Ensuring long-term success of personalised support for a young man with intellectual disability and harmful sexual behaviour: a Swiss case study

ABSTRACT Some people with intellectual disability who have sexually offended require long-term support with risk management. This paper demonstrates how least restrictive practices within a Swiss social care setting are utilised to support a young man with intellectual disability, mental health difficulties and persistent high risks. It is underpinned by the social model of disability, which directs attention away from individual pathology onto environmental support structures. Data was generated through qualitative interviews with the patient, forensic psychologist and social care provider and an in-depth analysis of the patient file. Current approaches to community support are synthesised and applied to this exemplary case. Hence, the analysis compares empirically based patterns from the case with the aims of risk management, person-centred planning and sexual offending treatment, highlighting the mechanisms that enable this support package to work. This results in a novel conceptualisation of sexual offending treatment success that encompasses environmental support structures. PRACTICE IMPACT STATEMENT The sociological focus of this paper emphasises the importance of learning amongst services on best means to provide collaborative, personalised support after sexual offending treatment. Achieving treatment success is no longer seen as a personal responsibility that lies with the person who has offended, and suggestions are made on means for managing individuals within social care settings. This makes vital reading for those involved in the de-institutionalisation of people with intellectual disability and persistent high risks.


Introduction
This paper introduces a young man from Switzerland who has sexually offended. He has intellectual disability, mental health difficulties, paedophilia, and high reoffending risks. He lives successfully in a social care setting and works as a kitchen assistant in sheltered employment. The young man has been with the same social care and employment provider for 15 years and restrictions that were initially placed on him have gradually reduced over time. This paper applies the social model of disability (Oliver et al., 2012) to highlight the mechanisms that contribute to the long-term success of this personalised support package, with a view to make recommendations on ways in which learning about what works can be transferred to other contexts. In short, this means that, instead of focussing on individual pathology, the focus is on how services and environments have had to change in response to the service user's complex needs.
To set the scene, the introduction provides an overview of three key practice areas that are synthesised in the young man's care package: risk management, person-centred planning and aftercare following sexual offending treatment.

Risk management
There are two types of offending risk factors: Static factors are historic or fixed (Lindsay et al., 2018). Dynamic factors can change, and when changed they impact on recidivism risk. They are subdivided into stable and acute dynamic factors. Stable dynamic factors, such as alcoholism, are expected to remain unchanged for months or years (Hanson & Harris, 2000). Acute dynamic factors, such as intoxication, change rapidly. Consequently, risk management efforts are aimed at the latter. Many individuals with intellectual disability who have persistent sexual offending risks require external support with managing risk (Griffiths et al., 2018).
In community settings day-to-day risk management is directed by social care staff. For instance, Boer et al. (2004) recommend that staff should monitor and be ready to respond to acute dynamic risk factors. However, many social care staff have little experience in working with people with forensic histories and they require additional guidance. Mosher (2010) claimed that there is surprisingly little evidence about this, but the international literature has evolved since. For example, De Santos' (2014) PhD thesis focussed exclusively on the role of social care workers during interventions into unacceptable sexual behaviour in people with intellectual disability in Scotland and Victoria (Australia). Stevenson et al. (2011) outline how care staff adopt a systematic approach to collecting data on the management of community access for individuals with high risks in Ontario (Canada) and McGrath et al. (2007) describe a community placement programme for people with intellectual disability with sex offending histories in Vermont (USA).
This paper adds to this international evidence base by introducing a person-centred service design within a social care setting in Switzerland. It furthermore demonstrates the practical use of assessing high risk care packages against Blasingame's (2016) seven areas of risk management, a simple tool that has thus far not been fully utilised. The seven areas are listed below: 1 Environmental contingencies include restricted access to a potential victim population and line of sight supervision when a person is in their proximity. 2 Coordinated case management refers to "concerted collaborative communication and shared responsibility between the multiple professionals and supervisory staff who form the individuals' risk management circle" (p. 242). 3 Staff competencies: Frontline staff need to receive sufficient training and support and they must be adequately motivated. 4 Psychiatric care: This includes the "prescription of appropriate psychotropic medication for those individuals who have dual diagnoses" (p. 243). 5 Cognitive-behavioural treatment (CBT), featuring relapse prevention and other sex offending specific training. 6 Law enforcement supervision: For those who have probation or parole status "it is imperative that all the members of the risk management circle understand the legal conditions imposed on the individual" (p. 243). 7 Victim advocacy: the individual and his support network are working towards the common goal of "no more victims" (p. 243).

