Challenges in Adopting Recovery-oriented Practices in Specialized Mental Health Care: “How Far Should Self-Determination Go; Should One be Allowed to Perish?”

Mental health services need to transform from a primary focus on symptom reduction to a recovery-oriented delivery. Research on recovery-oriented practices is mainly based in community mental health settings, while research on specialized mental health care remains scarce. In this article, we aim to identify and explore the experiences faced by professionals working in specialized mental health care units that aim to be recovery-oriented. Data were collected during seven focus group interviews with 45 professionals from four psychiatric hospitals and district psychiatric centers in Norway. We used reflexive thematic analysis to interpret the data. Three main themes emerged from the analysis: (a) disease-oriented structures, (b) negotiating roles and (c) risk management. This study identified the many tensions professionals face as they try to shift specialized mental health care toward a recovery-oriented paradigm. Specifically, professionals must balance managing risks and promoting self-determination. To succeed, it is not sufficient to implement practices that are characterized as recovery-oriented without also changing existing systems, structures, and frameworks. We suggest approaching recovery orientation through shared decision-making. This could contribute to the promotion of self-determination and increased inpatient safety in specialized mental health care.

The United Nations Human Rights Council (2017; 2020) has called for ''a revolution'' in mental health services with a deep commitment to human rights, dignity and non-coercive practices. Moving away from paternalistic mental health practices to those that support patient autonomy (Davidson et al., 2010) entails a paradigm shift in thinking, service orientation and utilization of resources (Glover, 2005). To make a change, recovery-oriented practices are highly recommended (Australian Health Ministers' Advisory Counsil, 2013;Mental Health Commision of Canada, 2015;New Freedom Commision on Mental Health, 2003).
The core characteristics of recovery-oriented practices are services that work toward articulating an organizational commitment to promoting citizenship, supporting personally defined recovery and developing working relationships (Le Boutillier et al., 2011).
Briefly, recovery-oriented practices can be understood as the services and professionals that work together with people with serious mental illness to support their recovery processes. In this article, recovery is understood as a multidimensional process that occurs in people's lives through interacting with others to live a satisfying and equally valued life as a full citizen in a community (Ørjasaeter, 2019; Ørjasaeter et al., 2018). However, adopting recovery as the guiding framework for service delivery has generated debate (Davidson et al., 2006). One critical point is the professionals' concerns about increased exposure to risk and liability (Davidson et al., 2006). Secondly, it is highlighted that there is a danger created by unreasonable expectations for patients. Lastly, it is pointed out that professionals may lose a focus on their role (Meehan et al., 2008).
Research on recovery-oriented practices is mainly generated in community mental health settings. Although there is increasing interest in specialized mental health care, research in such contexts remains scarce (Waldemar et al., 2016). A limited number of studies have attempted to examine recovery-oriented practices in such settings, and findings have identified different understandings of recovery, a lack of clarity about what constitutes recovery-oriented practices and challenges in implementing such practices (Aston and Coffey, 2012;Chester et al., 2016;Le Boutillier et al., 2011;Waldemar et al., 2016). Cleary et al. (2013) claim that in practice a recovery orientation is more often rhetoric than an integral aspect of practice. The lack of integration is linked to physical structures not appropriate to recovery-oriented practice (Cleary et al., 2013), ethical challenges like promoting selfdetermination (McKenna et al., 2014) and professional responsibility for managing possible risks to patient safety (Chen et al., 2011).
Strengthening self-determination and minimizing risk are core components of specialized mental health care (Perkins and Repper 2016). However, these components are assumed to be incompatible (Perkins and Repper 2016), although risk usually takes precedence (Aarre 2018). Traditionally, self-determination is understood as individual autonomy, in which human beings are considered as free individuals, with a right to make their own decisions without interference from others and with an ability to take responsibility for the consequences of their own actions (Mackenzie 2019). The United Nations convention on the Rights of Persons with Disabilities (CRPD), represents a paradigm shift in the understanding of self-determination, acknowledging that the right to self-determination is not subject to disability-based restrictions. The convention further refuted that people with disabilities lack the autonomy required to have human rights (Skarstad 2018). Self-determination is re-defined as a relational phenomenon that is exercised in relation to the environment, realized through relations (Skarstad 2018) and seen as both freedom and opportunity (Mackenzie 2019). Risk is usually found as a fundamental component of good mental health practices and often understood as the probability that something will happen that may have potential beneficial or harmful outcomes for the individual or the surrounding environment (Morgan 2000). Such events usually refer to behaviors resulting in suicide, self-harm, aggression and violence, and the neglect, abuse and exploitation by self or others (Morgan 2007). Professionals have traditionally had an important role managing risk through taking over control to reduce danger (Perkins and Repper 2016). A recent study by Jones (2020) states that risk management and recovery-oriented care are not mutually exclusive, and that they can coexist.
To strengthen patient self-determination, shared decision-making is gaining increased prominence in healthcare policies worldwide (Stacey et al., 2016). Shared decision-making is a process in which professionals and patients work together to select tests, treatments, management or support packages based on clinical evidence and patient's informed preferences. It involves the provision of evidence-based information about options, outcomes and uncertainties together with decision support counseling and a system for recording and implementing patient's informed preferences (Coulter and Collins 2011). In shared decision-making the patient's experiential knowledge and the professional clinical and scientific knowledge is integrated with the goal of making better decisions (Beyene 2020).
Specialized mental health care differs from community mental health services through their mandate of treating illness preferably within a short time. In addition, expectations of acting as experts and the professionals' special responsibility of protecting patients' lives and health also allow them to use coercion if voluntary measures have failed (Aarre 2018). When all mental health services are encouraged to work in accordance with the principles of recovery orientation, more knowledge of what this will entail in specialized mental health care will be needed.
The aim of this article is to identify and explore the experiences faced by mental health professionals when working in institutions seeking to offer recovery-oriented care. Therefore, our research question is: What do professionals experience in the development of recovery-oriented practices in specialized mental health care?

