexual Exploitation Material (CSEM) offenses. I will describe the steps that we followed in developing a workgroup to explore and apply the knowledge from research to our clinical practice. We adapted the term evidence-informed to acknowledge that the program relied on existing literature in combination with the experiences and expertise of our clinical team. A fictitious case that reflects an amalgamation of facts is presented to demonstrate the assessment and treatment processes. Implications for organizational consensus to conduct program evaluation and research are discussed.


Problematic Use of the Internet
Access to and regular use of the Internet have increased rapidly in recent years, with 5.11 billion unique mobile users in the world that grow daily 1 . Convenience, connectedness, affordability, creativity, education, entrepreneurship, and cultural exchanges are just a few of the common ways the Internet has changed every aspect of our lives. Similar to other technologies, however, the Internet can be misused, bringing harm to oneself and to others. There is growing concern that for some people, Internet use has become compulsive, excessively time consuming, costly, and ultimately disruptive to other important life domains. The Internet also has facilitated the proliferation of a new form of sex crime -the production, distribution, and consumption of child sexual exploitation materials (CSEM). Internet crimes against children are not limited to sexual abuse images, and other forms of abuse include online grooming of minors and sex tourism. Although these forms of abuse have been around for a long time, the rapid growth of the Internet has afforded unprecedented opportunity to produce, use, share, and sell these images. These opportunities are further advanced through its growing capacity to transmit information through websites, email, instant messaging, newsgroups, bulletin boards, peer-topeer file sharing networks, gaming, social networking sites, and other anonymized networks. Internet-related sex offenses are also facilitated through a concept that Cooper (1998) 2 introduced to the field as the Triple-A Engine: accessibility (24 hours a day, 7 days a week), affordability (free sexual material), and anonymity (perceived belief that online activities are anonymous).

Magnitude of the Problem
While the exact prevalence of child abuse images is unknown, estimates from arrest, victimization, and self-report findings suggest that viewing child abuse images is a serious problem and that the numbers are rapidly increasing over time. In 2019, the Internet Watch Foundation, a charity organization that works to remove child sexual abuse content on the Internet, processed about 132,000 online images containing child sexual abuse material -double the number identified online in 2016 3 . The National Center for Missing & Exploited Children 4 (NCMEC) CyberTipline -a centralized reporting system for the online exploitation of children -has received more than 65 million suspected reports of Internet child sexual exploitation since its inception in 1998. Through NCMEC, more than 18,900 of the victims depicted in abusive images have been identified by law enforcement. Takedown operations of websites that produce and distribute child sexual abuse images are reported regularly in media, with a recent one involving the arrest of 330 individuals for their involvement with a site that contained more than 250,000 unique child abuse videos 5 . In the United States, the number of men incarcerated for child sexual abuse images has nearly doubled since 2004, with most serving mandatory sentences of five years or more. In 2006, there were an estimated 3,672 arrests for crimes involving CSEM possession -more than twice the estimated 1,713 arrests in 2000 6 . There are also reports that the victims depicted in these images have become younger, and the images more brutal [7][8][9] .

Rehabilitation
In the United States, the criminal justice system is characterized as a swinging pendulum, alternating between punishment and rehabilitation eras 10 . One recent example is the shift from tough-on-crime toward smart-on-crime that resulted in early release and reentry programs. Today, correctional facilities in most developed countries recognize the value in rehabilitation and helping incarcerated individuals live a law-abiding life following their release from prison. The extensive work of Gendreau and colleagues 11,53 , and replicated across independent investigators 12,13 , is paramount in showing that punishment alone does not reduce crime, but rather that on average, rehabilitation does reduce recidivism, at least to a modest degree. Moreover, rehabilitative outcomes are enhanced with programs that adhere to Risk-Need-Responsivity (RNR) principles 14 . Psychotherapy and counseling are considered integral components of the rehabilitation process, and there are treatment programs designed for individuals who struggle with substance abuse, anger management, and domestic abuse, as well as those convicted of sexual offenses. Traditionally -and arguably still a dominant practice today -the treatment needs of individuals who commit Internet sexual offenses are addressed using standard, or contact ("hands-on"), sex offender treatment methods. With the increase in the number of Internet sexual offenses, however, questions have been raised as to whether individuals convicted of these offenses belong to a separate group of sexual offenders or if they are no different than those convicted of conventional sexual offenses, other than using technology to carry out their offending behavior 15 . Research has demonstrated that individuals convicted of CSEM offenses are different from those convicted of hands-on offenses in demographic, psychosocial characteristics, personality, and coping style 16,17 . Given these initial findings, a key question is how conventional sex offender treatment interventions apply to individuals convicted of CSEM?

Individuals Convicted of an Internet Sex Offense
Internet sexual offending comprises a range of crimes, including possession, distribution, and production of child sexual abuse images, sexual solicitation of minors, prostitution, and indecent exposure. However, the large majority -and arguably most concerning -involves child pornography. The heterogeneity of Internet sexual offending is complicated further by online, offline, and mixed offending modalities. Questions about the generalizability of the curriculum to address distinct and overlapping treatment needs arose in the early stages of the curriculum development. Our efforts to customize the program for individuals convicted of Internet sex offenses centered on maintaining adherence to RNR principles, as well as the emerging research demonstrating psychological difference among those who have committed Internet sex offenses 18 . Research thus far has identified three types of sexual offending against children with unique and overlapping characteristics: individuals convicted of CSEM (online) offenses, individuals convicted of typical (offline) sex offenses, and individuals convicted of offenses with both CSEM and contact (mixed) offenses. Individuals convicted of child pornography-only offenses are different from individuals convicted of mixed offenses and offline offenses against children, and their risk level can be best characterized as an intermediate between these two groups -whereas mixed offenders may benefit from conventional treatment models, CSEM offenders may have different criminogenic needs and responsivity styles.