Person-centred planning and positive behaviour support
In the 1990s Morris (1993) reported that the objective of community care planning was to fit people living with a disability into existing service provision. Since then, we have seen a gradual shift from a systemcentred approach towards a person-centred approach in some services. The aim of person-centred approaches is to tailor services around the individual, rather than enforcing one size fits all structures (Ratti et al., 2016). Person-centred approaches take into consideration the needs, desires, interests, preferences, understandings and lifestyle choices of the individual (Sanderson, 2000). This way of working has capacity to meet the social, economic and cultural rights of the person and it thus fits within a human rights based approach to disability (Degener, 2017). Evidence suggests positive changes following person-centred planning in respect to expansion of social networks, contact with family and friends, increase in choice, in scheduled activities and in community access (Robertson et al., 2007).
Within this framework positive behaviour support was developed for those who display harmful behaviours. This involves "recognising and responding to the individuality of each person, working to achieve meaningful personal outcomes, such as increased quality of life, and using empirically supported assessment and intervention strategies" (Grey et al., 2016, p. 256). Positive behaviour support assumes that behaviour is linked to and reinforced by environmental factors, hence behaviour change can be achieved "through manipulation of environmental factors as well as intervention with the focus person directly" (Blasingame, 2014, p. 487). Research has shown that this approach is effective in reducing the frequency and severity of harmful behaviours (MacDonald & McGill, 2013). Amongst others this paper will evidence how this way of working is successfully applied in the aftercare of a young man who has completed sexual offending treatment.

Sexual offending treatment for men with intellectual disability
Cognitive behavioural group therapy (CBT) for men with intellectual disability who have sexually offended teaches the men about sexual boundaries and the legal and social consequences of sexual offending. They learn to recognise their acute dynamic risk factors and triggers, such as boredom or contact with potential victim populations (e.g. Hordell et al., 2008;Lindsay, 2009;Malovic et al., 2018). Individualised risk management strategies are planned based on these personalised risk profiles (Hollomotz & Greenhalgh, 2020). A second objective is to enable planning towards a future of pro-social living. Very often men who have been identified as displaying harmful sexual behaviours come from disadvantaged backgrounds, characterised by erratic living situations, poor family relationships, neglect and abuse (Balfe et al., 2019). It is thus theorised that for them offending behaviour functioned to meet unmet needs and that "assisting individuals to achieve goods via non-offending methods may function to eliminate or reduce the need for offending" (Ward & Maruna, 2007, p. 108). Thus, Griffiths et al. (2018, p. 114) suggest that "the environments must provide optimal opportunity for offense-free interactions and activity, while providing appropriate supervision".
Evaluations of CBT group treatment for men with intellectual disability who have sexually offended mostly rely on recidivism data or psychometric testing pre-and post-treatment (Jones & Chaplin, 2020;Marotta, 2017;Schmucker & Lösel, 2015). This puts the onus for change on the individual and it does not reflect the reality that post treatment many men continue to be supervised within health or social care settings, where external support can help to prevent or in fact detect recidivism.
The importance of such external support was acknowledged by the pioneers of modern sexual offending treatment programmes for men with intellectual disability: One of the key principles underpinning the treatment outlined by Griffiths et al. (1989, p. 6) was that "the key to effective intervention is the development of environments that support change and prevent future problems". Haaven et al. (1990) went as far as suggesting that increased skills within the workforce involved in a person's care is an indicator for treatment success. This paper thus elucidates how such success can be achieved and maintained though building strong collaborative support structures in response to individual's risks, skills, and aspirations.

Methods
This research crosses disciplinary boundaries by applying sociological analysis to a research area that receives most attention from forensic psychology (Hollomotz, 2014). Sociology has much to offer to this area of research and practice, due to its capacity to make sense of complex social structures. The section on analysis will highlight how the social model of disability focusses attention on environmental and contextual issues. First, however, an overview of the research setting, sampling and fieldwork is provided.

Research setting and sampling
This paper presents a single case study from the canton Thurgau in Switzerland.
Here, Forio 1 , a forensic outpatient service provides treatment to men with intellectual disability who have sexually offended. Participants have medium to high risks, as assessed via Armidilo-S (Boer et al., 2013). By 2017, 33 men had completed the outpatient treatment group, which is described in detail in Hollomotz and Caviezel Schmitz (2018). From this group, eight men were selected for qualitative case study research. The aim was to achieve maximum variation in the sample and therapists selected cases they considered successful, unsuccessful and unusual. Initial contact was made by a former therapist and once a patient asserted an interest in taking part, a date was set to meet the researcher. There were some practical difficulties in arranging interviews. Eventually, six men participated in the study.
Pedro (not his real name) was selected to be presented here as a single case study for the following reasons: Despite Pedro's reported engagement with treatment his risks remain high. Fifteen years after completion the forensic service coordinate his case management, which makes this an unusually extensive aftercare period. However, the literature review highlighted that complex cases of men with persistent risks post treatment occur in other contexts (e.g. McGrath et al., 2007;Pritchard et al., 2016;Rea et al., 2014). Pedro's case is thus at the same time unusual within its context and common across contexts. It is also exemplary, in that, as will be shown, all seven areas of collaborative risk management (Blasingame, 2016) are satisfied as part of a personcentred support package.
This study received ethics approval from the Kantonale Ethikkommission Thurgau.