Method
This article is based on data from a Norwegian qualitative study on recovery-oriented practices in specialized mental health care and focuses on professionals' experiences on developing recovery-oriented practices in this context. The qualitative data is a joint product of the participants, the researchers and our relationship (Finlay 2002). As researchers, we have strived to be reflexive about our role, our professional background and our attitudes throughout the research process (Finlay 2002).

Setting
In this article, specialized mental health care refers to secondary care provided by health care specialists in psychiatric hospitals and district psychiatric centers. Treatment in these settings is provided when primary health care is no longer deemed sufficient for the patient's needs (Nymoen et al., 2020). Norwegian legislation and guidelines prioritize patient treatment in specialized mental health care in relation to the intensity and frequency of symptoms and functional level and the expected impact of treatment on the patient's quality of life (Norwegian regulations of priority between patients in health care, 2000; The Norwegian Directorate of Health 2015). Usually, specialized mental health care is considered the appropriate level of treatment if patients are assessed to have limited ability to function in daily life, have symptoms of psychosis and/or are at risk of self-harm or risk of harming others (Nymoen et al., 2020).

Recruitment and Participants
This research is based on a strategic sample (Patton 2015). We recruited institutions that reported working according to recovery-oriented principles, had employed peer-work supporters and supported medication-free treatment. Further, we emphasized the geographical spread of the institutions. Key persons from the mental health field and union representatives were contacted to identify relevant psychiatric hospitals and district psychiatric centers. For further information, we checked the websites of the institutions pointed out. There was no uniform policy on recovery orientation in the institutions. Thereafter, the managers at the selected institutions were invited to act as gatekeepers. They provided information about the study to employees in the unit and were encouraged to recruit a diverse range of informants who met the following criteria: (1) had at least one year of professional experience in specialized mental health care and (2) were therapists (e.g., psychologist or psychiatrist), milieu therapists (e.g., social worker or nurse) or other employees (e.g., music therapists, individual placement and support (IPS) workers, peerwork supporters, occupational therapists, physiotherapists, assistant nurses or carers) with clinical experience.
A total of 45 professionals were recruited from four different district psychiatric centers (DPS) and psychiatric hospitals in Norway. The sample consisted of 33 professionals from inpatient wards (general psychiatric units, medication-free units and psychosis units) and 12 professionals from outpatient clinics (Flexible Assertive Community Treatment (FACT), aftercare clinics and group psychosis clinics). The professionals primarily managed patients with schizophrenia, psychosis, bipolar disorders, severe depression and personality disorders with low functioning and high levels of symptom pressure. Of the professionals, 30 were female and 15 were male with significant variations in professions (see Table 1). Many had dual competences in the form of formal training in health or social sciences and experience of being a patient. Some of the participants also held a leadership position. The peer-support workers included were employed based on their experience as patients and were part of the treatment team. They contributed with hope, positive self-disclosure and role modeling.