Demographic and Clinical Characteristics
Differences between individuals convicted of Internet (online) vs. typical contact (offline) sexual offenses have been found in terms of demographic and developmental characteristics, as well as offending behaviors. In general, those convicted of Internet sexual offenses have been found to be overwhelmingly young, Caucasian males [19][20][21][22] . Recent studies have continued to find that individuals convicted of Internet sex offenses are significantly younger than their contact or mixed offending counterparts 20,[23][24][25] . Additionally, individuals convicted of Internet sex offenses are significantly more likely to report their relationship status as single -as opposed to married, divorced, or separated -than individuals convicted of contact sex offenses 20,[23][24][25] . Multiple studies report that individuals convicted of Internet sex offenses generally have stable employment 19,22 , with Aslan and Edelmann (2014) 23 reporting they are significantly more likely to have fulltime employment than individuals convicted of contact sex offenses. Similarly, those convicted of Internet offenses are significantly more likely to have completed a higher level of education than those convicted of contact offensessuch as college or postgraduate educations compared to a high school diploma 23,26 . In comparison to those convicted of contact sex offending, individuals convicted of Internet offenses have also been found less likely to have endured childhood sexual abuse 23,24 , have lower rates of substance abuse 20,22 , and fewer convictions for prior offenses 20,24,26 .
In comparison to mixed and contact offense groups, Internet offense groups have also been found statistically more likely to offend against mixed gender victims, though this may simply be a result of the breadth of content available on the Internet 23,24 . A meta-analysis by Babchishin and colleagues (2015) 16 found the available literature suggests individuals convicted of Internet offenses are less likely to have access to children, though more likely to have access to the Internet, and are more likely to have sexual interest in children when compared to individuals convicted of contact offenses -consistent with the idea that motivated individuals will offend in accordance with the opportunities available to them. Additionally, those convicted of Internet offenses are less likely to be antisocial, tend to have greater victim empathy, and have fewer cognitive distortions than contact or mixed offense groups 16,[24][25][26] . Some literature has also suggested that those convicted of Internet offenses are more likely to report homosexual or bisexual orientations than contact or mixed offense groups 16 .

Evidence-Informed Treatment Curriculum
The differences in the demographic and clinical characteristics of the offending pathways have raised questions on the unique needs of these groups. The extent that those convicted of Internet sexual offenses can benefit from programs that were meant for offline offenders is unknown. I will describe our experience of implementing an evidence-informed treatment curriculum for individuals convicted of Internet sexual offenses.

Method
As a first step, our agency developed a workgroup consisting of clinical supervisors who were assigned to operationalize the necessary components of an evidenceinformed program. Discussion sessions were held on key concepts and theories, barriers and challenges, and most importantly, implementation strategies. The second step entailed conducting a search of peer-reviewed studies using combinations of the following keywords: "Internet Sex*," "Child Abuse Image*," "Child Sexual Exploitation*," "CSEM," "Child Porn*," "Indecent Images of Children," AND "Treatment*," "Intervention*," "Rehabilitation*," "Sex Offender Treatment*." Studies were identified using the following electronic databases: PsycINFO, Medline, Pubmed, and Web of Knowledge. The contents of each article were reviewed and indexed based on their relevance to the assessment and treatment of individuals convicted of Internet sexual offenses. Books, particularly Internet Sex Offenders 27 , were used as well. We studied available material from the Internet Sex Offender Treatment Program 28 and the FCI Butner Sex Offender Treatment Program (https:// www.youtube.com/watch?v=P8FmrdtDo8I).
We used information from the following websites: Center for Sex Offender Management (CSOM; https://cepp.com/centerfor-sex-offender-management-csom/) and online PROTECT (https://onlineprotect.org.uk/). We also collaborated and consulted extensively with Professor Derek Perkins and Dr. Hannah Merdian, the authors of online PROTECT, as well as with Dr. Eke and Dr. Seto. Finally, we attended Research and Treatment Conferences sponsored by the Association for the Treatment of Sexual Abusers (ATSA) and sought out paper, symposium, and poster sessions that were relevant to Internet sexual offending.