Fieldwork
The case study is "an empirical method that investigates a contemporary phenomenon (the 'case') in depth and within its real-world context, especially when the boundaries between phenomenon and context may not be clearly evident" (Yin, 2018, p. 15). As recommended by Yin (2018), multiple sources of evidence were consulted. The patient file offered insights into Pedro's childhood, own victimisation and offence history. It included psychological assessments, reports about treatment progress and notes taken at case management meetings. A semi-structured interview was conducted jointly with Pedro's forensic psychologist (FP) and his residential group manager (RM). They explained how risks are managed and remembered how this had changed over time.
Pedro was interviewed separately. He reflected on his experience of attending treatment and receiving external risk management support, his satisfaction with this support and his social care placement. His interview was conducted in an accessible manner. As outlined by Hollomotz (2018), this included question phrasing in plain language and a focus on topics that were relevant to Pedro's direct experiences. The interviews were fully transcribed and added to the case study database, alongside notes taken from the patient file. The researcher began to write the paper in English, keeping citations in original German and they translated these once a full draft had been created.
During a subsequent visit to the forensic outpatient service, the researcher discussed the draft paper and its emergent themes with the forensic psychologist, thus conducting a partial member check of the data (Guba & Lincoln, 1982). As a result, some details were refined. What is more, the researcher's prolonged engagement at this research site across five separate visits enabled them to immerse themselves in the setting. They were able to observe the transferability of the approach described here, as practice across several cases and provisions at this service were underpinned by similar mechanisms to those drawn out in the discussion of Pedro's case.

Analysis
At the beginning of the results section a brief overview of Pedro's diagnostic labels and treatment progress is provided. However, the aim is to move the reader's gaze beyond individual factors and to explain how needs arising from Pedro's impairments are responded to by his social environment in the context of ongoing risk management. This is in line with the social model of disability, which makes a distinction between impairment and disability (Oliver et al., 2012). Impairments, such as intellectual disability and mental health difficulties, are located at the individual level. However, impairments do not have to become disabling. Disability only occurs when practices and environments are ill equipped to respond to an individual's impairment related needs. For example, in the section on treatment progress, Pedro's forensic psychologist will comment that Pedro's intellectual disability affects his ability to acquire independent risk management skills. Much of this paper is focussed on explaining how this is responded to by his environment, who support Pedro in areas he cannot manage himself. Thus, it will be demonstrated how Pedro's support team work on preventing his impairment effects (his inability to manage risks independently), from leading to disability, which could manifest in Pedro being "locked away" from social intercourse. The same framework is applied to make mental healthcare support structures explicit. It will be demonstrated how even risk management is conducted in a "holding" manner in response to Pedro's fragile mental health.
Critical realists argue that the reality of events in the social world can only be understood if we identify the underlying structures that generate those events (Bhaskar, 1989). To explore those and in line with the interpretivist tradition (Mason, 2018) this researcher seeks to make sense of the world through the eyes of a man with intellectual disability and professionals involved in his care. Puzzling is that Pedro has made limited progress towards becoming able to manage his own risks, but that professionals and Pedro himself believe him to be a "success". This means that further conditions have to be at play in determining this outcome.
The analytical technique of pattern matching (Sinkovics, 2018) was used to compare empirically based patterns from Pedro's case with the aims of risk management, person-centred planning and sexual offending treatment. These three areas of practice are thus synthesised to explain the mechanisms that enable the support package to work. Beyond this, the analysis makes sense of why these work well together. The author zigzagged (Emmel, 2015) back and forth between the data and the outcome of presumed "success", seeking to explain the causes for these high levels of satisfaction. In doing this, they consulted with Pedro's forensic psychologist, listened to a wider network of stakeholders (as summarised by Hollomotz & Talbot, 2018, 2021 and conducted some increasingly purposeful literature searching, as recommended by Cooper et al. (2020). As a result the discussion draws out underpinning "deeper" processes, which generate the observed relationship between the support package and its outcome of presumed "success" (Dalkin et al., 2015). Through this process a novel understanding of "success" is developed.