Data Collection
To document the professionals' experiences of recovery-oriented practices, we conducted seven focus groups from four psychiatric hospitals and district psychiatric centers in different parts of Norway. Focus groups provide participants an opportunity to discuss and interact with views and experiences other than their own (Savin-Baden and Major 2013), have the potential to make more aspects of a phenomenon visible and are suitable for collecting data that can be used to develop practices (Lerdal and Karlsson 2009).
Inspired by Kvale and Brinkmann (2015), a semistructured topic guide was prepared in advance. When conducting the focus groups, the participants were encouraged to speak freely about their experiences and followed up by encouraging them to elaborate on the context and events and sought examples where possible. To avoid breaking the flow of the conversations, we brought notebooks to jot down themes and questions we urged the participants to explore later in the focus group.
Each focus group consisted of 4-8 participants with different professional, work and experience backgrounds and lasted between 97-108 min with an average of 104 min. The interviews were recorded on a digital audio recorder and transcribed verbatim. In this article, participants' names were changed to numbers and professions to protect their anonymity. The second number in the parentheses indicates the focus group in which the participant took part (1-1).

Data Analysis
We conducted a six stage reflexive thematic analysis Clarke 2019, 2021;Byrne, 2021). After the first stage of familiarization with the data, we generated initial codes. These codes were grouped and collated into potential main and subthemes. Next, the themes were reviewed. Some themes were refined, merged or deleted. Clear definitions and names were generated for each main and subtheme. Formulation of the themes was a long-lasting analytic process. In the last stage, the article was produced. However, writing was an integral part of the analysis, not something that only had to be completed at the end of the process (Clarke et al., 2015).

Ethical Considerations
The Norwegian Centre for Research Data (NSD, 2020/567002) gave ethical approval for this project. The project was conducted according to the principles of the Helsinki Declaration (World Medical Association, 2013). Participation was based on written informed consent. Participation was voluntary and informants could withdraw at any time without having to provide a reason.

Findings
In exploring professionals' experiences in the development of recovery-oriented practices, three themes emerged from our analysis: (a) disease-oriented structures, (b) negotiating roles and (c) risk management.

Disease-Oriented Structures
The theme ''disease-oriented structure'' dealt with how professionals related to the structures in specialized mental health care. According to the participants, the disease-oriented approach of care is central, and as with somatic medicine, diagnostic manuals are an essential aspect of provision. Traditionally, as professionals, their core tasks include: clinical assessments (based on disease history, clinical interviews, observations, and tests), the determination of a diagnosis and the provision of evidence-based treatment. This approach contrasts with a recovery-oriented approach that focuses on the patient's biography, personal meaning, growth and discovery.
A mental health inpatient ward is not basically a recovery-oriented business. To make this become Peer-support worker 5 Social worker 5 Psychiatrist 5 Other (music/occupational/physio therapist, Individual support and placement worker)