Operationalizing Key Concepts and Theories
Evidence-informed: Nowadays the term 'evidencebased' has become a popular term that is used broadly, but without specificity on the level of evidence. For example, a prison rehabilitation program called Insight Garden Program, includes the following statement on its website: "we are an evidence-based program, with a less than 10% recidivism rate (rate of return to prison) which results in the healing of lives, families, and communities." 29 . The program is likely to have therapeutic benefits, including increase in engagement and motivation to change, but where the evidence is derived for this program is unclear. The 'gold standard' to determine a program's effectiveness is a double-blind, randomized, controlled trial where an outcome from one treatment is compared to an outcome from a controlled treatment. Although double-blind studies are nearly impossible in psychotherapy outcomes, randomized non-blinded controlled trials are the preferred method for verification of evidence-based interventions.
We adapted the term evidence-informed to signify that our program relied on the existing literature in combination with the experiences and expertise of our clinical team. This is consistent with the literature that argues that "evidence-based" and "evidence-informed" do not denote the same thing and should not be used interchangeably 30,31 . We aimed to make clear that workgroup members and organization administrators understand that evidence-based implies data in effectiveness -or in this context -reduction in recidivism. To the best of our knowledge, no program has yet shown or examined what works in reducing recidivism among Internet sexual offenders. Therefore, treatment programs designed for Internet sexual offenders can be best characterized as research-informed.
Child Pornography, Indecent, or Abuse Images, and Child Sexual Exploitation Material: There are concerns and objections in the field on the use of the term "child pornography" 32 . Children cannot consent to sexual activity, and the content of the material is considered equivalent to crime scene photos or films of children being sexually abused or exploited. As such, indecent or child abuse images would more accurately reflect what these images portray. However, given that Internet offenses against children are not limited to images or films, and may include stories, solicitation, and we favored using child sexual exploitation material (CSEM), which encompasses a wider group of offenses. It is worth acknowledging CSEM is also not an all-inclusive term, and leaves open questions on situations where an individual may use the Internet to read about sex tourism and travel to a particular country to sexually abuse children. However, given that the majority of clients in our program were convicted of child abuse images, the use of CSEM was considered a good fit. We limited the use of the term child pornography to contexts where it would be awkward or confusing to avoid -for example, referencing articles where the term was used originally, names of assessment tools, or items within a risk assessment tool.

Risk Assessment:
A central issue in program development was choosing a risk assessment tool for estimating the likelihood of committing a sexual offense in the future, as well as for guiding treatment decisions. There are multiple limitations in the availability of suitable risk assessment tools for individuals convicted of Internet sex offenses, including considerations for the specificity of Internet offenses, short follow-up periods, and small sample size.
There are two commonly used actuarial risk assessment tools for individuals convicted of Internet sex offenses: The Risk Matrix-2000 33 (RM2000) and the Child Pornography Offender Risk Tool 34 (CPORT). The RM2000 is a statistically-derived risk assessment tool intended for males aged at least 18 years who have been convicted of a sexual offense 33 . The tool was developed in the United Kingdom with males aged 18 years and older who were convicted of a sexual offense after the age of 16. An Internetonly offense group was not included in the original sample, but has been present in subsequent samples 35,36 . The international version scoring manual provides guidance on how to use RM2000 with Internet-only offense groups. Briefly, static information is used to classify individuals into four different risk bands for sexual (RM2000/Sexual) or other violent offenses (RM2000/Violence), which is representative of that group's likely rate for reconviction of future sexual offenses. The RM2000/Sexual scale consists of seven items: age, sexual offending history, criminal (nonsexual) offending history, sexual offense against a male victim, sexual offense against a stranger, having a twoyear live-in romantic relationship, and non-contact sexual offense history. The scoring algorithm for the international version is a two-step process.
Step 1 involves summing the first three items and translating the score into one of the following risk categories: below average, average, above average, or well-above average. Step 2 involves summing the last four items, and then using that score to shift the initial risk category one or two levels upward. If two or three of these items are present, the initial risk category is raised one level (e.g., from below average risk to average risk). If all four of these aggravating factors are present, the initial risk category is raised by two risk levels (e.g., from average risk to well-above average risk).
Barnett and colleagues (2010) 35 examined the predictive validity of a modified version of the RM2000 scales and reported moderate (AUC = .7) predictive accuracy (albeit large confidence interval: .43 to .97) for all subgroups of sexual offending, including individuals convicted of Internet offenses (making or possessing indecent images of children). These findings, however, are not consistent across all studies. Wakeling and colleagues (2011) 36 , for example, reported poor predictive validity (AUC = .5) for sexual reconviction among those convicted of Internet offenses only, and moderate accuracy for those convicted of mixed offenses. Overall, findings indicate that modified versions of RM2000 are relevant for better understanding recidivism among individuals convicted of Internet sexual offenses, but a tool designed specifically for this population has yet to be developed. 34 is a more recently developed structured risk assessment tool used to assess the likelihood of future offending in men convicted of CSEM possession. The scoring manual and updated reference material for the CPORT are available online through Research Gate: https://www.researchgate. net/project/Child-Pornography-Offender-Risk-Tool-CPORT. The CPORT consists of seven items: age at the start of the police investigation, prior criminal history, prior or concurrent contact sexual offending, prior or concurrent failure on conditional release, evidence of pedophilic or hebephilic sexual interests, more boy than girl child pornography content, more boy than girl other child (e.g., nude) content. The Correlates of Admission of Sexual Interest in Children (CASIC) 37 , a separate questionnaire consisting of six items, can be used to supplement scoring for evidence of pedophilic or hebephilic sexual interests. CPORT items are summed for a total score that can range from zero to seven. Combining results from the development and validation samples, the CPORT has demonstrated good predictive accuracy for five-year sexual recidivism and online recidivism 34 . At present, there is insufficient research to assign risk level for CPORT scores, and interpretation of scores is limited to percentile ranking. Recidivism probability estimates from logistic regression for five-year time frames may also be used, but the stability of these are unknown.

