Positive risk management: Staff perspectives in acute mental health inpatient settings

Aims: To explore inpatient staff's understanding and implementation of positive risk management. Background: Risk management is an essential skill for staff working in acute mental health inpatient settings. National policies advocate the use of positive risk management as a form of collaborative, recovery-focused risk management. However, little is known about how staff understand, operationalize, and use positive risk management in practice. Design: Qualitative reflexive thematic analysis study. Methods: The authors recruited a purposive sample of healthcare professionals working in acute inpatient settings ( N = 16) in 2019 across three National Health Service Trusts in the North-West of England. Participants completed semi-structured interviews which were analysed using reflexive thematic analysis. Results: The analysis generated three themes: (a) within staff barriers; (b)


| INTRODUC TI ON
Assessing and managing risk is essential for staff working in mental health settings (Hawley et al., 2010). Risk, defined as the 'nature, severity, imminence, frequency/duration and likelihood of harm to self or others' (Department of Health (DoH), 2009, p61), is common in mental health settings due to high rates of self-harm (James et al., 2012) and attempted and completed suicides . The Healthcare Quality Improvement Partnership (2019) reported 91 completed suicides in 2017 by service users in inpatient settings. Additionally, Iozzino et al. (2015) reported that almost one in five people admitted to a mental health inpatient ward had committed an act of violence.
The DoH (2009) describe positive risk management (PRM) as 'risk management, which improves the service user's quality of life and plans for recovery, while remaining aware of the safety needs of the service user, their carer and the public' (DoH, 2009, p9). Although definitions of PRM have differed across documentation (Just et al., submitted), policymakers generally consider PRM to be a collaborative, strengths-based, person-centred approach used to aid recovery.

| Background
Research on PRM has demonstrated benefits in terms of proactive practice (Greenhill & Whitehead, 2011), service user collaboration (Robertson & Collinson, 2011), maximizing autonomy (Hall & Duperouzel, 2011), the relevance of incorporating strengths-based approaches (Damme et al., 2017;Wylie & Griffin, 2013), and positive risk-taking (Birch et al., 2011). Further benefits of PRM aspects have included the utility of person-centred approaches (Moerman, 2012), relational security (Birch et al., 2011;MacInnes et al., 2014), shared responsibility (Kaliniecka & Shawe-Taylor, 2008), and quality-oflife improvements (Van Damme et al., 2016). Research into service user involvement and collaboration has been limited to small sample sizes (e.g., Hall & Duperouzel, 2011) and study numbers (Eidhammer et al., 2014). However, the available research suggested that including service users were beneficial. For instance, Hall and Duperouzel (2011) piloted a collaborative risk screen and found that the screen maximized participants' sense of autonomy, empowerment, and trust. Langan and Lindow (2004) suggested that collaboration aided the understanding of triggers and helped to develop a therapeutic relationship that was important in risk management. Deering et al. (2019) systematic review supported the notion of collaboration and relationships within risk management. Their results indicated that service users valued interpersonal relationships, feeling heard, and being part of the process.
These aspects were central to helpful risk management practices.
Despite the evidence for implementation, staff have not always applied PRM in practice. Research demonstrated that only 38% of outpatients (Prokešová et al., 2016) and 50% of inpatients were involved in their risk management plans (Bowers, 2011). Additionally, only around 10% of carers and family members were involved in service users' care plans (Bowers, 2011). Price et al. (2017) examined service users' views on helpful risk management strategies and found that service users reported staff favouring harm minimization approaches in escalating crises. Clifford et al. (2018) found that even when clinicians endorse the concept of PRM, in practice, they still use risk-averse approaches due to uncertainty and worry. Downes et al. (2016) results illustrated that not all mental health nurses agreed with taking planned risks. This could suggest that consensus on how to conceptualize PRM was challenging to establish among clinicians and it therefore remains unclear how clinicians understand PRM.
Research has suggested that staff-related barriers to implementation are sometimes driven by staff perceptions and concerns.
Risk management research has illustrated that staff are concerned about the therapeutic relationship within risk management (Downes et al., 2016), a lack of guidance, support (Reddington, 2017), and the ability to balance safety and quality-of-life needs (Clifford et al., 2018). Downes et al. (2016) found that mental health nurses believed that while risk assessment could aid professional decision-making, some expressed concerns that the process could be dehumanizing, while others did not agree with the role of positive risk taking within recovery. Robertson and Collinson (2011) found that staff held positive attitudes towards working collaboratively and positive risk taking, defined as the taking of planned risks that encouraged growth (Morgan, 2004), yet had concerns over the public's perception, possibly explaining why staff might disagree with positive risk taking. The drawbacks of having risk-based decisions judged by others were evident in the broader literature (Logan et al., 2011), as practitioners were worried about accountability (Reddington, 2017) and professional litigation (Carson, 1997) in the event of serious incidents. These factors were likely to contribute to guideline implementation, yet little was known about whether these factors were evident in practitioners' views of PRM. To date, there have been no studies exploring how clinicians perceive, understand, and operationalize PRM.

