Therapists’ Experiences in Their Work With Sex Offenders and People With Pedophilia: A Literature Review

This article presents a review of the literature that pertains to the experiences of therapists who work directly with child sex offenders and/or people with pedophilia. We draw together results from studies that attempted to identify how therapists experience such work and how they were personally impacted by it. Usually, such studies are embedded within one of the following theoretical frameworks: Secondary traumatic stress, compassion fatigue, vicarious traumatization and burnout. Most literature on the topic has therefore sought to determine to what extent and why, work-related stress responses may occur among these therapists. The aim of this paper is therefore to provide insight into this, arguably, important line of research, while evaluating the current knowledge as well as providing recommendations for future research efforts.

negatively impacted by their work could be more likely to participate in such studies, as they may find them useful and relevant.
Based on their interpretive phenomenological analysis, Friedrich and Leiper (2006) also concluded that sex offender treatment providers experience a considerable amount of negative feelings, due to their work. In 2012, Sandhu, Rose, Rostill-Brookes, and Thrift also conducted a study using interpretive phenomenological analysis, in which similar results were obtained. It is, however, unclear to what extent positive experiences occurred as well. Although the authors stated that the participants were asked about positive aspects of work, no information hereof was included in their results.
Scheela's interview-study from 2001 generated similar results. According to her analysis via the constant comparison method, the participants felt like it was exciting to work in what was described as a new, challenging field. Interestingly, such findings are, however, seemingly overlooked in most of the literature in this field. Whether or not the majority of sex offender therapists, in fact, experience their work as mainly negative is therefore uncertain.
Finally, some studies of relevance to this issue have been inconclusive, reporting either mixed results or a lack of cut-off scores, making it difficult to evaluate the results (Adams, 2017;Crabtree, 2002;Steed & Bicknell, 2001;Thorpe, Righthand, & Kubik, 2001).
In other words, these clients are perceived to be especially difficult to work with, due to such behavior.
Categories like "sex offenders" may, however, be too heterogeneous to discuss as a single entity (Goldstein-Dwyer, 2014). These explanations could therefore lead to unjustified generalizations.
Furthermore, it seems to be a common belief that sex offenders are especially difficult to successfully rehabilitate, thereby, potentially, explaining negative responses among treatment providers. Defining "difficult", "rehabilitation" and "success" in this context is, however, not an easy task (Chudzik & Aschieri, 2013). Although recidivism rates are not true offence rates, as many instances remain unreported, they could still provide insight into this issue. While recidivism rates vary be study (e.g. as a result of measurement variation), sex offenders have higher rates of general recidivism than sexual recidivism (Hanson & Morton-Bourgon, 2004). However, they have lower rates of general recidivism compared to non-sex offenders (Langan, Schmitt, & Durose, 2003).
As such, sex offenders may be easier to successfully rehabilitate than non-sex offenders. In addition, treated sex offenders have lower rates of any recidivism compared to untreated sex offenders, indicating that treatment is, in fact, successful in many cases (Hanson, Bourgon, Helmus, & Hodgson, 2009). Consequently, difficulties in relation to rehabilitation, alone, cannot account for therapists' distress.
That many argue that sex offender therapy is difficult because of factors within the client, as opposed to the therapist, is, however, unsurprising. According to all of the introduced theoretical constructs, therapists' distress should be seen as a reaction to something external to the therapist (Clarke & Roger, 2007). What is surprising, on the other hand, is that none of the quantitative studies on this topic explicitly examined the inferred relationship between offender characteristics and therapists' distress (Moulden & Firestone, 2007). It is therefore unclear if the assumed problem stems mainly from the client's characteristics, exposure to, or knowledge about, sexual offences against children, a mixture of these, or something else entirely.
Studies using qualitative methodologies did, however, provide insight into this. Elias and Haj-Yahia (2016), who argued that existing research has merely sought to describe the consequences of work with sex offender, while neglecting how therapists perceive and cope with them, made the latter the aim of their research. According to one participant, negative impact was perceived to be a consequence of being confronted with their clients' offences: "The emotions I carry with me are difficult. I think this intensive exposure to sexual deviance scars the psyche" (p. 10). This is in line with the argument put forth by Kottler and Markos (1997), claiming that, One of the most unique challenges for therapists is working with pedophiles. Because of the reprehensible nature of the predatory behavior towards children, therapists need to be especially aware of how working with this type of offender will personally affect them" (p. 74).

