Subtheme 1.1: impact of limited care options and capacity

Whether service users felt accepted or rejected when presenting to crisis care was a salient part of their experience. Specialist crisis care for people with CEN was rarely described, and there were many descriptions of service users being referred to community services that they did not qualify for or that did not meet their needs.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Morris, Smith and Alwin22,Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} In a crisis, a loss of ability to understand or articulate feelings is common, further compounding the difficulty for service users in expressing complex and changing needs.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10,Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} Service users reported that they needed crisis services to be accessible to enable them to seek help before the crisis becomes very severe.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Helleman, Goossens, Kaasenbrood and van Achterberg10,Reference Dunne and Rogers21,Reference Morris, Smith and Alwin22} When service users felt listened to and felt they were able to trust clinicians, conversations could be key to overcoming a crisis.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10} During the brief admission intervention, early conversations with a nurse in which the nurse was active and the service user felt listened to helped to create a sense of safety and reduce service users’ levels of emotion.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10}

Crisis teams and other alternative care options were seen as valuable,^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3} yet A&E was perceived by family carers, professionals and service users as often the ‘only option’ or ‘last resort’ during a crisis.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} When community resources or alternative crisis services were inaccessible or unavailable, A&E became the only option for care.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Vandyk, Bentz, Bissonette and Cater6,Reference Dunne and Rogers21} Service users said that they felt forced to use A&E when their community worker was away, when the crisis was too severe for community care or they were on a long waitlist for alternative care.^{Reference Vandyk, Bentz, Bissonette and Cater6} Staff in a variety of settings felt that their own service environment was not suitable for this patient group, with both A&E staff and general practitioners reporting that people with CEN would be better served by a service other than their own.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} There was the perception among staff that there were inadequate resources for this group of people, or that current services were inadequate for them.^{Reference Commons Treloar27} Staff reported that could feel uncertain as to how to respond to very dysregulated service users, and that conflicts could arise among treating teams when there were different options for how to proceed with care.^{Reference Commons Treloar27}

Although A&E often seemed to be the most readily accessible crisis care service, A&E staff reported facing tremendous difficulties managing CEN and caring for these service users. They described lacking the resources to deal with high levels of distress, as the A&E environment can be overburdened and chaotic.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} The behaviour of some people with CEN was perceived as sometimes challenging in A&E, and thought to be a reaction to long waits,^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} limited freedom or poor treatment from staff.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10} Comorbid substance misuse was also a challenge. A&E staff reported that they lacked confidence in caring for people with comorbid problems such as substance use, which can be common in service users with CEN (Table 2).^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Vandyk, Bentz, Bissonette and Cater6}

Table 2 Summary of themes and subthemes, with illustrative quotes

A&E, Accident and Emergency.

Because of these issues with suitability and accessibility of services, service users reported that they often felt help-seeking was ‘futile’. Professionals echoed this view, saying that for service users with CEN there is a chronic and frustrating ‘back and forth’ cycle of repeated emergency department use.^{Reference Vandyk, Bentz, Bissonette and Cater6,Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23}

The degree to which service users found care to be accepting or rejecting appeared to vary according to the setting. In the papers that explored experiences of brief admission, service users reported that knowing that the brief admission was accessible to them if needed made them feel safer when in the community.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} In contrast, A&E, although accessible, was considered a ‘last resort’.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23}

Subtheme 1.2: carers experience rejection and lack of engagement from healthcare systems

Family carers reported that their support needs were not considered adequately, their requests went unanswered, their relative's diagnosis and needs tended to be poorly communicated to them,^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Dunne and Rogers21} and professionals often refused or were reluctant to involve them in treatment.^{Reference Dunne and Rogers21} Carers sought out psychoeducation on their own and sometimes acknowledged that their understanding of CEN was poor, yet they felt they were expected to make important treatment decisions in the face of contradictory and inconsistent advice.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3} When asked about the possible reasons behind their exclusion from care at the time of a crisis, carers felt that staff were unaware of the critical role they played in the service user's life.^{Reference Dunne and Rogers21} A collaborative relationship with professionals was reported to be preferable for carers, providing greater communication and guidance on how to manage crises.^{Reference Dunne and Rogers21} Carers also wanted staff to recognise them as individuals outside of the caring role, and be more considerate of how the caring role affects all aspects of their life.^{Reference Dunne and Rogers21}