The Child Pornography Offender Risk Tool (CPORT)
While static risk assessment tools estimate risk level and likelihood of recidivism, dynamic risk factors identify treatment targets that are amenable to change. Comprehensive risk assessment involves combining both static and dynamic targets in a structured, or empirical manner. There are currently no validated risk assessment instruments specific to the treatment needs of individuals convicted of Internet sex offenses, and whether the available tools address the needs of all sexual offending groups remains an open question. The Stable-2007 38 and the Sex Offender Treatment Intervention & Progress Scale (SOTIPS) 39 are two of the commonly-adopted dynamic risk assessment tools for individuals convicted of Internet sexual offenses.
The Stable-2007 38 consists of 13 items (significant prosocial influences, negative emotionality, social isolation, intimacy deficits, impulsivity, problem-solving skills, lack of concern, hostility towards women, sexual self-regulation, sexual deviance, sexual preoccupation, emotional identification with children, and cooperation with supervision) organized under five sections: social influences, intimacy deficits, sexual self-regulation, general self-regulation, and cooperation with supervision. Total scores on the STABLE-2007 can range from 0 to 26 for individuals with child victims and 0 to 24 for individuals without child victims. The total score can be classified into three risk categories: low (0 -3), moderate (4 -11), and high (12). However, the Stable-2007 manual uses the following precautionary language for use with individuals convicted of child pornography-only offenses: a) the tool may be used as a clinical guide to identifying treatment needs and supervision targets, b) the tool should not be used to estimate recidivism rates or to assign nominal risk categories, and c) users should explicitly acknowledge that the tool is not validated for child pornography-only offense groups. A recent meta-analysis that included a sample with a high number of CSEM-exclusive male subjects found the STABLE-2007 to have good predictive accuracy for a new child pornography offense 40 . These preliminary findings are promising to further investigate the generalizability of the Stable-2007 items to individuals convicted of Internet sex offenses.
The second commonly-used risk assessment tools for CSEM clients -SOTIPS -consists of 16 items (sexual offense responsibility, sexual behavior, sexual attitudes, sexual interests, sexual risk management, criminal and rule breaking behavior, criminal and rule breaking attitudes, stage of change, cooperation with treatment, cooperation with community supervision, emotional management, problem solving, impulsivity, employment, residence, and social influences) that are inclusive of sexual and general criminal offending. The items are organized under six sections: sexuality and risk responsibility, criminality, treatment and supervision cooperation, self-management, social stability, and support. The SOTIPS manual rules on target population is similar to the Static-99R 41 , where there is a distinction between category A and category B sex offenses, with possession of child pornography falling under the latter and not meeting the scoring criteria. The authors have adopted a precautionary use of the test for CSEM clients that is consistent with the Stable-2007 40 .
Another tool that might be relevant, especially in differentiating CSEM clients with and without nonsexual criminal histories, is the Level of Service/Case Management Inventory (LS/CMI) 42 . The LS/CMI consists of 43 questions grouped into eight criminogenic needs sections: criminal history, employment/education, family/marital, leisure/ recreation, companions, attitudes, alcohol/drugs, and antisocial patterns. While no research has directly examined the utility of SOTIPS or LS/CMI for identifying treatment needs in clients with CSEM-exclusive offense histories, the validation studies included a wide range of offending groups, including those with CSEM offenses.
Finally, the Kent Internet Risk Assessment Tool-2 43 , although not commonly used in forensic evaluations, is used to assist law enforcement agencies with case prioritization and workload investigations. Reportedly, it aims to identify, individuals who represent the highest risk of contact offending by prioritizing those who share features with mix offenders. The items are organized into four evaluation steps: previous behaviors, access to children, current behavioral facilitators, and other factors.
To the best of our knowledge, a risk assessment tool that could be used to identify online offenders who will re-offend by committing a contact sexual offense is not yet available.
In our clinical setting, we have implemented the use of the CPORT along with the Stable-2007 40 or SOTIPS during the intake process. The decision process in selecting these tools was not as systematic as it could have been. In hindsight, using principles from organizational decision-making could have saved us a lot of time and positively affected the perceived credibility of clinicians toward the organizational leadership. We went back-and-forth on our decisions, eliciting criticism for being indecisive. Nevertheless, the final selection of the instruments, was made according to the peer-reviewed literature, availability of the test developers to consult and offer training opportunities, and consensus within the organization clinical leadership. It is worth pointing out that access to the literature, especially for nonacademic organizations, is equally important as the published literature itself. For example, the CPORT/CASIC Project Page on ResearchGate facilitated networking with others, finding training handouts and other resources, accessing articles, and having updates about published studies using the CPORT.
One challenge that our organization encountered is its lack of risk categories, therefore making it difficult for clinicians to report risk in a conventional format. Although the scoring guide includes information on the distribution of scores -specifically, where raw scores may be in terms of percentiles -this is not always as intuitive to interpret for some of our colleagues in the criminal justice system Consequently, the RM2000 36 is used whenever a clinician must render a decision about risk level. LS/CMI is used as clinically needed, typically when the client has a history of nonsexual offenses, and/or when the identification of antisocial behavior, attitude, and patterns are relevant for case formulation. Clinicians and colleagues at large have provided favorable feedback in the use of these measures to determine risk and identify targets for intervention.