10
Total 45 recovery-oriented, you must work actively if you think that is what the inpatient ward should be. (Psychiatrist, 2-1). Participants experienced being trained to ''evaluate the other.'' They have a duty to maintain a medical record for every patient. However, they suggested that creating such notes based on their conversations and observations during the day could conflict with their understanding of recovery-oriented practices. They emphasized the importance of sharing their interpretations with patients and exploring the degree of common understanding. Among other factors, this could help to avoid pathologizing normal behavior.
We try to observe and ask the patient questions, such as, 'When I experience you like that, what do you think about it?' However, when a patient is admitted to a mental health inpatient ward, the individual is often defined without being given the opportunity to explain the whole situation (…) In the medical record, health professionals could state, 'The patient appears agitated.' It might stop there without providing more information. However, it might be a reason for the annoyed behavior, for example, a call received from the daughter that affected the patient. Then, a natural behavior in that situation is defined as something unusual and related to illness in the inpatient setting. (Nurse, 3-5).
Especially those working in inpatient specialized mental health care, stressed that professionals often interpret what they see and hear in light of the patient's diagnosis, potentially looking for confirmation of a diagnosis. However, the participants sought to meet the patients differently; they emphasized a ''recoveryoriented hearing perspective.'' According to one of the participants, this was something that patients noticed.
Some inpatients are provided with experiences that make them amazed in the way they are greeted by us, because they perceive our meetings differently in some way. They make it clear that they experience our listening to them (…) Probably, everyone thinks they listen, but there might be a difference. It could be that we hear things in different ways. So, what kind of 'hearing perspective' is recovery-oriented? I think that a recovery-oriented hearing perspective is one that tries to hear what is being said and does not try to translate this into another language. The other language is the medical psychiatric language; that's what I think. It is the traditional medical psychiatric thinking that is in opposition to a recovery-oriented perspective. (Psychiatrist, 2-1).
The participants explained that it was almost impossible to overlook illness and symptoms, as these were considered the starting point or ''admission ticket'' to treatment in specialized mental health care. Nevertheless, the participants made an active choice regarding how much time and space the illness and symptoms would need during treatment. Therefore, they downscaled the use of numerous assessment forms about the patient's symptoms and functions.
We try our best within the framework we are part of, the hospital, and in a hospital, there exist some expectations that we sometimes focus on illness. But we try very hard to focus on the [patient's] resources. (…) Rarely do we rate our patients using PANNS (Positive and Negative Syndrome Scale) or register GAF (Global Assessment of Functioning) scores in our inpatient ward. (Psychologist, 3-1).
In contrast, several underlined that their professional education had trained them in a disease-oriented rather than recovery-oriented approach. We are trained in the diagnostic system and the disease model; if the patient is provided with that diagnosis, the type of treatment automatically follows. (Psychologist, 4-4) Such training was particularly true if participants were trained as psychiatrists, psychologists, or nurses. The study of nursing is characterized by a focus on symptoms and symptom relief. (…) The recovery perspective is the opposite, located on the other outer edge. (Nurse, 1-1).
This theme shows that both professionals and the system are prepared for the disease model and that adopting recovery-oriented practices entailed downplaying aspects on which they have been trained to focus.

Negotiating Roles
Participants emphasized that adopting recovery-oriented principles in specialized mental health care implies a shift in roles, both for the professional as an expert and the patient as a passive recipient. According to the participants, when engaging in recoveryoriented practices, patients are less passive and more accountable, which in turn means that professionals take less control and responsibility for treatment.
When we move on to a practice where inpatients themselves should define what is helpful to them and where we, as professionals, are not supposed to take control, we must think differently than we used to in psychiatry (…) To perform recovery-oriented care, I need to practice not taking too much responsibility and to try to figure out a way to do that. I learn while I walk the road. Because this is quite new, we do not have anything that could support us or guidelines written down. As (far as) I know, there exists no formal education to work in this new way. (Nurse, 7-1).
As the participants had been accustomed to taking the lead in all treatment planning and processes, they found it challenging to let go. Some of the participants revealed that they had previously limited the patients' opportunities to make decisions and initiated various restrictions on the patients if they became insecure. Now, to a greater extent, they allowed patients themselves to control the process. However, this did not happen without concerns and an increased need to reflect on what they were willing to endure arose.
We must decide for ourselves how much uncertainty we are willing to face. It varies. This is something we discuss. Yes, we can agree that we must endure some uncertainty and that it is individual. However, if the patient leaves the ward and says that he or she is going to jump from the 'city bridge,' can you bear to hear it without acting? Can you bear to say, 'It's your life?' Or do you have to run after the patient? It varies as to how professionals experience such situations, and it could also vary within the same professional in different situations. (Psychiatrists, 2-1).
The participants developed partnerships with patients in which both parties brought important expertise to the planning and delivery of mental health care. They emphasized the importance of having dialogues before, during and after treatment, so that it became natural for both patients and professionals to share their experiences. Regardless of the patient's level of functioning and symptoms, they tried to meet the patients' preferences and facilitated finding a balance together.
I have patients who are seriously ill, have psychosis and who have the lowest medicine dosage. I chose to keep the low dosage of medications if I see that the patients are okay. There is no need to prescribe a higher dose if together we have found the right balance. I also have patients who are totally medication free. I mean, it is possible to be psychotic and still not use any medications. But again, we must ask, 'How is the patient?' Does the patient feel okay being psychotic? (Psychiatrist,(4)(5).
Participants highlighted that patients were unfamiliar with taking an empowered role.
Some patients come with helplessness. They just want to be helped, want us to decide, govern and take the whole responsibility. (…) They would like the recipe. But it is not so simple; it is not quite a readymade sandwich list that I can present to the patient. (Psychiatrist, 1-4).
Participants emphasized that patients need time to get used to the idea and the consequences of an empowered patient role in the treatment follow-ups. Although participants noted that most patients increased their involvement and responsibility in treatment and appreciated this new patient role, they pointed out that there were some who had been in a disempowered patient role for so long that they neither saw nor wanted to take on a different position. The participants underlined that an essential part of recovery was the extent to which a person was able and willing to take on this responsibility. If the person was unable to take ownership of his or her life, recovery would be difficult to achieve. To succeed, a collaboration with the patient is required.
Some are so shaped by the patient role that they are unable to get out of it. It brings too much resistance, self-stigmatization and problem-focusing. If it is too overwhelming, then there is difficulty finding a corner where we can start working together. (Peer-work supporter, 5-6).
Although participants highlighted that patients should be placed ''in the driver's seat'' in their treatment, they emphasized that professionals should not relinquish all responsibility or refrain from using their expertise to guide patients. Participants underlined the need to balance their professional expertise with the patient's competence in managing long-term serious mental illness.
I think about how we define professionalism. Some say that it is a not-knowing position, but I think we have an expertise, and this should not be underestimated. I think patients are provided with a feeling of safety when they know that we have an expertise. I have not experienced that they feel overwhelmed because I have expertise. I explain that I have knowledge and inform them about what I can do or what I am not able to do, as long as I balance my expertise with the expertise they have. (Psychologist,.
Developing recovery-oriented practices entails that the roles of patient and professionals are in motion and need to be renegotiated. This necessitates restructuring work for both parties.