Carers faced many practical issues, including financial strain resulting from the impact of the caring role on their ability to work. Resources to support carers were seen as inadequate, and emotional support was felt to be lacking.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Dunne and Rogers21} Some carers described experiencing a ‘parallel crisis’ with their own distress when the service user was self-harming or attempting suicide.^{Reference Dunne and Rogers21} If carers experience these painful crises repeatedly with no support offered, the caring role can become so painful and overwhelming that they see no choice but to walk away.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Dunne and Rogers21}

Subtheme 2.1: developing and maintaining the therapeutic relationship

Clinician characteristics appeared to have a large influence on crisis care quality. There is a strong convergence among service users and carers about the interpersonal qualities they valued. Across the board, non-judgemental and validating support was the most helpful aspect of the therapeutic relationship for both service users and carers.^{Reference Vandyk, Bentz, Bissonette and Cater6} Service users emphasised the importance of openness, friendliness and warmth when presenting to crisis care.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10,Reference Barr, Jewell, Townsend and Grenyer24} Service users also wanted staff to be genuinely engaged and interested in them, to feel seen and heard, and be treated with dignity ‘like a person’ instead of a ‘patient’.^{Reference Morris, Smith and Alwin22,Reference Borschmann, Barrett, Hellier, Byford, Henderson and Rose29} Staff who consistently exhibited these characteristics were able to build a trusting relationship with the service user.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} Strategies such as telling jokes, making small talk and participating in informal conversation helped put service users at ease and in a better position to trust clinicians.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10}

Service users were most encouraged by staff who took a recovery-focused, optimistic and hopeful approach. Staff who were non-judgemental, normalising and gave service users a ‘chance to recover’ rather than emphasising the perceived chronic or ‘life-long’ nature of CEN or their diagnosis were highly valued.^{Reference Borschmann, Trevillion, Henderson, Rose, Szmukler and Moran9,Reference Morris, Smith and Alwin22} Service users also discussed the importance of being viewed as an individual by the staff rather than just another person ‘moving through the system’. Flexible and holistic care was lauded as the ideal by service users, but many reported that they received treatment that was overly standardised and did not take their individual needs or history into account.^{Reference Burke, Kells, Flynn and Joyce26} Staff recognised that interactions with service users, particularly when service users were very emotionally dysregulated, could bring up uncomfortable emotions in themselves, including a sense of frustration and helplessness.^{Reference Burke, Kells, Flynn and Joyce26,Reference Commons Treloar27} In some staff this resulted in reflection and a request for more training, and in some it resulted in the expression of stigmatising beliefs that people with the diagnosis of borderline personality disorder were wasting health service time.^{Reference Commons Treloar27} Staff that participated in a workshop considering clinician connections with service users reported that a greater awareness of their own internal states and the impact of their own emotions on the interaction was helpful in developing good therapeutic relationships.^{Reference Burke, Kells, Flynn and Joyce26}

Subtheme 2.2: setting and negotiating a framework for treatment

Interpersonal relationships with carers and staff can either greatly relieve or greatly add to the service user's suffering in a crisis. Service users and staff both stressed the importance of negotiating a clear framework for care. For example, for people using a brief admission intervention, care was most helpful if information on how and when they could contact nurses were provided early on and communicated clearly to service users.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10} Being consistent and non-judgemental, maintaining professional boundaries and communicating clearly were considered to be among the most important staff characteristics. Yet many service users and carers experienced poor communication from healthcare staff, which hindered the development of a trusting relationship.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Morris, Smith and Alwin22}

Although staff understood and endorsed needs for sensitive, clear and considerate communication, some staff reported finding it difficult to strike a balance between having boundaries and being empathetic.^{Reference Burke, Kells, Flynn and Joyce26} Some staff were concerned they would be reinforcing ‘bad behaviour’ from service users by being validating and empathetic.^{Reference Burke, Kells, Flynn and Joyce26} Boundaries between service users were even harder to enforce and were less defined. Some service users became overinvolved in the care of others and then felt responsible for conflict de-escalation, which was inappropriate to them as they were meant to be focusing on their own recovery.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10}