Therapist-Focused Explanations
Many "moderating factors" within therapists have, however, been theorized. Yet, few studies incorporated these factors into their research. The role of treatment providers' individual differences, in relation to the experience of work-related stress, has therefore not been studied widely or systematically (Sandhu & Rose, 2012). In the following section, we discuss research that includes such therapist-focused explanations.
Demographic Variables -Studies that included therapist-factors tended to focus on demographic variables, such as age or gender, which were correlated with test-scores on questionnaires (Crabtree, 2002;Steed & Bicknell, 2001). No clear patterns have emerged so far. Generally, such studies may provide basis for further research but their utility in terms of explaining individual differences appear to be low.

Therapists' Experiences With Sex Offenders 502
Coping Mechanisms -Some authors argue that the use of either positive or negative coping-mechanisms may mediate or moderate the effects of the work-content. According to Wallace, Lee, and Lee (2010), who administered a questionnaire to 232 "abuse-specific counselors", the use of "active" coping strategies were linked to lower levels of burnout. Similarly, Thorpe, Righthand, and Kubik (2001), demonstrated that "positive coping mechanisms" were linked to improved work performance. These results were based on the administration of a questionnaire, which is thought to capture coping strategies and burnout potential, among 17 participants.  (2016), who conducted an interview-study that was analyzed via grounded theory, illustrated that therapists' perceptions about sex offenders personally affected them in their work. According to one participant, "A sex offender is a damaged person. He didn't start out that way" (p. 1160). As such, the participant was able to empathize with her clients, driving therapy forward.
Previously, it was noted that an increasing amount of therapists work with sex offenders and people with pedophilia. Based on the following results, however, it would seem that, overall, few therapists chose this line of work, perhaps due to negative attitudes towards these types of clients. No more than 4,7% of the 86 German therapists that participated in a survey conducted by Stiels-Glenn (2010) indicated that they were willing to accept people with pedophilia for treatment. Furthermore, only 3,5% indicated that they would work with child sex offenders. Almost half of the participants that specified why they would not provide treatment for this group explained that such choices had been made due to negative feelings or experiences concerning these types of clients. The remaining half stated that they did not have enough knowledge about this client group to conduct treatment competently.
Lack of Support -Some therapists may decline such clients in order to prevent judgment from others in their community. Results from the previously introduced study by Scheela (2001), for example, suggested that sex offender therapists that spoke in opposition to the dominating, negative societal attitudes towards this population were viewed as the enemy as well. Some therapists, who chose to work with these populations, may therefore feel uncomfortable sharing what they do for a living with others, in fear of being misunderstood (Kottler & Markos, 1997). For these reasons, among others, collegial support among sex offender therapists is thought to be an important moderator of the effects of such work (Clarke & Roger, 2002).
Value Conflict -In order to conduct successful sex offender therapy many, potentially conflicting, demands require careful consideration. Since therapy with these populations is often court ordered, a frequently asked question is, for example, how to define clear boundaries between therapy and punishment (Chudzik & Aschieri, 2013). It can therefore be assumed that many therapists are faced with a dilemma when considering to whom they owe their loyalty -society or the offender? i Another widely debated question concerns which approaches to sex offender treatment are the most effective and ethical. This relates, for instance, to what kind of stance therapists should take towards such clients. One school of treatment specialists contended that the first step in treatment should be to break denial and obtain disclosure by adopting a "confrontational approach" towards clients (Nori, 1992). In the 1990s this was the standard procedure in most sex offender treatment groups (Nori, 1992). Recommended therapeutic measures included various kinds of aversion therapy (Priest & Smith, 1992), for example, by pairing an arousing image with electrical shock (Quinsey, 1973). Drug therapy or "chemical castration", which is still widely used in some countries but may entail intolerable side effects, is also an available tool (see Panesar, Allard, Pai, & Valachova (2011) for a detailed description hereof). Castration has also been used in extreme cases (Nori, 1992).
Although these strategies have been widely criticized (Sandhu & Rose, 2012), discussions are still ongoing.
Therapists' Experiences With Sex Offenders 504 The abovementioned examples are included in this review because they highlight unique ethical dilemmas inherent in this line of work. While it is outside the scope of this review to discuss these questions in detail, we argue that such ambiguity may prove to be relevant in explaining potential negative effects from working with sex offenders, as such dilemmas may be strenuous to cope with. However, no researchers have, to our knowledge, focused on this potential relationship. Additional research that includes ethical aspects of these therapists' work is therefore necessary.