Risk Management
The theme risk management focuses on ethics, responsibility and legislation. Professionals in wards, particularly psychiatrists and psychologists, who had an extended responsibility for treatment in specialized mental health care, were primarily concerned about the question of risk management. The vast majority of participants dealt with issues related to compulsory interventions. We must not forget that we work with people who are seriously ill, and that they can take serious, bad actions. As professionals, you cannot be passive in acute phases; you might need to act. (Psychiatrist,(4)(5). Examples of such interventions could be isolation, mandatory medication or transfer to a psychiatric intensive care unit (PICU), all of which the patient might oppose.
The participants reported many discussions in their units about when it would be ethically appropriate to intervene. They emphasized the importance of ensuring patients' integrity. However, they perceived a limit to a patient's autonomy to make her or his own decisions and their own ethical duty as professionals to protect the patient's life and health. ''How far should self-determination go? Should one be allowed to perish?'' (Psychologist, 3-4) The Norwegian Mental Health Care Act (1999) provides psychiatrists and clinical psychologists in specialized mental health statutory authority to intervene without patient consent in order to prevent injuries if lenient interventions have previously been attempted. Participants saw this legal responsibility as demanding.
It is a balancing act because you want empowerment, patient participation and decisions made by the patients themselves. At the same time, we are responsible for the legal aspects. That is quite difficult. If someone gets hurt or injured and the situation gets quite dangerous, I must take control and make decisions. (Psychologist, 5-4).
When participants had to initiate decisions that limited a patient's right to decide for her or himself, they were concerned regarding the damaged to a relationship they had built over weeks, months or years. However, they highlighted the importance of the implemented changes from 2017 in the Mental Health Care Act, which (1) increased the right of the patient to make decisions about his or her own health; (2) allowed the patient to express his or her own desires and aspirations before decisions were implemented; and (3) provided the right for the patient to evaluate implemented measures with professionals. The new legislation related to compulsory interventions contributed to protecting the patient's integrity and dignity.
It established a very good protection of integrity after coercive interventions. It involves a retrospective conversation about patient integrity and autonomy and questions about coercive interventions experienced by the individual involved, for example, 'How was it for you when I had to do that?' Previously, we did not do things like that. (Psychologist,.
In conversations about compulsory interventions, the participants stated that they took patient collaboration seriously and emphasized the importance of exploring their patients' reactions and how the compulsory interventions may have undermined their relationship.
You must talk about it afterwards. Ask what the compulsory interventions did to the relationship when I put on that (legal) hat. Before I intervened, I might have been in team meetings emphasizing that things are going well, that we must follow your goals and that it is your decision. Suddenly, I took a different position; I intervened and decided that you had to go to the hospital (PICU). What did such an intervention do to you? What did my actions do to our relationship and your trust in me? (Psychologist,(3)(4).
Participants agreed with the need to keep compulsory decisions to a minimum but that such actions were not necessarily at odds with recovery-oriented practices. Rather, participants highlighted a greater focus on why such decisions were made and how professionals approached these. They were open to agreeing that some situations could have been resolved in other ways and stressed the importance of being part of an interdisciplinary team that is interested in both seeing and understanding the patient.
A lot of the coercion used in the ward is because we as milieu therapists became scared and had a lack of understanding of our patients' struggles. (…) However, it is important to recognize that coercion can sometimes be an important and necessary tool. My point is that if we are not aware of our own understanding of the other person, then we risk exercising coercion that is not appropriate. When we strive for more interdisciplinary collaboration using a common language, then we open pathways for the patient to get better, whether it is now or later. It is an opening for getting better even though we have used coercion or had to be paternalistic during one phase of their process. This is not possible if we do not open up for the other person to come forward. (Nurse, 1-5).
When participants were in situations that required them to take control, they experienced dichotomous thinking, where it quickly became a question of whether their practice really could be recoveryoriented. Is there a possibility that we are recoveryoriented this week, but next week we are not? Sometimes we become very paternalistic where we need to take full control. (Nurse, 4-2). Although the participants underscored the goal of adopting a recovery-oriented approach, they occasionally used practices that were in conflict with the core elements of recovery thinking. Therefore, some found it more fruitful to talk about degrees of recovery-oriented practices in specialized mental health care. They had to acknowledge that this was not always possible or that they did not succeed sufficiently in applying such practices in these units Is there a clear limit to this (when we are recoveryoriented or not)? I do not experience it that way. I experience it in such a way that we constantly strive to take steps toward being as recovery-oriented as we can. Sometimes we are probably less recoveryoriented than at other times or at other points. But I think it becomes inappropriate to think so paradoxically about it -like (it is) either-or. (Psychologist, 3-1).
The participants experienced a cross-pressure between ensuring patients' integrity and autonomy and professionals' legal duty to intervene. Often, they reported situations where it was difficult to find other possibilities, and their decisions had consequences in terms of the patient's trust and relationship with the employees and the system they represented.