The negotiation of responsibility in a collaborative relationship appeared to vary by setting. During a brief admission, participants described benefit from being helped to take responsibility for their care,^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} by avoiding losing control and coercion. This was in contrast to in A&E departments, where staff's perception that the person with CEN could take control of their illness could drive negative attitudes and frustration.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} Service users reported experiencing A&E as noisy, chaotic and stressful, and often felt unsafe.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} Many service users expressed a preference for access to alternative crisis spaces, such as cafes, respite homes or rehabilitation centres.^{Reference Barr, Jewell, Townsend and Grenyer24} Service users also expressed the need for more freedom and autonomy in their care, but this was rarely provided in A&E.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} As a means to help address this issue, some service users endorsed the development of joint crisis plans, which detailed how they wanted to be treated by staff in a crisis and helped them feel more in control and satisfied with their care.^{Reference Borschmann, Trevillion, Henderson, Rose, Szmukler and Moran9,Reference Helleman, Goossens, Kaasenbrood and van Achterberg10}

Subtheme 2.3: challenging interpersonal interactions and negative staff attitudes can exacerbate presenting problems

Service users reported many experiences of marginalisation, stigma, discrimination and negative attitudes among healthcare staff.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Vandyk, Bentz, Bissonette and Cater6} Common experiences were staff becoming dismissive and reluctant to engage, or having their care withdrawn completely based on their diagnosis or when presenting with problems such as self-harm.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Vandyk, Bentz, Bissonette and Cater6,Reference Morris, Smith and Alwin22} Service users felt they had been deprioritised for treatment or ‘put in the back of the line’ once staff discovered their diagnosis. Service users who had been identified and labelled as a ‘frequent flyer’, or someone who often presents to A&E, reported particular experiences of discrimination.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23}

Although some A&E staff did feel sympathy toward patient distress, service user background heavily influenced this.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} There was an attitude among staff that patients with CEN were time-consuming^{Reference Commons Treloar27} and taking time away from patients with acute medical problems, and this was felt to be particularly frustrating if they had a history of repeat admissions.^{Reference Vandyk, Bentz, Bissonette and Cater6,Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} Other problematic attitudes from staff included a belief that service users with CEN are ‘just choosing to be difficult’,^{Reference Vandyk, Bentz, Bissonette and Cater6} they are ‘manipulative’,^{Reference Commons Treloar27} ‘attention-seeking’ and ‘using the system’ in a way that it was not intended.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} Service users also reported issues of diagnostic overshadowing and feeling ‘defined’ by their diagnosis.^{Reference Morris, Smith and Alwin22} Many service users reported apprehension about receiving a diagnosis because of stigma, feeling fearful that a permanent label would lead to a curtailing of freedoms or loss of control, leading them to conceal or minimise their presenting problems.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3} Staff also reported that once a person was given the diagnosis of borderline personality disorder this could lead to a loss of ‘objectivity’ in staff assessments of the person.^{Reference Commons Treloar27}

Negative experiences elevated emotional distress and reinforced low self-worth with service users. They described feelings of rejection, abandonment, shame, isolation, helplessness and guilt after presenting to crisis care, especially A&E.^{Reference Vandyk, Bentz, Bissonette and Cater6,Reference Helleman, Goossens, Kaasenbrood and van Achterberg10,Reference Burke, Kells, Flynn and Joyce26} Interactions with staff did not have to be overtly hostile to be negative; feeling ignored, misunderstood or dismissed also worsened negative emotions in the service user.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10} Service users were less inclined to seek help after a negative experience with crisis care to avoid feelings of shame.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} This is potentially dangerous as future suicide attempts may be more resolute to avoid readmission.^{Reference Vandyk, Bentz, Bissonette and Cater6}

Subtheme 3.1: transitioning with a clear recovery plan

Service users explained that having a preventative strategy in place and knowledge of effective crisis resources provided a sense of security and reassurance after discharge.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10,Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} Service users were helped by the knowledge that they could access care when needed, be provided with support to continue any intensive therapies, and could stop future crises from getting worse with a quicker return to baseline functioning in daily life.^{Reference Helleman, Goossens, Kaasenbrood and van Achterberg10,Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} However, many service users reported that links to out-of-hospital support were limited, particularly if they were usually high-functioning before the crisis occurred.^{Reference Vandyk, Bentz, Bissonette and Cater6} This lack of access to continuing support meant that these service users had difficulty continuing in their recovery, leaving them more vulnerable to further mental health crises.^{Reference Vandyk, Bentz, Bissonette and Cater6}