| Aims
The aim of the study was to explore inpatient staff's understanding and implementation of PRM.

| Design
A qualitative reflexive thematic analysis study to understand how inpatient staff implement and understand PRM.

| Sample/participants
The lead author recruited an expert purposive sample of 16 participants over 6 months (July 2019 to December 2019) via face-to-face presentations and poster advertisements distributed in eligible clinical services across three National Health Service Trusts in the North-West of England.
Participants were eligible if they were as follows: 1. currently employed as a registered mental health nurse, social worker, occupational therapist, or support worker/healthcare assistant, 2. able to converse in English proficiently, 3. working in an adult acute inpatient or Psychiatric Intensive Care Unit (PICU) service for at least 3 months.

| Data collection
The lead author screened participants who expressed an interest in participating against the eligibility criteria who then provided written informed consent. Participants completed a semi-structured qualitative interview. The topic guide (see Figure S1) included questions about the meaning, experience, barriers, facilitators, and knowledge of PRM. Participants received £10 reimbursement for their time.
The first author provided participants with information sheets and consent forms. Potentially identifying details were removed during the transcription process to ensure confidentiality and anonymity while adhering to data protection procedures. Patient and public involvement (PPI) was a key part of the study design and planning. A Community Liaison Group was consulted to ensure that community and service-user perspectives were considered in the study.

| Data analysis
The author used reflexive thematic analysis (Braun & Clarke, 2006;Braun et al., 2019) to analyse the data. This consisted of six stages: data familiarization, initial code generation based on inductive coding, identifying initial themes, reviewing themes, defining and naming themes, and producing a report.
The first author led the analysis and the second and third authors reviewed and contributed to each phase by discussing the codes, potential assumptions, distinctiveness of themes, and interpretative quality. The authors chose a semantic analytical approach to coding and theme development as this incorporated the entirety of the rich dataset. The lead author kept a reflective diary and notes from the interviews, aiding the analytic process.

| Reflexivity statement
The first author was a trainee clinical psychologist with experience in working in acute inpatient settings and managing risk. The first author had no training in PRM, which allowed them to conduct the analysis from a relatively non-expert standpoint. However, the first author engaged with relevant research literature prior to data analysis, creating the potential for bias. The second and third authors were academic and clinical practitioners, who worked with vulnerable adults with mental health difficulties and had used PRM in their clinical roles. The team reflected on the impact that this might have had on the current research, in terms of what they view as optimal practice. The authors acknowledge the potential for bias due to the lead author having engaged with relevant research literature before data analysis.

| Rigour
To ensure the analysis was sufficiently rigorous and minimized the potential for bias, the research team conducted frequent discussions regarding the coding of data, potential assumptions, and possible omissions. The lead author led the analysis with support from the research team. Initially, all authors coded the same data extract and compared the codes generated to ensure interpretations were derived from the data. The authors discussed the initial themes to consider the relationship between codes and themes, the different levels of themes, and how they related to the research question. The authors' refined themes further by considering internal homogeneity of the themes and external heterogeneity across themes. Additionally, to assess the methodological quality, the lead author completed the Critical Appraisal Skills Programme (CASP, 2018) checklist for the present study.