Toward Recovery-Oriented Specialized Mental Health Care
Our findings highlight that developing recoveryoriented practices in specialized mental health care requires a different approach to risk management and new ways of collaborating in treatment. The process of adopting a new approach and way of working creates tensions. The findings show that professionals within specialized mental health care who seek to be recovery-oriented need to balance between two concepts with strong positions: self-determination and risk management. We illuminate these concepts by discussing what is prominent when they are presented with dilemmas in specialized mental health care. It is not unique that professionals have to deal with risk management and self-determination when providing healthcare. However, it becomes particularly challenging to balance these concepts in the context of specialized mental health care. Instead of moving from professional-led to patient-led decision-making in specialized mental health care, we suggest approaching recovery-oriented practices through shared decision-making.

Self-Determination
In the mindset on which recovery orientation is based, the right to self-determination is central and highlighted as a goal. According to professionals in our study, the promotion of self-determination in people with serious mental illnesses and the professionals responsible for facilitating it were emphasized as ethically right and important. Simultaneously, professionals experienced self-determination as demanding, as it has been emphasized as a core value to a limited extent in specialized mental health care. Traditionally, people with serious mental illnesses who have received treatment in specialized mental health care, especially in wards, have not been empowered. They have learned through experience that professionals are the experts who have explained what they need, identified goals, determined the course of treatment and prepared their treatment plans (Slade 2009). Some patients have experienced that their ability to make decisions is limited in development. They have felt unsafe in making their own decisions and have been relieved when others take control.
According to professionals in our study, patients' lack of experience in having the right to make informed decisions about their health and illness in specialized mental health care makes the increased focus on self-determination even more challenging. Patients with serious mental illnesses need new experiences in exercising self-determination, and this is a skill they have to learn and develop. Although professionals underlined the importance of supporting patients in exercising self-determination, they lacked confidence and training to facilitate such processes. As they advocate a cultural shift from ''treating'' to ''learning and enabling'' , they had to break free from just having an expert role and develop professional roles in accordance with supporting patient self-determination. It is essential to provide sufficient time and effort to make this possible, as it will take time for patients who previously had limited self-determination to start making decisions for themselves.
Many patients will need support and training to be able to make the best informed choices for themselves. It is precisely this support and training that professionals can facilitate (Perkins & Repper, 2016), but it requires that they dare to let go. This is quite challenging because professionals might be unsure whether patients with serious mental illness can take on this responsibility alone, as they never had the opportunity to develop this skill. If professionals adopt a traditional understanding of self-determination as individual freedom, i.e., that the person is considered an independent and free individual to make his/her own decisions (Mackenzie 2019;Skarstad 2018), there is a danger that professionals may consider that patients with serious mental illness lack the ability to make their own decisions and rather create some ''selfdetermination-free zones'' to either protect the individual or the surrounding society. Thus, the problem becomes one in which self-determination as a human right disappears, and professionals adopt a position of power where they assign or deprive individuals of this right (Guddingsmo 2020).