Theoretical Framework
Our treatment framework was based on the work of Merdian et al. (2018) 44 , which through the use of interviews and thematic analyses, had identified seven subordinate themes affecting an individual's pathway to CSEM offending. Brief descriptions of these seven pathways are described in Table 1. These themes were structured into seven treatment modules that included a clinical guide, client handout, and sample goals and objectives. The program also included integration of etiological models that considered offending behavior as a consequence of motivational and facilitative factors [45][46][47][48] 51 Good Lives Model, were adopted as well.

Pathway Description
The Developmental Context This pathway refers to key learning experiences from developmental years that could be considered distal factors contributing to eventual offending behavior. This pathway can be broken down into two subthemes: lack of connection with others and a vulnerability to socio-emotional breakdown -such as poor familial relationships -as well as the role of sexual needs within and separate from intimate relationships -which may include childhood sexual abuse or early exposure to pornography.

Individual Propensities and Psychological Vulnerabilities
This pathway refers to the main motivational and facilitative psychological vulnerabilities that have been found to be directly linked to CSEM offending behavior. These may include negative internal statessuch as continued reinforcement of low self-esteem -reliance on dysfunctional coping skills, and/or an underlying sexual interest in children.

The Internet Environment and Behavior
This pathway refers to the role of the online environment as a situational facilitator of the offending behavior. Two subthemes predicate this, the first being the nature of the Internet -in terms of accessibility and anonymity -and the second being the psychological meaning of the offending behavior -such as the act of searching for this material, as well as chatting/sharing with other CSEM-involved individuals.

Personal Circumstances
This pathway refers to an individual's circumstances directly proceeding their initial CSEM offending behavior. This is often a lack of social connection or sexual outlet, the breakdown of a relationship, general feelings of failure, or another significant life stressor.

Permission-Giving Thoughts
This pathway refers to cognitions that lower inhibitions toward the CSEM offending behavior and are perceived as reinforcers for continuing to offend. Specific thoughts are often that the offending behavior is not harmful, that it is simply a victimless crime, and/or that the content does not depict real children.

The Consequences for the Individual
This pathway represents experiences directly following the offending behavior, which is in response to negative perceptions of the individual's personal circumstances and the inability to cope. This can be divided into two subthemes, the first described as a negative reinforcer -such as viewing this behavior as an escape from everyday life -and the second as a positive reinforcer -such as viewing this behavior as a means of sexual gratification and connection.

Desistance
This pathway supports the idea that individuals may have engaged in attempts to self-manage their offending behavior prior to arrest. This falls into two subthemes: increased social engagement -such as elevated self-confidence or ability to engage socially -and confidential help and education -such as relabeling the content as child abuse images instead of child pornography. integrated offense-related vulnerabilities and personal circumstances within the Internet environment. It also provided the flexibility for structuring the program into a preexisting setting (e.g., group hours, number of sessions, individual and group meetings), instead of the other way around.