| FINDING S
The lead author completed 16 face-to-face semi-structured interviews lasting between 16 and 100 min (mean 63 min, Table 1).
One participant requested to end the interview after 16 min but consented to the inclusion of their data in the analysis. Participants were from three National Health Service Trusts in the North-West of England (N = 1, 9, 6). All interviews were audio-recorded and transcribed verbatim by the first author.
Thematic analysis generated three key themes and nine subthemes relating to participants understanding and implementation of PRM (Figure 1). The authors used extracts from the transcripts to demonstrate the interpretative adequacy of the analysis, with further examples available in Table 2.

| Theme 1: within staff barriers
Theme 1 refers to staff being selective in their use of PRM based on their own beliefs, biases, and influences. Staff did not feel able to apply PRM for all service users under all circumstances consistently and were influenced by several factors. Some staff recognized these influences, while others were unaware of the impact these had on their risk-based decisions.

| Beliefs-based practice
Beliefs about mental health dictated staffs' use and practice of PRM.
Practitioners used diagnoses to determine risk and suitability for PRM. Most practitioners considered diagnoses as part of the risk management strategy. Some practitioners held beliefs that people's behaviour was a direct result of their mental health. Staff believed it was valid to use diagnoses to determine risk and that some diagnoses were riskier than others: 'their diagnosis can impact on the behaviours they engage in, so it can impact how someone's risk might be different to others' (16).
In contrast, a few practitioners did not hold beliefs about risk and diagnostic constructs. It appeared that by using diagnosis as a starting point, some practitioners were able to understand the service user's point of view. It was, therefore, apparent that there was a contrast in practitioners using diagnosis to determine risk and suitability for PRM, while other practitioners used diagnoses to inform a holistic understanding of service users: 'She (sigh) in her EUPD she gets quite feelings of abandonment, you know, neglect, you know she feels disempowered with services, she doesn't feel they're supporting her …. She doesn't always see the positive sides of what how she's managing the risk' (10).

| Emotional decision-making
Practitioners' anxieties and fears influenced their use of PRM.
Staff were concerned about the prospect of accountability in the event of adverse outcomes. The psychological burden of making a wrong decision felt overwhelming and influenced all other decisions. Thirteen of 16 participants talked about stress and anxiety related to making risk-based decisions. Additionally, eight participants discussed the personal impact that making such decisions had on them, such as a lack of sleep, difficulties in switching off, and agitation. Practitioners were often preoccupied with the worst possible outcome, which was twofold: harm to the service user and repercussions for staff: 'Maybe that's more self-preservation for the person who is entrusted with managing those risks …. I guess that might come from a selfish point of view' (5).

| The impact of experience
Participants' experience and seniority affected the practice of PRM.
Junior staff talked about being unsure what PRM was, yet were keen to implement it, while complacency was a concern for experienced staff. All newly qualified staff spoke about feeling unsure how to use PRM. They stated that there was no training provided on risk management which required staff to learn on the job or from other people's approaches and being reliant on their own initiative. Newly qualified staff highlighted the need to look at different perspectives and remain compassionate towards service users, while acknowledging that fading empathy could occur when managing risk daily: 'you got to make sure you don't lose your compassion …. I think, sometimes when you're exposed to risk so often maybe it's possible to become desensitised' (1).
A hierarchical system was evident which newly qualified staff found hard to challenge. New staff did not feel empowered to challenge established processes and senior clinicians, despite acknowledging the importance of doing so. Newly qualified staff felt intimidated by more senior staff, which meant that they stood back and watched improper practice take place, feeling unable to speak up. This example illustrates how a newly qualified practitioner struggled with this dilemma when helping a more senior colleague take vital readings: 'I step back because they've been here longer than me …. even though I think it's wrong, I go with what they say …. But I didn't have the confidence to tell her ….
because I'm only a baby, I shouldn't tell her that …. I let them deal with it …. The way they take risks, the way they speak to people is wrong' (9).
The above quote illustrates how practitioners held the belief that they could not challenge improper practice. Not being able to challenge practice was evident for many practitioners; all newly qualified staff and two other participants described how extensive experience could result in complacency and risk aversive practices, which they struggled to challenge.

| Theme 2: Within service user barriers
Staff considered specific service user factors to influence PRM implementation. These included staff perceptions of service users' level of risk and trustworthiness. Given the potential consequences of serious incidents occurring, staff described needing to make judgements about whether a service user was suitable for PRM.