Results
This section begins with some historic and diagnostic information about Pedro. The subsequent four sub-sections cover treatment progress, mental healthcare, risk circles, environmental contingencies and supporting pro-social conduct. The results are loosely structured around Blasingame's (2016) seven areas of risk management and they will be drawn out more explicitly in the subsequent discussion.

Historic and diagnostic information
Pedro's patient file includes several comprehensive assessments. The following offers a summary. Firstly, Pedro has diagnoses of intellectual disability (IQ 56), antisocial personality disorder (F60.2 in ICD-10-CM, 2018), major depressive disorder (F33) and paedophilia (F65.4). Specifically, he has a rigid sexual preference for boys aged 5-7. How these conditions affected Pedro's treatment progress is discussed in the subsequent two sub-sections. In addition, the patient file provides detailed information about Pedro's history. His mother experienced intellectual disability and severe mental health problems, which resulted in several hospital admissions. Pedro's father suffered from alcoholism. Although he left the family home shortly after Pedro's birth, Pedro remained in contact. Services became involved when Pedro was six years old. It emerged that, in early childhood, Pedro had experienced child neglect, unavailability of attachment figures and sexual violence within the family home. The exact nature of the sexual offences and identities of the perpetrators remain unclear, with the biological father, an uncle and older siblings implicated.
Aged seven Pedro moved to a boarding school for children with additional educational needs. Aged twelve he disclosed that he had been sexually abused (statutory rape) by his brother and other older boys over a prolonged period. He admitted committing similar offences towards younger children. Once the abuse at the school was uncovered and investigated, a total of 37 victims and 24 perpetrators were recorded. These were dealt with through the guardianship office, which imposed restrictions on Pedro, for instance in terms of where he was to live and with whom. As was the case with the intra-familiar sexual abuse experienced by Pedro, there were no criminal prosecutions. Pedro moved to a foster family and continued his schooling. Aged 16 he attended the outpatient treatment group for juveniles who had sexually offended.
When Pedro turned 17 his treatment and schooling were coming to an end, and he needed to plan towards his transition into adult services. Due to the restrictions placed on Pedro by the guardianship office, which remains involved to this day, independent living was not an option. Switzerland does not have dedicated long-stay forensic provisions for people with intellectual disability (Georgescu & Styp von Rekowski, 2018) and a solution was sought within existing adult social care structures. Eventually, Pedro became the first user with known forensic history to receive services from a large social care provider, which maintains 40 beds within a residential home and 85 beds within community based supported living groups. Initially, he moved into the residential home. In addition, the service offer 155 sheltered employment and 10 apprenticeship places. Pedro started a sheltered apprenticeship, completed this successfully and was kept on as an employee. Fifteen years on Pedro continues to use the provider's residential and employment services. Restrictions that were placed on him have gradually reduced over time, as will be described shortly. Today Pedro lives in a supported living setting. He still visits his former foster family for respite care on some weekends and in his holidays.

Treatment progress
The Forensic Psychologist (FP) asserts that the overarching aim of their work with Pedro is what Blasingame (2016) terms victim advocacy: We need to ensure he feels safe in two ways. On one hand, due to his victimisation history he needs to feel secure and trust that he will not be victimised again. On the other hand, because he has perpetrated, he needs to be protected from reoffending. This has been successful as Pedro has never again been victimised or offended. The FP and RM described Pedro as a compliant participant who engaged well with treatment. Pedro emphasised that this was an important milestone: "I was able to educate myself about how to cope with things I have always found hard in life." The FP asserts that Pedro has got as much as he could get out of the group treatment, which is the acceptance that he must stick to a system of rules. […] "No sex with children". Pedro understands and superficially sticks to the boundaries of sexual behaviour (FP).
However, she adds that in relation to Kohlberg's (1969) moral development theory, Pedro is at stage 1: He displays good conduct to avoid punishment. Pedro was unable to link his own victimisation experiences to those of his victims and he was unable to develop victim awareness or victim empathy (Williams & Mann, 2010). The significance of this is that progress in these areas is often used as an indicator of treatment success (e.g. Craig, 2019;Young et al., 2015). However, Pedro's care team believe that treatment was nonetheless beneficial. The FP understands that some of the remaining shortcomings in self-management skills are caused by Pedro's impairment: "He lacks the cognitive abilities to accept and manage his sexual preference." Moreover, the FP believes that, "if he could move freely and go to playgrounds, he would. The risk that he would offend is high". In response to these risks, ongoing coordinated case management is required, as explored in the section on risk circles. These also need to address Pedro's mental healthcare needs, as described below.