Case Study
A fictitious case is presented to demonstrate the intricacies of the treatment program where an assessment of risk, criminogenic need, and responsivity factors are examined in the framework of the pathways model. Mr. Musa contacted the postal inspector via email after seeing his message posted through Yahoo! Groups which stated that the undercover inspector had some "action photos" of subjects in their early teens and younger. The postal inspector provided graphic descriptions of available videos, and Musa responded with four image files of prepubescent males engaged in sexual conduct and a request for one of the tapes from the postal inspector's list depicting 12-year-old baseball players and cheerleaders engaged in sexual conduct. Mr. Musa also offered to trade "a DVD for a DVD." Over the following days, Mr. Musa continued to email the undercover postal inspector child sexual exploitation images, particularly prepubescent male content, and also let him know that he had sent the postal inspector a DVD in the mail with a two-hour reel of sexually explicit photos and short video clips. Once the postal inspector received the DVD one week later, it was reviewed and discovered that the DVD contained approximately 700 images and 40 videos. Out of these, 85 images and 15 videos were determined to constitute child sexual exploitation material. These files depicted primarily prepubescent males, some as young as five years of age, engaged in sexual conduct with other prepubescent males or adult males, including acts of fondling and anal penetration.
Two weeks later, the postal inspector reinitiated contact with Mr. Musa, and they exchanged emails for one month before Mr. Musa indicated he was once again interested in trading child pornography, and that "LimeWire" and "Bearshare" were good sources. One week later, a warrant was obtained authorizing the search of Mr. Musa's residence. The postal inspector alerted Mr. Musa that he would be receiving a DVD in the mail. When the package arrived at his home and he signed the delivery receipt, postal inspectors appeared and began their warrant search. Mr. Musa admitted to officers that he began viewing child sexual exploitation content in approximately 1999, and that he considered himself as "being someone who is addicted to looking at these images" and that he found them to be "a thrill to watch." He denied engaging in any sexual contact with a minor. Mr. Musa was ultimately held responsible for being in possession of 600+ images of child sexual exploitation material.
Prior to the index offense, Mr. Musa had no juvenile or adult court adjudications. He currently reports that his index offense was a misunderstanding, and that he was not aware of the illegality of child sexual exploitation materials at the time he was viewing it. He believed that since it could be so easily accessed online, especially back in the early years of the Internet, that it must be regarded like any other adult pornography. He restates that it began in 1999 when he was searching for adult pornography images. After visiting a website, he claims to have received an email with a link to an additional pornography site. Out of curiosity, he viewed the site and was surprised to find images of children. He believed the children in the images were willing participants, and that viewing the content was a victimless action. He adamantly asserts that had he not received the link to this website, and were it not so easily accessible online, he would otherwise never have viewed this kind of material. He admits to having fantasized and masturbated while viewing the child abuse images.
Social History: Mr. Musa was raised by his two parents and has five siblings. He states he was raised in a positive family environment, and that he always shared a close relationship with his parents and siblings. Both of his parents maintained steady employment, and their financial situation was consistently stable. Mr. Musa denies experiencing any mental or physical abuse during his childhood.
Mr. Musa reports a normal academic progression through elementary, middle, and high school with no difficulties, never having received any special education services. In fact, he reports he always felt ahead of his peers, often receiving top grades without having to put in considerable effort. Because of this, he felt that he never had any close or real friendships during his school years and described all of his social relationships as mere acquaintances. Mr. Musa graduated from high school and subsequently attended some community college, having enrolled in photography and health classes out of personal interest. However, he chose not to pursue any kind of degree.
Between the ages of eight and 11, Mr. Musa reports having sexual "connections" with a male cousin in his twenties. Mr. Musa never spoke about this experience until the index offense, and by that time the cousin had passed away. At the time, he did not perceive these acts as sexual abuse or exploitation, and does not recall feelings of shame, fear, or distrust of adults. Mr. Musa now understands the abusive nature of these acts, but he neither feels resentment toward his cousin, nor thinks of himself as a victim of sexual abuse. Mr. Musa does not believe this experience caused him to seek child sexual exploitation images, but he could see how it may have affected his deviant attitude towards this content, that viewing these images was harmless.
Mr. Musa reports his first consensual sexual interaction was at the age of 18, and the first time he had sexual intercourse was at the age of 21 with a girlfriend of approximately six months. Mr. Musa has been in brief relationships with women and men of his own age, never lasting more than 12 months. He identifies his sexual orientation as same sex, and explains that his previous romantic experiences with women were obligations to please his family. Although he enjoys strong ties with his parents and siblings, he continues to maintain his sexual orientation as a secret. He believes disclosure of his sexuality would be another disappointment to his family and feels he has already put them through enough upset. Mr. Musa reports that he has never been married, has fathered no children, and is presently single.
Mr. Musa reports maintaining steady employment since attending community college, and prior to his arrest, he spent 17 years working for a division of a popular chain grocery store. Although a senior employee, Mr. Musa reports dissatisfaction with his salary. He felt he was not receiving the payment he deserved at that point in his career and noted that he did not receive any kind of raise or promotion during the last three years of his employment. Mr. Musa was terminated from this job due to the index offense.
Mr. Musa reports his first interactions with alcohol at age 17, however, he did not enjoy the feeling of being "out of control" and only continued to consume it in order to ease anxiety in social situations. He reports consuming one-to-four alcohol drinks once or twice monthly. He admits to using marijuana about once a year, which began approximately five years prior to his index arrest. Mr. Musa has no other drug use history.
Mr. Musa reports that his lack of substantial friendships in childhood caused him a lot of loneliness and anxiety as he has entered his adult life. However, he never attended any mental health treatment for this. No other mental illnesses run in his family, to his knowledge. In terms of physical health, he currently suffers from Type II diabetes, acid reflux, and chronic pain due to repetitive movements that are in the beginning stages of arthritis.

Risk Factors The Child Pornography Offender Risk Tool (CPORT):
The CPORT is a seven-item structured tool to assess the likelihood of future sexual offending over a 5-year fixed follow-up. The items include: age 35 or younger at index offense, any prior criminal history, any prior or index contact sexual offense history, any conditional release failure, admission of pedophilic or hebephilic sexual interests, greater content featuring boys as compared to girls in child pornography viewed/possessed, and greater content featuring boys as compared to girls in other/nude child materials viewed/possessed. Mr. Musa presents with 3 out of 7 risk indicators: admission of pedophilic interests, and more boy-focused content in both child pornography material, as well as nude/other child material.
For a person with this score and no contact offense history, the observed five-year recidivism rate is 12.2% for any sexual offense and for any child pornography offense. In the combined normative dataset from Eke et al. (2019) 52 , roughly 15% of individuals had the same CPORT score, 69% had a lower score, and 16% had a higher risk score.
In other words, out of 100 individuals involved with child pornography material, 69 would be expected to have a lower score, 15 would have the same score, and 16 would have a higher score.

RM2000.
Mr. Musa's risk for sexual recidivism was also assessed using RM2000, a static actuarial risk assessment tool. It is worth pointing out that the reason for proctoring RM2000 was to derive a risk level. The RM2000 items include: age, sexual offending history, criminal (nonsexual) offending history, sexual offense against a male victim, sexual offense against a stranger, single, and non-contact sexual offense history. Mr. Musa's history places him in the Average category on the Sexual-Scale (Revised), where he obtained a total score of 2 for sexual offense against a male victim and single status.