Organizational-Focused Explanations
Some researchers investigated more general aspects of sex offender therapists' work, such as their caseload, which was found to be the only significant predictor of burnout in a questionnaire study by Adams (2017).
Similarly, Shelby, Stoddart, and Taylor (2001) identified the setting to be the only significant predictor of burnout in their study. In more detail, providing therapy for sex offenders in secure settings (such as prisons) was associated with greater stress among the 86 therapists, who participated in their study. This may be because secure facilities usually house sex offenders that are thought to be more dangerous. Consequentially, therapists that provide therapy for this group of clients may feel unsafe in doing their work. These findings suggested that the work setting could be the most powerful moderator, with regards to work-related stress among sex offender therapists.
As such, we should expect sex offender therapists to be exposed to general work-related stressors common across occupations. In other words, identified stress symptoms among sex offender may not (only) result from working with sex offenders. In some cases, then, negative impact may be better explained by drawing on theory from general work psychology.

Discussion
In this review, we aimed at outlining the existing literature concerning sex offender therapists' experiences and the effects of this line of work on these. These studies were, however, not without theoretical and methodological limitations, some of which will be assessed in more detail in this section.

Sampling Bias?
Many research contributions within this field treated "sex offender therapists" as a homogenous group.
However, therapists working with sex offenders come from a variety of educational backgrounds and provide counseling in various contexts. Searching for general experiences and impacts among these practitioners may therefore prove difficult.
Additionally, to our knowledge, research to date has been conducted exclusively on English-speaking therapists, with the exception on one study (Elias & Haj-Yahia, 2016 Another sampling bias may be that most researchers drew their participants from organizations such as USAbased ATSA (Association for the Treatment of Sexual Abusers), which could further limit our possibilities to make any generalizations based on results from these studies.

Theoretical Preconceptions?
Despite evidence to the contrary, working therapeutically with these clients was mostly portrayed as strenuous work in the discussed literature. One possible explanation for this apparent paradox may be that most researchers are aware of the results that were published by Farrenkopf (1992). Arguably, this awareness may have resulted in a heightened sensitivity towards negative experiences among therapists. Researchers in this field should therefore be aware of this potential confirmation bias.
Moreover, most researchers, including Farrenkopf (1992), based their research on one of the aforementioned theoretical constructs, in which therapists' distress should be seen as a result to something external to the therapist, making the experience of distress among these therapists, to some extent, inevitable (Clarke & Roger, 2002). Currently, there is, however, little evidence to support this claim. It is therefore problematic that it is being inferred that therapists' symptoms are, by nature, a consequence of their work.
Nevertheless, such research may be challenging to conduct for a number of reasons. As mentioned, categories like "sex offenders" and "pedophiles" include heterogeneous individuals, making it difficult to determine whether or not it would be meaningful to search for a unified "nature", which, in turn, could account for therapists' "symptomology". Similarly, "therapists" may not react uniformly to the demands of their work. Taken together, the time may therefore be ripe to explore other theoretical perspectives, in relation to sex offender therapists' work experiences, than those associated with work-related stress.

Methodological Preferences?
Because the majority of research in the field is based on questionnaire studies, these studies seem to fail to address important aspects of the experiences of treatment providers within this field. While the validity of several of these questionnaires have already been criticized, in relation to the lack of profession-specific questionnaires (Clarke & Roger, 2002), the following questions should also be considered in relation to these: What kind of knowledge can be gained from "measuring" stress-related symptoms and comparing different professions in relation to these? How should "stress" be operationalized in this context? What can these questionnaires tell us about causation? What does it mean to have "an experience" and can they be quantified in a meaningful way? And, finally, what can they teach us about the meaning of our experiences, individual differences, and contextual and dynamic aspects? In reference to the latter, longitudinal studies would be of value, as little is known about how "impact" may change over time, but do not currently exist.
In addition, no research to date included reference groups outside the field of sexual abuse, to our knowledge. found that levels of vicarious trauma did not differ between groups. According to the authors, "This is not to suggest that there is no relationship between vicarious trauma and client population. It is possible that the Therapists' Experiences With Sex Offenders 506 clinician groups in this study were too similar (…) For example, sexual offenders may also share details of their experience of sexual abuse victimization as a child" (p. 66).
Additionally, the results of these questionnaires, which are typically based on one of the aforementioned theoretical constructs, where usually correlated with, for example, demographic variables thereafter. While correlations may provide some useful insight into a phenomenon, claims of causality should be made with great caution (Coolican, 2009).