Risk Management
Risk management is seen as the cornerstone of mental health care (Slemon et al., 2017). In our study, professionals clarified that working in specialized mental health care entailed clear expectations and duties for them to effectively assess and manage possible risks patients pose to themselves and others. This also involves interventions against the patients' will if they are considered to have serious mental illness, lack consent competence or pose a real danger to their own or others' lives. However, self-determination is one of the key values in recovery-oriented practices (Farkas et al., 2005) and appears to be opposite to risk management (Roychowdhury 2011).
When professionals emphasized self-determination as a significant value in the treatment, but simultaneously have had an ethical and legal duty to identify risks and take preventative action to avoid risky behavior, they experienced putting their relationship with the patient into play. They found it quite demanding to make decisions for the patient and take control of the treatment. In fact, they experienced it as a danger to the therapeutic relationship with the patient, which was previously considered as being in a position to help and support the patient in the recovery process. Even though professionals realized that making the decisions and taking control of the situation could prevent danger to the patient or his/ her surroundings, their actions could also have significant costs related to the destroyed hope and trusting relationship of those they intended to help (Perkins and Repper 2016). However, professionals tried to reduce the risk of damaging the relationship by initiating dialogues about different understandings of the situation, interventions and the relationship. Gaining understanding and recognizing each other's perspectives could make it possible to continue a trusting relationship even when disagreement of the course of treatment exists.
Risk is an everyday experience, an intrinsic part of living with mental illness and a necessary component of rehabilitation (Jones 2020). Like Perkins and Repper (2016), professionals in our study were aware that they could neither ''make someone safe'' nor ''make someone recover.'' As professionals, they had to be aware of how much uncertainty they were able to endure. They revealed that the limit differed among the professionals and was crucial in choosing to take control. It was essential that patients themselves be held accountable if they were to take responsibility for their lives. If the patients were to gain experience, professionals had no choice; they had to open up to a certain degree of risk tolerance. They had to accept that risk cannot be reduced to zero (Morgan 2004), and no risk-free course of action exists (Perkins and Repper 2016). However, professionals in our study were aware of the consequences of making incorrect assessments and that it could prevent them from letting patients decide for themselves.
The fact that the culture of blame is not given up creates challenges because many professionals are afraid of legal, organizational and professional repercussions Jones 2020;Morgan 2007). It becomes taxing to promote selfdetermination when there is low tolerance for any chance that something might go wrong. As Jones (2020) states, risk is neither dichotomous nor fixed but rather a dynamic and fluid construct that is variable and prone to external factors. In line with Slade (2009), our study shows that adopting a recoveryoriented approach in specialized mental health care created new ethical dilemmas in relation to professional accountability, particularly in the relationship between risk and self-determination.