Carers reported that discharge from emergency services was often sudden, confusing and unexplained.^{Reference Barr, Jewell, Townsend and Grenyer24} Often the service user was sent home without a clear treatment plan or appropriate support in place.^{Reference Barr, Jewell, Townsend and Grenyer24} Thus, carers felt their involvement in treatment was limited, as the plans for discharge were often not communicated to them despite their importance to recovery.^{Reference Lohman, Whiteman, Yeomans, Cherico and Christ3,Reference Barr, Jewell, Townsend and Grenyer24}

Subtheme 3.2: Negotiating collaboration and service user responsibility

Service users wanted opportunities to self-refer for support and maintain some autonomy when in crisis; for example, being able to self-refer to a safe place when feeling suicidal, where hospital admission or police involvement would be unlikely to ensue.^{Reference Vandyk, Bentz, Bissonette and Cater6} Despite these desires to take more responsibility for their care, service users often experienced a lack of autonomy, particularly in A&E and in-patient settings.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} They described experiences of constant supervision and coercive measures, which undermined their ability to take responsibility for their own treatment and recovery.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} More novel approaches, such as self-referral for brief admission, may help combat this, as service users who had experienced both brief admissions and regular admissions reported that they were encouraged to take an increased amount of responsibility during the brief admission.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} These service users explained that they embraced and appreciated the requirement to be more involved in their care planning.^{Reference Eckerström, Flyckt, Carlborg, Jayaram-Lindström and Perseius25} Advanced crisis plans were identified as one way that people with CEN can have more control over their own care during a crisis, in which participants coproduced crisis plans during periods of stability.^{Reference Borschmann, Trevillion, Henderson, Rose, Szmukler and Moran9} The importance of individualisation of care was highlighted, as participants showed a wide range of different preferences, and the majority took the opportunity to include refusals for specific treatments, such as medication and involuntary detention.^{Reference Borschmann, Trevillion, Henderson, Rose, Szmukler and Moran9}

Subtheme 4.1: training, knowledge and confidence

Clinicians at all levels of experience across services (community mental health teams, A&E, brief admission) reported lacking confidence in treating CEN, expressing that they often felt ‘helpless’ and ‘incompetent’ interacting with service users presenting with emotional dysregulation.^{Reference Burke, Kells, Flynn and Joyce26,Reference Commons Treloar27} Treating psychological distress successfully was a major area of concern, as clinicians felt their knowledge and skills were lacking in supporting service users during emotional crises.^{Reference Burke, Kells, Flynn and Joyce26} Clinicians reflected that uncomfortable emotions could be brought up by interacting with very dysregulated service users, and staff could often feel unsure of how to respond.^{Reference Commons Treloar27} This lack of confidence and understanding often led to frustration among healthcare staff, as they did not know why these service users expressed intense emotions and challenging behaviour.^{Reference Burke, Kells, Flynn and Joyce26}

Service users and carers thought professionals needed to be more educated about CEN, as knowledge was seen as a central influence on staff attitudes toward people with a diagnosis of personality disorder.^{Reference Vandyk, Bentz, Bissonette and Cater6,Reference Dunne and Rogers21} Misunderstandings could lead to negative beliefs and attitudes, which is damaging to service users seeking help. In contrast, further training around CEN helped change staff perceptions of service users, so that they were able to take a more positive view of their interactions.^{Reference Burke, Kells, Flynn and Joyce26} Professionals agreed that there was a major need for further training and opportunities to practice necessary skills for working with these populations, and most were keen to take part in this.^{Reference Burke, Kells, Flynn and Joyce26}

Sub-theme 4.2: emotional support and boosting morale

Many professionals discussed difficulty in staying calm and regulating their own emotions when working with people in crisis. Burnout and high work stress were reported as significant risks when working with service users with CEN.^{Reference Burke, Kells, Flynn and Joyce26} This was exacerbated if staff did not have the proper skills and training to know how to support these service users in crisis situations.^{Reference Burke, Kells, Flynn and Joyce26} This is especially true for A&E staff, who often work in an already overburdened environment with competing demands on their time that may preclude the opportunity for support.^{Reference Spence, Bergmans, Strike, Links, Ball and Rhodes23} Staff agreed that increased communication, good team support from peers at work and clinical supervision^{Reference Commons Treloar27} would help ease their workload burden, improve patient care and increase their confidence, whereas poor and unsupportive workplace cultures contribute to burnout and poor practice.^{Reference Burke, Kells, Flynn and Joyce26}