Mental healthcare
In the 15 years following group treatment Pedro experienced at least five episodes of severe mental health crises. On three occasions he was close to hospitalisation, due to suicide risk. During his first few months in the residential home Pedro woke at night with vivid suicidal thoughts. The staff he approached felt overwhelmed. It was therefore established that We needed to be briefed on issues relating to mental health. How do I respond to a person who has these kinds of fantasies? It was important to have a contact person who was able to advise and share the risks (RM).
A mental healthcare strategy was put in place: Pedro received individual therapy with FP for three years. Today, Pedro continues to see his psychiatrist every three months. This is standard practice for patients who are prescribed high levels of Quetiapine. 2 Pedro's emergency mental healthcare strategy specifies that, if needed, he is to turn to a designated staff who can arrange urgent appointments with the forensic outpatient service or the psychiatrist. Appointments are usually offered within one week. In addition, Pedro explained that he uses the traffic light system that was introduced in group treatment to assess his emotional wellbeing and ruminating thoughts green = ok, orange = getting risky, red = dangerous/out of control, compare to Hollomotz and Caviezel Schmitz (2018): "I need to visualise the traffic light, so I know where I stand. 'Am I green, or more orange or have I crossed into red?' It helps me understand where I'm at." At orange Pedro would ask for help: I have quite a few people around me. With my problems […] I need enough support persons, who are so strong that they can steer me and my thoughts and my past, to steer it with me, so I get on a good path. The staff need to be as strong as the feelings that are sometimes buzzing around inside me.
Pedro's most recent two episodes of mental health crisis were less critical than his earlier ones. The RM believes that this is because the support circle has learned how to prevent crises: Relationships with staff have a lot of capacity. We have a team that has been consistent for years and this team can hold him. The thoughts still come, and they still flood him, but we have more capacity to bring him back.
To "bring him back" staff remind Pedro of the pro-social life goals he has achieved, such as new skills acquired, positive relationships and freedoms he enjoys, as discussed shortly.

Risk circles
As described by Griffiths et al. (1989), the overall responsibility, decision making and planning of Pedro's support during treatment was shared with a strong system of support persons through "risk circles". What is unusual about this case is that these have continued to meet bi-monthly for the past 15 years and plans are to continue these indefinitely. A similar long-term arrangement is in place for only one other patient at the forensic outpatient service. In all other cases the aftercare period is limited to one year. The membership base of Pedro's risk circle has been consistent. As illustrated in Figure 1, it includes representatives from the social care organisation and external members. External are the forensic psychologist, psychiatrist and legal guardian. Internal are the managers and keyworkers from Pedro's residential group and sheltered workplace, as well as the adult protection link worker. 3 These members feed back to the wider staff teams within the workplace and residential group. Pedro's foster family do not attend risk circles, but they are regularly updated by FP on a need-to-know basis.
Risk circles regularly review and update Pedro's intervention plan, which is akin to the Stop & Go Intervention Plan introduced by Blasingame (2014). It lists goals linked to avoiding high risk situations, such as boredom, alongside approach goals, such as suggestions of things to keep busy with. Risk circles reflect some of the practices observed during positive behaviour support (Bartle et al., 2016). In line with this approach the forensic outpatient service aims to include family members in risk circles. However, unusually, Pedro's risk circle does not include his family, as relationships are strained, and contact is a trigger for Pedro's low moods.
According to the RM there is a risk of becoming complacent with Pedro. Risk circles offer important reminders: The care team needs the long-term consistency. When we have staff turnover the new staff need training. We benefit from external feedback and from being reminded that things are working well, because we are doing a good job and not because he has been cured.
Risk circle meetings are split into two parts: At first, the professionals provide updates on Pedro's conduct and discuss concerns. Direct care staff ask for advice. For instance, the residential group keyworker brought two DVDs to the last meeting, as Pedro had requested to watch these and the keyworker was unsure whether they were appropriate. The risk circle advised that these were not suitable for Pedro, as there were children among the main cast. For Pedro, this material increases risks of having sexual thoughts about children, which can spiral into heightened anxiety. Pedro attends the second half of the risk circle, and he can bring requests to the meeting, such as asking permission to visit new community venues, which have not yet been approved. Meetings are also used to communicate to Pedro what is and is not acceptable behaviour.
Guardianship can be useful for reinforcing boundaries (Hill-Tout et al., 1998). Consequently, the legal guardian must formally sign off changes to the intervention plan. As such, risk circles formalise law enforcement supervision, even though Pedro has never been formally charged. Nonetheless, the FP makes clear that during meetings We maintain a positive atmosphere. If he did something that was not acceptable, then he did something that was not acceptable. But we would not say that he is a bad person. The key message: "We accept you as you are" stays (FP).
Usually, if Pedro displayed a behaviour that required an intervention, this would happen straight away. The support team would not wait for the next risk circle to bring this up. Keeping the risk circles positive to some extent, even at difficult times, is key considering Pedro's fragile mental health. The risk circle thus becomes a "holding environment", within which criticism is raised in as honest, friendly and supportive a manner as possible. This is driven by the belief that a supportive atmosphere is more likely to keep Pedro engaged (Schmelzle, 2015). "Holding" within risk circles thus consists of fundamental "caregiving activities and processes that facilitate growth and development" (Winnicott, 1965, p. 33). As there was a considerable delay in publishing this paper during unprecedented times the author e-mailed FP to request a brief update on how and whether these practices were able to continue during the Covid-19 pandemic. Box 1 summarises the reply.