STABLE-2007.
The STABLE-2007 measures empirical risk factors that are routinely addressed as part of correctional rehabilitation (i.e., criminogenic needs) for adult males convicted of sexual offenses against a child or non-consenting adult. The STABLE-2007 consists of 13 items related to psychological, interpersonal, and sexual functioning, that are added together to create a total score. No current research has examined the extent to which the STABLE-2007 provides accurate assessment of criminogenic needs among men whose only sexual convictions involve possession of indecent images of children. The STABLE-2007 was used as a clinical guide for identifying treatment needs and supervision targets.
The following need areas have been identified as being of SIGNIFICANT CLINICAL CONCERN for Mr. Musa: Capacity for Relationship Stability, General Social Rejection, and Deviant Sexual Preferences. The following need areas have been identified as being of SOME CLINICAL CONCERN: Sex Drive/Preoccupation and Sex as Coping.
Level of Service/Case Management Inventory (LS/ CMI). The LS/CMI is an instrument that measures the risk and need factors of convicted adolescents and adults. It is used to aid in the treatment planning and management of individuals in correctional related agencies. Mr. Musa obtained a LS/CMI Section 1 TOTAL Score of 5. His risk/ need level is low. The following areas were assessed as having the most concerning level of risk and needs and should be primarily targeted in treatment and supervision: currently unemployed and absence of recent participation in an organized activity.

Protective Factors
Mr. Musa has the benefit of some protective factors, including a positive and stable relationship with his parents and siblings, who have all remained supportive of him throughout the offense. Despite some reliance on alcohol in social situations, there are no historical indications of alcohol or drug dependence. Additionally, Mr. Musa demonstrates satisfactory compliance and understanding of his supervision conditions and the Courtmandated treatment requirements. Mr. Musa is currently unemployed, but he has a stable work history and is actively seeking employment. He considers work and financial independence as an important aspect of his identity.

Risk Level
In the absence of research to combine static and dynamic risk factors to derive a single risk level, reliance on RM2000 along with his problems relative to sexual deviance and poor interpersonal competence have been used to assess his overall risk level. Mr. Musa appears to be at an average risk of reoffending. He has few identifiable criminogenic needs and is expected to respond positively to services. Eighteen months of individual psychotherapy is recommended: weekly sessions for the first 12 months, and a decrease to twice a month for the last six months. This would be equivalent to approximately 60 hours of therapy.

Responsivity Factors
Mr. Musa finds the therapeutic process helpful in facilitating self-awareness and making changes. His motivation to reflect, present, and provide and take feedback is high. Mr. Musa's beliefs about human sexuality are deeply rooted in cultural norms and expectations. He may also be sensitive to a therapist's cultural awareness and his/her ability to address issues without assumptions and judgments.

Case Formulation
Developmental Context: Mr. Musa reports growing up in a secure environment with responsive and warm caregivers. Despite a positive family environment, from a young age, Mr. Musa has felt inadequate in his social abilities and experiences difficulties in forming lasting relationships. The treatment goal for this theme would be for Mr. Musa to identify the strengths and challenges of attachment experiences, and their impact on his adult relationships. Treatment objectives may include: a) processing childhood feelings and memories associated with his primary caregiver, b) processing his desire and ability to trust others, as well as demonstrating himself as trustworthy, c) reflecting on how strong ties with his parents may generalize to feelings of security in adulthood, d) exploring the importance and value placed on relatedness, and e) identifying the activities involved in achieving relatedness.

Individual Propensities/Psychological Vulnerabilities:
Mr. Musa's motivation for online offending could be hypothesized as stemming from sexually deviant fantasies and interests. It is worth noting that he was engaged in the viewing and possession of these images for over six years, a substantial devotion of time, energy, and interest. Feelings of loneliness and anxiety, as well as the inability to establish emotional and sexual relationships with ageappropriate adults, may have facilitated these offending behaviors. Sexual preoccupation and his own sexual victimization as a child are also likely to have increased the risk of offending. The treatment goal for this theme would be for Mr. Musa to develop an understanding of how individual propensities and psychological vulnerabilities increase one's risk of offending, especially when combined with dysfunctional coping skills. Treatment objectives may include: a) participation in the assessment of sexual interests using psychophysiological measures, b) exploration of childhood experiences that may trigger and maintain his sexual interest in children, c) reflection on the consequences of his sexual interest in children, d) consultation with a psychiatrist and exploration of the use of medication to reduce sex drive, and e) identification of ways in which his sexual interest in children may have been related to seeking common life goals.
Internet Environment: Accessibility, affordability and anonymity, commonly known as the "Triple-A Engine," of the Internet permitted the acceleration of Mr. Musa's online offending behavior. He recalls he relied more and more on the Internet throughout his index offense for interpersonal needs and the regulation of negative emotions associated with his loneliness. He reports frequently turning to the Internet to escape from the pressure and discomfort of his life at the time. Additionally, his perception of anonymity through the Internet contributed to risky explorations, such as seeking younger images. He felt more emboldened to act on his natural interests and dispositions. The treatment goal for this theme would be for Mr. Musa to develop the knowledge and skills necessary to manage his online behavior. Treatment objectives may include: a) sharing thoughts and feelings on using the Internet to escape from his negative feelings, b) reflecting on the role that anonymity, accessibility, and affordability played in his online interactions, c) eliciting feedback on whether his online behavior fits the self that is presented in the group, d) building an empathic response in recognizing children in the images as victims of child abuse, and e) developing and implementing strategies to block or avoid high risk sites.
Personal Circumstances: At the time of the index offense, Mr. Musa reports experiencing persistent feelings of inner emptiness and dissatisfaction with his employment situation. One particular life event that occurred close to the time of the offense was a loss in a jobrelated promotional opportunity. He states that he viewed himself as a failure with a dead-end job and no prospects for a career. This experience was generalized to other aspects of his life, especially interpersonal needs, where he felt that he was incapable of developing and maintaining intimate relationships. He resorted to dealing with these emotions primarily through the viewing of pornography and masturbation, which ultimately amplified his risk for seeking illegal images. The treatment goal for this theme would be to raise his awareness on how life event(s) are influenced by schemas. Treatment objectives may include: a) reflecting on what other significant things were going on in his life, b) identifying and processing of self-defeating statements that may recur throughout his life, c) building alternative beliefs replace maladaptive schemas, and d) developing a visual picture of possible healthy coping responses.