| Determining PRM suitability
Staff regarded the use of PRM being dependant on service users' level of presenting risk. Staff described PRM as a tool that aided personal recovery, although this could only be possible for service users with minimal or reduced levels of risk. During crises, staff associated PRM with adverse outcomes (e.g., harm to self or others); therefore, staff said that they would wait until the risk reduced before attempting PRM. Staff would typically only implement PRM following a harm minimization approach. For example, this quote illustrates how staff only felt able to take positive risks, such as going out on leave, after a person had been 'stabilised', rather than as a tool to stabilize service users: 'If someone is stabilised … can we maybe look at giving them back such a thing, can we look at maybe giving them some escorted leave …. as their mental health gets stabilised you can …. the risk is less' (5).
When asked to describe how the participant might stabilize risk, they explained: 'I suppose it's when we manage when they're going to attack us and we calm them down' (9). The quote illustrates that staff's perceptions of threat from service users, or during crises, affected whether collaborative approaches, such as PRM, were considered suitable. It is noteworthy to mention that many practitioners felt that they were constantly managing crises.

| Trusting the untrusted
Staff talked about PRM being a matter of trust between practitioners and service users. However, they also reported how difficult it could be to trust service users, potentially creating an impossible prerequisite for PRM. Of the 12 participants who spoke about trust, 11 reported feeling unable to trust service users, some to any degree: 'As far as patients go, I don't trust patients' (3). Staff talked about being deceived by service users to retain or acquire access to leave or specific items. They reported not being able to trust service users because the potential for negative consequences was too great: 'I mean they can say to you yeah, I'm fine, I've got no thoughts to hurt myself …. And you don't know what they're going to do .… they could decide to go and hurt themselves' (14).
Despite reporting a lack of trust of service users, six practitioners said they used PRM to demonstrate that they did trust them. Staff reported that trusting service users enabled their work to be recovery-focused, which in turn increased autonomy and quality of life. They described how trust provided the basis for a therapeutic relationship where staff took positive risks more readily. The following quote shows the incremental nature of this relationship, as positive risk increased so did trust: 'I feel that the more positive sort of risks you take ….
The more trust is gained' (7).
It was important to note that of the six practitioners who used PRM to demonstrate to service users that they trusted them, five also struggled to form a trusting relationship. Practitioners appeared conflicted, not trusting service users to take positive risks but also knowing that positive risk taking was essential to recovery. One participant illustrated this dynamic eloquently when talking about trust and the related challenges: 'I think you have to accept that and give people the chance to try things and do things on their own' (14).

| The need for support and verification
Staff perceived successful PRM implementation as reliant on a sense of shared responsibility and being able to verify their decisions with senior colleagues. Shared responsibility was closely linked to concerns over accountability. Practitioners felt anxious about risk and felt the need to discuss and consider alternative perspectives with colleagues. The need was, in part, due to the uncertainty around risk management, as staff could never be sure of outcomes, yet felt responsible for them.
Staff were nearly unanimous, with 14 of 16 participants stating that being supported by their colleagues had been essential to practicing PRM. They believed that PRM needed to be a team approach.
The following quote depicts how helpful staff found being able to create a shared view: 'I can literally ring or email the X (control and restraint) team or the safe wards team and ask them for advice and they'll come down and be involved in care plans' (11).
The participant not only valued the advice but also the practical involvement of others and suggested that lone risk-based decisions felt too challenging and overwhelming, yet with the support of others, that uncertainty had become more manageable.
Staff described a hierarchical structure in services where certain viewpoints were more valuable than others. This related to practitioners' need to discuss concerns with senior colleagues for a sense of shared responsibility and to manage their anxieties, feeling more accountable if senior colleagues were not involved: 'when I did the safe self-harm care plan, you know I wasn't a charge nurse then, so the charge nurse supported me …. And then X (ward manager) was on board, the consultant was on board' (10).