Conclusions
We can conclude from the literature discussed in this review that some therapists who work with sex offenders are negatively affected by their work. There is, however, also sufficient evidence to determine that a considerable amount of these therapists find satisfaction in their work.
Overall, therapists working with sex offenders and people with pedophilia provide a vital contribution to our society. Enhancing our understanding of the challenges these therapists may face is therefore of great importance. Yet, in order to support treatment providers in their work, it should also be recognized that it is of equal importance to understand its rewards. As discussed in this review, most studies to date have, however, solely focused on determining the negative impacts of this line of work. Furthermore, a pursuit to identify, measure, and separate the negative impacts from the positive may be a questionable endeavor because such features, in fact, could be interrelated and coexist. It was, for instance, precisely the challenging nature of sex offender treatment that made it exciting to work with as well (Scheela, 2001).
Furthermore, we argue that "experience" must not be reduced to a concern with symptomology, as valuable aspects of human experience are left out. Instead, individual differences as well as dynamic and contextual features must be taken seriously, in order to achieve a holistic understanding of the issues in question. While not dismissing the value of questionnaire-based research, qualitative strategies may be better suited for this endeavor. Interpretive phenomenological analysis offers a promising avenue to this end, as a result of its consistent focus on the experiential content of consciousness (Smith, Flowers, & Larkin, 2009). Another line of research that might add important insights into the phenomenon, albeit from a different theoretical approach, is discursive psychology (Potter, 2012), which, amongst others, addresses issues of identity construction in talk (MacMartin & LeBaron, 2007).

Notes
i) To what extent these positions are really different is also a relevant discussion, albeit outside the scope of this paper.

Funding
The authors have no funding to report.

Quantitative: Self-Report (Maslach Burnout
Inventory and the COPE Inventory along with demographic questions).
Although, "the purpose of the study is to explore the level of Caseload size was found to be a significant predictor of the depersonalization aspect of burnout -DP (but not EE and PE).
Coping, gender, and years of experience were not significant predictors of burnout.
Carmel & Friedlander  Quantitative: Self-Report (PTSD symptomology, family and peer support, work load and supervision hours).
As a group, the participants experienced low levels of distress.
While support was significantly predictive of lower levels of distress, workload was not.
Farrenkopf (1992) An investigation of the impact on 24 therapists from their work with sex offenders.
Quantitative: Self-Report (Impact of working with sex offenders, personal coping strategies and gender differences). Furthermore, more than 85% of the participants reported moderate to high levels of compassion satisfaction, which indicates that many of the participants found their work satisfying.
Moore (2016) The study aimed at uncovering countertransference issues among seven therapists working with male sex offenders.
Qualitative: Interviews analyzed via the consensual qualitative research approach.
The participants were provided questions that examined countertransference reactions with regard to positive feelings, negative feelings, and sexual feelings. Quantitative: Self-Report (Fatigue-Self-Test for helpers, Impact of Events Scale Revised and years of experience).
None of the therapists were found to exhibit symptomology at a clinically significant level on the IES-R scale. From the results of the other scale, however, the authors concluded that, "as expected, working with perpetrators of sexual abuse was found to have a negative impact" (p. 5). In more detail, a mean score of 30.24 and SD of 13.6 were identified -within the potential range of 0-115. It is, however, difficult to evaluate these results, since no cut-off scores were provided. It is, furthermore, indicated that the test tends to err on the side of over-inclusion (Steed & Bicknell, 2001, p. 3 Overall, sex offender treatment providers did not obtain high scores on the MBI scales (as subscale scores were well under 27 and 13; at 19,6 and 8.21, respectively). Nevertheless, they scored higher than the mental health norm, yet lower than the social service worker norm. Moreover, sex offender treatment provider scored higher on "personal accomplishment" than the norm for both other groups -suggesting that sex offender treatment providers generally feel accomplished in their work.
Furthermore, those working with incarcerated sex offenders obtained higher scores on burnout measures than outpatient treatment providers. Quantitative: Self-Report (Professional Impact Questionnaire).
The level of burnout differed between the groups of professionals. Clinicians showed intermediate scores (on the "emotional impact" category, which is thought to reflect potential for burnout). While the authors hypothesized that the scale can "measure burnout experienced by professionals" (Thorpe, Righthand & Kubik, 2001, p. 198), there is no clear evaluation of the overall level of burnout among clinicians.
Similarly, indication of cut-off scores was not provided.
Nevertheless, the use of positive coping strategies was linked to decreased burnout potential. Quantitative: Self-Report (Impacts of Events Scale and trauma history).
Groups did not differ significantly in overall VT-score.
For the total IES, the mean score was in the moderate range, and 52% of the sample scored in the clinical range.
Nevertheless, according to the authors, no cut-off scores are available for this questionnaire. Moreover, the response rate was 23%, which raises questions concerning reliability.