Permission-Giving Thoughts:
Mr. Musa's reported thinking patterns that facilitated his offending behavior reflect denial and disbelief in his role of child abuse. He remembered telling himself that viewing the images caused no harm to the children, that there was no direct physical contact and the children in these images did not reveal any distress or indication of sexual abuse, and even that the children in the photos were smiling and appeared to be enjoying themselves. The treatment goal for this theme would be to explore how these thoughts might have aided to justify his offending behavior. Treatment objectives may include: a) developing a list of his permission-giving thoughts, b), exploring the origins of his permission-giving thoughts, c) reframing these permission-giving thoughts through alternative explanation, and d) seeking input from group members to shift absolute thinking patterns into something more balanced.
The Consequences for the Individual: Mr. Musa's reaction toward his offending behavior seems to have changed from a positive mood and feelings of excitement, to overwhelming dread and the exacerbation of his loneliness. He said the best way that he could describe the process was a loss of control. The treatment goal for this theme would be to learn how to use his strengths in pursuit of prosocial pleasurable activities. Treatment objectives may include: a) completing the Values in Action inventory of strengths, b) developing a list of activities to use his signature strengths to their fullest potentials, c) using his signature strengths in novel ways, and d) developing a tolerance for a certain degree of negative mood and boredom.
Desistance: Mr. Musa has a positive bond with treatment providers, and as he likes to say, the "collective wisdom" of the group. He feels that reflections and feedback from other members have provided him with the tools to understand and regulate his own thoughts, emotions, and behaviors. The treatment goal for this theme would include further developing desistance strategies. Treatment objectives may include: a) preparing to tell his life story with an emphasis on his crime-free identity, b) developing a list of potential barriers to desistance and his strategies to overcome these barriers, c) exploration of prosocial activities or programs to develop his sense of belonging, and d) commitment to developing a good life that is free of sexual offending.

Staff Training and Development
Staff training and continuous professional development both play a key role in the implementation of new programs. Our first strategy was to ensure there were manuals and materials available for the new program. We developed clinical guidelines, client handouts, group activities and exercises, as well as sample treatment goals and objectives that corresponded to each offending pathway. These materials were organized and made available through an internal SharePoint site. Trainings were provided both face-to-face and through livestream, as well as prerecorded webinars. Experts in the field were invited to train staff in the area of risk and treatment of online sexual offending. These trainings were received well and followed by enthusiastic question-and-answer sessions. Staff rated case examples and practical information as the most pragmatic features of these trainings. Common feedback from clinicians included a desire for "less theory and more practical information." We also provided regular internal trainings. Anecdotal feedback from staff seemed to suggest that one-to-one interactions with supervisors who were informed and excited about the new program were critical in their implementation efforts. Although we had assessments in place to measure staff members' pre-andpost knowledge, skills, and attitudes, regretfully we were unable to follow through with a systematic evaluation.
Nevertheless, efforts are being made to continue to evaluate training outcomes with new employees.

Research and Program Evaluation
A major challenge in implementing a new program involves attaining organizational consensus, as well as a commitment to research and program evaluation. Three research areas that would be necessary to advance the program from research-informed to evidence-based treatment are: fidelity evaluation, client feedback, and recidivism outcomes. Fidelity measures that incorporate the clinician's own expertise and style, while still adhering to the core theory and principles of the program, is essential to evaluate the effectiveness of the program. Client feedback gauged through surveys or other processes could inform program administrators and clinicians about clients' perceptions and experiences, especially what they may have found most or least beneficial. Finally, there is a need for outcome research using a Randomized Controlled Trial (RCT), otherwise, we will never achieve consensus on 'what works' in the rehabilitation of individuals convicted of Internet sexual offenses, or arguably, whether low risk clients need any treatment.

Conclusions
I hope our experience in the development and implementation of this program will encourage others to explore new ways of integrating research into their practice. The experts with whom we have had the privilege of consulting and collaborating have inspired us to do our best and always question the merits of our decisions. We continue to have dialogues with program administrators and clinical justice partners in support of research and program evaluation. Recently, a doctoral student in Clinical Psychology from Royal Holloway, University of London, attained approval to conduct a program evaluation to explore whether individuals who participate in our program consider it helpful, and if so, in what ways. We encourage our colleagues to promote evaluation of their treatment programs and take steps toward developing an evidence-based program for online offending.