| Collaborative process
The second subtheme pertains to engaging service users and their carers throughout the risk management process. Fifteen practitioners spoke of the value of involving service users in their risk management plans, as this had ramifications on service users' quality of life, feelings of empowerment, levels of autonomy, and motivation. Staff reported that collaborative working ensured that service users felt empowered, as they had opportunities to take control of their lives and make decisions for themselves, better equipping them for life in the community. Practitioners also explained how PRM was easier and less demanding when done collaboratively: 'giving them a bit of a say. Because I think when people feel more in control, they're more likely to engage with us' (13).
Working in partnership with service users was seen as key to fostering independence and a positive self-identity. This practitioner talked about a service user who required substantial support, but who staff also needed to provide with a sense of autonomy and control: 'he still really likes being involved with it as much as he can …. he still likes to be able to like chop his own mushrooms and do that, but he might need a bit of supervision with like appliances …. And I think it's just part of your identity' (1).

| The multiple perspective-taking predicament
Staff talked about collaborative care being important but finding this very hard to put into practice. Staff described collaborative risk management as consisting of service users being compliant with plans that they had created. Participants described viewing PRM from their perspectives, struggling to consider others' frames of references: 'it's not collaborative. Them are the decisions we're making …. you don't like work and ask what would work for you …. We go at them with this is how it is, this is what we do' (5).
Where staff involved family members of service users and other practitioners, this had been to gather information. Practitioners were keen to involve others and found this valuable. However, due to the demands placed on them, practitioners struggled to consider different ways to incorporate service users and their carers into the process of PRM. There was an implicit assumption that their decisions, based on their own perspectives, were appropriate for others: 'It's quite empowering in the sense that you're making a decision for someone, but you know you're doing it because you feel it's going to benefit them' (8).
Risk-based decisions were usually made by clinical teams as opposed to service users, as illustrated by the following participant talking about whether to grant a service user leave: 'the team kind of discussed and … we felt that at that moment in time it wasn't appropriate' (16).
Practitioners were often keen to involve service users and acknowledged that they did not use PRM enough. It was evident that staff felt conflicted, wishing to provide a service user-led collaborative risk management process, but struggling to implement this in daily practice.

| Competing demands
The fourth subtheme refers to the time pressures that staff faced when trying to use PRM, including conflicting demands and time constraints. Staff discussed how other demands took precedence over PRM. Twelve of 16 participants identified lack of time as the reason that they did not implement PRM: 'you might kind of like take … less positive risks …. or you might avoid risk altogether …because you think I've not got time to make it safe, so I'll just avoid it for the moment' (1).
Staff regarded PRM as requiring sufficient time to implement and they were less likely to put it into practice when time was limited.
Practitioners' time constraints were also related to their fears and worries of accountability, thus reporting that PRM would be unsafe without giving it the necessary time and attention it required.
Eight participants identified demands that took priority over PRM, including medication, service targets, new referrals, and paperwork, resulting in staff not fully implementing PRM initiatives.
For instance, one participant developed information packs for staff about individual service users, including their likes, dislikes, and hobbies, intending to provide a holistic understanding of the service user to inform risk management: 'But it never took off. Because we didn't ever get that started. But it's a really good pack, it's in the office ….
But they were really good …. They never took off' (10).  (Carson, 1997;Reddington, 2017), staff worried about accountability, or investigations, in the event of a serious incident. Our results also echo previous work on the emotional impact on staff when working with service users who are suicidal, with staff reporting fear of being blamed (Awenat et al., 2017). Clifford et al. (2018 note that the ability to balance safety and quality of life needs was a barrier to guideline implementation. The current research supports this, as often staff were well-intending, however, reverted to a harm minimization approach when faced with conflicting demands.

| D ISCUSS I ON
The results provide insight into why risk management planning is not always collaborative between staff and service users (Bowers, 2011), despite research demonstrating its benefits (Greenhill & Whitehead, 2011). A hierarchical system influenced by staff experience, seniority, and emotions was evident. Newer staff members did not feel able to voice their concerns regarding current practice, increasing the potential for malpractice and the development of a culture where mistakes were not addressed (Mannion et al., 2019).
The current findings suggest that staff are keen to use PRM but may not fully understand and implement it in practice.
Furthermore, staff do not routinely receive any formalized PRM training and subsequently find it challenging to navigate risk-based decisions. Their own experiences predominantly influence staff decision-making and they commonly feel uncertain about engaging in PRM. Previous research indicates that lack of training is a barrier to guideline implementation (Fischer et al., 2016). As training increases competency (McNiel et al., 2008), it is essential to provide PRM training to new members of staff to increase their practice knowledge and confidence. More experienced staff would also benefit from regular refresher training, in line with national guidelines (DoH, 2009) to reduce the potential for complacency in risk management practices, which the current study indicates as a possible barrier to PRM. Previous findings suggest that training effects are unlikely to continue unless support is available to staff (Haberlin, 2012).
A key implication of the current research is the importance of support from colleagues and organizations. Training alone is unlikely to be sufficient in altering practices (Richards et al., 2004 2