Box 1. Covid-19 update
During Covid-19 lockdowns the social care organisation did not allow external visitors. Pedro was unable to access his regular therapy sessions, as only crisis intervention sessions were allowed to take place, but Pedro did not need these. A few weeks into the first lockdown risk circle meetings continued via video conference and once restrictions eased, they returned to their face-to-face format, following applicable social distancing and hygiene guidance.

Environmental contingency
When Pedro moved into the residential home his guardianship order required that he needed to remain in his bedroom at night. To facilitate this, the addition of en-suite facilities was commissioned for Pedro's room, something that was not available to other residents, who used shared facilities. The home had 24-h staffing and an alarm was activated at Pedro's bedroom door overnight. During the day, Pedro had to remain under eyes-on supervision. He was not allowed off site, except when visiting his foster family. After this arrangement had worked for a few months, a plan for community outings was drawn up. The requirement for eyes-on supervision when visiting populated areas still applies today and there are restrictions on Pedro's movements. He is not permitted to visit swimming baths, playgrounds or other venues with a high likelihood of encountering children.
Eight years ago, the risk circle decided that Pedro could move into a supported living group: He has made progress, which is why it was possible to move him to a less restricted flat. We would not have done this without evidencing first that he would stick to the rule of staying in his bedroom at night (RM).
Similar occurrences of granting increased community access and privileges in response to compliance with risk management strategies are described by Pritchard et al. (2016) and Stevenson et al. (2011). Service planning started by acquiring a suitable property in a small hamlet close to the provider's main site with its residential home and sheltered employment facilities. Next, social care support was designed around Pedro's needs. The living group has expanded over the past eight years. Today, it is shared by seven men, most of whom have forensic histories. This group does not have a night staff. Pedro's door is no longer alarmed. "He has as much independence as possible. But we know that he will always need supervision" (RM). The patient file evidences that decisions to reduce supervision were data driven, as recommended by Stevenson et al. (2011): Observations on Pedro's conduct were reported at risk circles and here decisions were made to gradually alter the intervention plan and environmental structures.
Over time it was thus possible for Pedro to take small steps towards increased independence. Pedro now walks the rural route from his home to the sheltered workplace at the main site without staff. As Pedro sets off home staff notify the workplace and vice versa. If Pedro does not arrive within a designated time (20 min), staff would search for him. In seven years, this happened twice. On both occasions Pedro was held up talking to a farmer. However, the professionals assert that Pedro is closely monitored, and staff are always ready to respond to any changes in mood or presentation, as Pedro's risky thoughts can at times "flood him". The RM recalls an incident, which took place six months prior to the interview: Pedro went on holiday to a campsite in Spain. You cannot prevent the presence of families at a campsite. One staff noticed that Pedro kept sitting on a bench outside the bungalow, so she sat on it. She observed that this location gave her clear view of a child, who was frequently walking from his parent's to his grandparent's bungalow. Thus, Pedro had selected this spot deliberately and if you know him you know how his gaze changes. He has a very peculiar facial expression and you must be very firm and bring him out of his trance.
The FP and RM agree that Pedro requires an intensive support package to manage risks. "I think he is kept safe by the people around him. The system needs to work, or he goes off the rails" (RM). RM discusses the example of Pedro going to a supermarket. For outings like these to populated areas, Pedro knows that he must ensure that he remains within staff's eye-line: "With new staff he has tried to walk into a different aisle, so the staff cannot see him anymore and has to actively go looking for him" (RM). This indicates a lack of generalisation of relapse-prevention behaviours (Griffiths et al., 1989;Rea et al., 2014), in that Pedro follows the agreed rules with familiar staff, but his repertoire of prosocial behaviours does not include the generalisation that he needs to observe the same rules with unfamiliar staff.