Shared Decision-Making
To adopt a recovery-oriented approach, professionals in our study highlighted a need to involve and hold patients in specialized mental health more accountable in treatment. We suggest that shared decisionmaking could be helpful as a collaborative approach to support self-determination and increase safety among patients. Shared decision-making places the patient at the center of care and equalizes the traditionally asymmetric power relationships between patients with mental illnesses and professionals (Beyene et al., 2019). As shared decision-making is seen as an intermediate position between a clinical-led, paternalistic approach and a patient-led, informed choice approach (Chong et al., 2013), patients' informed preferences, as well as clinical evidence are both recognized to reach a mutual agreement on the best course of action (Coulter and Collins 2011). Building on a shared decision-making process in specialized mental health care can prevent dichotomous thinking where the patients are either left alone to make the decisions or, more commonly, the professionals make the decisions for the patients. As professionals in our study expressed, professionals and patients bring different but equally important forms of expertise to the decision-making process. Drawing on both clinical and patient expertise and preferences ensures a broader and more complete picture of the situation, risks and course of actions and has the potential to provide better outcomes when patients are active in managing their care and contribute to full involvement in decisions affecting them (Slade 2017).
Professionals in our study were concerned with maintaining their clinical expertise. Working with a recovery-oriented focus did not mean setting aside their clinical expertise; rather, they utilized their competence to reinforce the patient's treatment goals. One necessary part of their job was using their expertise concerning risk, while maintaining their ethical responsibility to intervene. Sometimes they experienced patients who could not take care of themselves, and as professionals, they had to take the main responsibility for a period of time. However, when the situation changed, they expected the patient to be more active and self-determined in treatment (Beyene et al., 2019).
Recovery-oriented practices call for a move away from risk management to a shared responsibility to promote safety (Perkins and Repper 2016). In our study, supportive and trusting relationships were seen as a prerequisite to promoting safety. Such relationships involved a true collaboration in working toward patients' goals and aspirations. Through open dialogues, both parties could experience a real opportunity to understand the other's world and decision basis, which made it possible to seek a course of action that accommodated both parties' agendas. This does not mean that patients and professionals always come to a shared agreement. Sometimes a professional has to intervene to ensure life and health. However, open and honest dialogues about safety before, during and after an intervention could be helpful in preserving a trusting relationship and enabling patients to regain self-control to get on with their lives and pursue their goals and aspirations. Creating a culture that addresses safety and opportunities rather than risk, allows for an easier, more productive and collaborative starting point (Perkins and Repper 2016).
A relational understanding of self-determination fits well with a shared decision-making approach. When individuals are seen as mutually dependent and inextricably linked to society (Mackenzie 2019), decision-making processes that are shared and supported by others are acknowledged. As a result, the patient's capability for self-determination is socially and situationally shaped. This could prevent situations where patients are understood as self-determined only if they are independent of others when making decisions in their own lives (Skarstad 2018). When self-determination is constituted as a human right achieved through supportive social relationships (Skarstad 2018) and not as a gift assigned by professionals, the patient's opportunities to exercise self-determination increases (Guddingsmo 2020). Thus, people with serious mental illnesses could use their skills and resources, take responsibility for what they want and explore possibilities through supportive professionals.

Concluding Remarks
This study argues that it is more fruitful to discuss degrees of recovery orientation rather than defining practices as recovery-oriented or not. It might not be realistic to eliminate coercion from specialized mental health care delivery, as professionals are legally required to protect the patients, the careers and the general public from harm. A full recovery orientation entails a massive paradigm shift in thinking, service orientation and utilization of resources. Such transformation will take time and will require at least a generation to materialize in any substantive way (Davidson et al., 2006).
This study illustrated that professionals must balance conflicting epistemologies, ethical considerations, legal dilemmas and role expectations. Professionals adopted a pragmatic approach and worked to change the system from within to enable a larger system transformation in the future. However, such efforts could be regarded as tokenism and illsustained (Roychowdhury 2011) as they sought to implement a radical concept into such a strong disease-oriented tradition as specialized mental health care. Moving practices toward a recovery-oriented paradigm where the promotion of self-determination becomes the guiding star requires that current structures, frameworks, legislation, standards for professional responsibility, guidelines and service delivery must be changed in order to embrace a new paradigm. We suggest approaching shared decision-making with a relational understanding of self-determination built on genuine and supportive relationships between patients and professionals. Finally, we propose an emphasis on safety plans instead of risk management.

Implications
This study highlights the tensions professionals experience when seeking to adopt recovery-oriented practices in specialized mental health care. Future research should focus on tensions between risk and selfdetermination. Specifically, the process toward how decisions are made and by whom should be explored. Increased knowledge about shared decision-making is needed. Further, it is important to illustrate examples where patients and professionals work together to make patients stay safe. This should not be about preventing patients from being exposed to risk, but rather how to increase patients' capacity to make good choices for themselves through open dialogues, trusting relationships and support from professionals. Similarly, risk assessment and risk management should be an open part of practice, where patients are actively involved to ensure a shared engagement with patient safety. medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.