Supervision and reflective spaces
Managing risk places a huge emotional burden on practitioners. Practitioners are keen to do the right thing. However, when harm occurs, practitioners require a supportive environment which encourages future PRM use. Moreover, if staff can reflect on their attitudes and factors influencing decision-making, it is more likely that they feel able to employ PRM. Supervision and reflective spaces, therefore, need to be developed with the underlying assumption that risk is interpreted differently, and open discussions of these differences are encouraged to foster understanding and implementation. Supervision and reflective spaces are thus paramount for effective PRM. 3

Culture change
The implementation of any new practice includes becoming aware of which unhelpful approaches are being used and feeling able to challenge these. This requires an organizational culture change. Having a mixture of newly qualified and more experienced staff is likely to help. However, staff need to feel supported and confident to challenge existing practice. Encouraging a culture that embodies the need to consider staff well-being, the importance of relationships, and staff feeling supported is vital for the successful implementation of PRM.

Consistency across documents
If policies and guidelines are not consistent in their message of what PRM is and how staff can implement it, staff are unlikely to feel confident and able to implement it in practice. It is therefore crucial for documents to be clear, consistent, and explicit in their message, making it more likely that practitioners will implement guidelines.

5
Balancing demands Results indicate that across documents and practitioners, challenges are evident due to the need to balance challenging demands. Little guidance is provided on how practitioners can achieve the balances as this is likely person specific. However, only providing staff with awareness does not result in practical changes. Therefore, staff need experience of managing the complexities, as well as support from their peers, teams, and organizations.

6
Relationships are paramount Services have become more concerned with relational security, illustrating the importance of a good therapeutic relationship between staff and service users. The current findings support the need to prioritize relational security and provide further impetus for services to focus on this aspect of risk management.

7
Service user led Arguably one of the most crucial findings of the results are the challenges that practitioners face in being service user led, which are reflected in national policies and guidelines. There is an awareness that risk cannot be eliminated and that PRM is suitable for all service users under all circumstances. However, documents and practitioners face challenges in implementation. Staff need to feel supported to discuss their concerns for collaborative working, as well as needing to re-prioritize collaborative working as a mandatory requirement of PRM. (Hubbard et al., 2017). Encouraging a culture that embodies the need to consider staff well-being, the importance of relationships and helping staff to feel supported, is vital for the successful implementation of PRM. For a full list of clinical implications, please see Table 3.

| Limitations
The recruitment ended after the lead author completed and transcribed 16 interviews, bringing a potential for 'premature closure' (i.e., not having sufficient data to identify meaningful themes; Connelly & Peltzer, 2016). The lead author, who had a thorough understanding of the data, discussed potential codes and themes regularly with the research team ensuring that potential data omissions were unlikely.
Once the authors completed data analysis, they identified the generated themes as being valid based on the codes and their application across the data set. However, one participant discussed broader global system implications, which could have been explored further.
The limited number of participants in specific professional groups prohibited an analysis of differences in beliefs between professional roles. It would, therefore, be useful for future research to analyse the importance of broader systems and differences between professional groups. It is possible that there may have been a selection bias, where participants with certain attitudes towards risk management were overrepresented in the sample, which may be a limitation of this research.

| CON CLUS ION
In conclusion, inpatient staff expressed a firm intention to provide person-centred risk management processes. However, PRM often brought them into discord with themselves, other people, and perceived systemic barriers. Results illustrate the need for ongoing training and support for inpatient staff when practicing PRM safely and in line with national guidelines.

ACK N OWLED G EM ENTS
We are grateful to the participants of this study who took time from their busy clinical roles to share their views and experiences.

CO N FLI C T O F I NTE R E S T
No conflict of interest has been declared by the authors.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/jan.14752.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.