Supporting pro-social conduct
The professionals commented that Pedro does not offend: RM: Because he would lose too much, I think.

FP: Yes.
RM: He would lose everything.
FP: Yes, exactly. Because he has built something up, something that is important to him. This is his world. This is his life, with all the possibilities. It offers him everything! Pedro confirmed that it was important to him that his behaviours remain within agreed boundaries: "I attended the group to learn about my risks.
[…] What can I do? Where are my boundaries? […] That's what I took away from the group." Pedro aspires to lead a life on par with that of other users of Swiss social care services, despite the considerable restrictions that are imposed on him. He described meaningful positive relationships, listing approaching a dozen names of individuals whose company he enjoys. This includes peers and staff at home and in the workplace. Pedro works as a kitchen assistant in the canteen on the main site of the social care provider. The RM asserts that the kitchen ensure that Pedro has opportunities to take pride in his work. For instance, Pedro attended a praline making workshop at the local college and established himself as the "praline expert". When Pedro's mental health dips, staff remind him how much he is valued at home and at work and this helps to improve Pedro's mood.
Overall, his meaningful (sheltered) employment, good relationships with staff and peers and living in a group where he feels safe and enjoys some freedoms result in Pedro achieving a good quality of life, which the professionals say motivates him to uphold pro-social behaviours. For instance, Pedro recalled his thoughts during a recent incident when a peer infuriated him: I thought: "I could get him in an armlock. I could yell at him and call him all sorts … I could even beat him up." But before all the risks came, before he could drive me totally mad, I made my way to a member of staff. I got help. Because I know my risks -if I can't do it on my own, I need to get help.

Discussion
The Swiss health system performs very well with regard to a broad range of indicators (De Pietro et al., 2015), but this is not the chief reason why Pedro's support package is so comprehensive. Regardless of context, individuals with persistent high risks post treatment require long-term, structured support. Pedro's case illustrates a point raised by Griffiths et al. (1989) and Haaven et al. (1990): The onus for change is not just on the individual, but environments need to increase their skills and adapt their ways of working to support individuals to manage their risks and achieve a good, prosocial life. To illustrate how this is done in Pedro's case this discussion assesses this support package against the seven areas of collaborative risk management outlined by Blasingame (2016), which were introduced at the start of the paper. In the following, each of the seven terms is italicised for ease of reference.
Pedro's risk circles coordinate case management (Blasingame, 2016). One function of these interagency meetings is that keyworkers can seek advice and discuss care staff's concerns. A further function is information sharing. In that way, risk circles deliver on staff competencies (Blasingame, 2016). This ongoing information exchange between clinical specialists and social care is commendable. For instance, social care staff in De Santos' (2014) study reported difficulties in accessing advice and input from health professionals. Similarly, social care staff in Robertson and Clegg's (2002) study reported difficulties in accessing information about individual's risk profiles and asked for more guidance. Community teams at a research site in England evidenced that such information is essential for successful community resettlement from secure hospital (Hollomotz, 2021).
Risk circles formulate, regularly review and update Pedro's intervention plan. Figure 2 illustrates the triangulation of three equally weighted attributes at the centre of such planning: Pedro's skills are considered and utilised, giving him leadership wherever possible. For instance, Pedro walks to and from his workplace independently. Skills are carefully weighed against risks, which gives rise to external risk management support. For instance, Pedro remains within eye-line supervision in settings where risk of child presence increases. Finally, planning is driven by Pedro's aspirations. For instance, he was able to realise his dream of moving into a supported living group. He is also supported to access community-based activities, such as the praline making workshop and he continues to visit his foster family, thus upholding key relationships.
This process of designing support from the ground up mirrors some of the key features of personcentred planning: Support structures are formalised by Pedro's intervention plan, which reflects what is important to him. Moreover, "the plan results in ongoing listening, learning, and further action" (Sanderson, 2000, p. 7). It is altered in response to requests from Pedro, changes in his environment and observations about his conduct and mood. Moreover, to some extent the plan reflects what is possible, not just what is available (Sanderson, 2000). Both, the care provider and the forensic service had to reimagine their provisions to enable them to respond to Pedro's needs. However, Pedro's support package remains within existing residential and occupational support structures, enabling Pedro to lead a life that aspires to be on par with that of other users of Swiss social care services. Person-centred planning has the potential to think beyond and to dream up new provisions (Sanderson, 2000). Maximising person-centred planning to its full potential could be an aspiration for the future.
Pedro's intervention plan is underpinned by three of the areas of risk management outlined by Blasingame (2016). First, even in the absence of parole requirements, the legal guardian's power to authorise changes acts in lieu of law enforcement. Second, it sets out environmental contingencies, which have changed over time. Finally, the mental healthcare strategy includes that Pedro is prescribed psychotropic medication and he regularly sees the prescribing psychiatrist. Emergency appointments can be requested. Figure 2 further illustrates that Pedro's risk circle and care teams aim for consistency in the ways staff work with him across and within settings. Pedro behaves differently with new staff, so reinforcing that the same rules apply is important. Consistency thus links to boundary setting, in that boundaries are consistent across contexts and to some extent across time, as any easing of restrictions must be earned through display of good conduct and testing. There is mutual trust between Pedro and his support circle, who rely on Pedro following the rules set out in the intervention plan. In return, Pedro can rely on the support circle "holding" him. Here, this is used as "an elastic metaphor that alludes to the provision of an optimally responsive environment that meets the patient's need for affective attunement" (Slochower, 2013, p. 22). This is in line with the principles of trauma informed care, in that a non-threatening environment is established that facilitates trust and emotional safety (Levenson, 2014).

Conclusion
Pedro was unable to learn independent risk management through group sexual offending treatment, even though adaptations were made to make this accessible for individuals with intellectual disability and even after he received additional personalised risk management support for the past 15 years. Pedro's therapist described independent risk management as lying beyond his intellectual capabilities. Pedro also requires support to manage his mood and a comprehensive mental healthcare strategy is in place.
However, looking at his case through a social model lens reveals that much support is provided to prevent Pedro's impairments leading to disability. Pedro is able to lead a life close to on par with other social care users in his local area, accessing similar work, living and community leisure opportunities, as far as his risks permit (i.e. playgrounds and swimming baths are off limits). Sexual offending treatment had been used, not just to treat Pedro, but also to "treat" the environment. Through treatment serving as an extensive assessment period the support network gained vital knowledge about Pedro's risks and triggers and best means of supporting him. Learning about Pedro continued beyond treatment and to this day.
On the flipside, the social model can be used to argue that the care package has not yet gone far enough. Pedro leads a life that aspires to be on par with other social care users in his country. He does not lead a life on par with the general non-disabled population in Switzerland. To achieve equality in this sense he might want to consider moving out of his group setting and into the community. He would leave sheltered employment and instead participate in the free labour market. Yet, his high risks, combined with structural inequalities persisting within the labour market (e.g. Trezzini et al., 2021;van der Zwan & de Beer, 2021), mean that these are unlikely to be considered realistic options. Examining these inequalities further is beyond the scope of this paper, but this sidenote is an invitation to consider future possibilities beyond what has been achieved so far.
Back to the reality of what has been accomplished. Figure 2 congers up an image akin to a safety buoy. This metaphor works well and in multiple ways: Pedro's support circle "hold" him when his selfdestructive or sexual thoughts are at risk of "flooding" him. Moreover, Pedro is kept safe when he perceives risk from outside his support circle. At times of crisis the support circle keeps Pedro afloat by maintaining an accepting and positive atmosphere. It thus achieves "holding", which is characterised by mutual trust, consistency and boundary setting. The outcome is that the twofold aims of "no more own victimisation" and "no more offending" are upheld alongside a consistent approach to preventing mental health crisis. This is the chief reason why both, Pedro and his support network consider Pedro to be a success.
Although cases of individuals with persistent high risks post treatment tend to be unusual within their context, they occur across different contexts, where they use up significant resources, which is why learning about Pedro's case should be of interest to international readers and especially those involved in the de-institutionalisation of people with intellectual disability and persistent high risks. Those seeking to implement structures around a person with complex needs should bear in mind that creating Pedro's support package took perseverance through ongoing observations, listening and learning. It was built from the ground up by putting Pedro's needs, skills and aspirations at the centre and developing personalised support around him. There is no straightforward way of copying practices across to other settings and the means for achieving equilibrium through buttressing person-centredness with strong structures and relationships will differ for each individual. However, this paper demonstrates the practical use of structuring care packages around Blasingame's (2016) seven areas of risk management, a simple tool that has thus far not been fully utilised. It is recommended that this tool is applied to other settings and that, as demonstrated in this paper, this is combined with a drive to achieve person-centred planning. Notes 1. https://www.forio.ch/ 2. Pedro takes 100 mg/day (PMR:50-100 mg/day) to counteract symptoms of sensory overload. 3. Large Swiss social care organisations typically employ a dedicated adult protection link worker who supports the prevention of violence and sexual exploitation. Amongst others they offer staff training, act as a first point of contact and network with relevant organisations (Verband für anthroposophische Heilpädagogik und Sozialtherapie, 2021).