Aims

To systematically review and synthesise qualitative literature on the experiences of service users with complex emotional needs (CEN) of community mental health care, and their views about what constitutes good quality care.

Search strategy and selection criteria

The CRD handbook guidance (https://www.york.ac.uk/media/crd/Systematic_Reviews.pdf) and the PRISMA reporting guidelines were followed [18, 19]. The protocol was prospectively registered on PROSPERO (CRD42019142728). The present review was part of the NIHR Mental Health Policy Research Unit’s work programme on CEN, which included four systematic reviews (alongside the current review are reviews of qualitative studies of clinician experiences, quantitative studies of service outcomes, and economic evidence of cost-effectiveness). The protocol for the wider programme of work was also registered on PROSPERO (CRD42019131834). A single search strategy was used for the whole programme, and articles relevant to each review retrieved from the resulting pool of papers. The protocol was developed by the review team in collaboration with a working group of lived-experience researchers and subject experts.

Searches of MEDLINE (January 2003—December 2019), Embase (January 2003—December 2019), HMIC (January 2003 –December 2019), Social Policy and Practice (January 2003 –December 2019), CINAHL (January 2003—December 2019) and ASSIA (January 2003—January 2019) were conducted. The search strategy was supplemented with forward and backward citation searches of included articles. An additional search of EMBASE and MEDLINE (January 2003-November 2019) was performed to identify related systematic reviews, and the reference lists of relevant reviews were checked. Grey literature was identified through web searches and through searches of the above bibliographic databases. The full search strategy was peer reviewed using the PRESS checklist prior to searching and is available [20], including a search narrative [21], in the (S1 Text in S1 File).

Citations retrieved during searches were collated in Endnote and duplicates were removed [22]. As a single search strategy was used for a wider programme of work, initially titles and abstracts were independently double screened for all reviews simultaneously. Full text screening was then performed for citations that were potentially eligible for this review by AE and LSR. All papers thought to meet inclusion criteria and 20% of ineligible papers were double screened. In cases of disagreement or uncertainty, consensus was achieved through discussion with senior reviewers (SJ and SO).

We included primary research studies published since 2003, when “Personality Disorder: No Longer A Diagnosis Of Exclusion” was published [6], as papers that are older than this may be less relevant to current needs. No limits were placed on the language or location of publications. Eligible studies were those that:

Included recognised qualitative data collection and analysis methods. Written data from questionnaires were included if a recognised qualitative analysis method was used such as thematic analysis. Mixed-method studies were included if the qualitative data were reported separately to the quantitative data.

Reported data from adults (aged 16 or over) with a “personality disorder” diagnosis. We also considered for inclusion papers focusing on care provided for complex emotional needs described as repeated self-harm, suicide attempts, complex trauma or complex PTSD, and emotional dysregulation or instability: we made this decision as we were aware that otherwise some papers may be missed because authors and/or participants are reluctant to use the term “personality disorder” for the reasons outlined in the introduction. When the above search terms resulted in retrieval of potentially relevant papers, a group including a senior psychiatrist reviewed study context, inclusion criteria, and sample description to assess whether the majority of the sample fitted the clinical picture associated with “personality disorder”.

for such studies we considered in each case whether the sample appeared to consist mainly of people with long-term difficulties similar to those that may result in a “personality disorder” diagnosis.

Data extracted for this review related to care provided by community based mental health services, including primary mental health care services, generic community mental health teams, and specialist services for people with complex emotional needs. Data related to care from residential, forensic, crisis services, or from specialist services for different conditions, such as substance misuse clinics, were excluded.

A more in-depth description of the eligibility criteria is contained in the (S2 Text in S1 File).

Quality assessment and analysis

Data on the key characteristics of eligible studies were independently extracted by two reviewers (AE and LSR) using an Excel-based form. Quality assessment of included papers was performed using the Critical Appraisal Skills Programme (CASP) Qualitative Checklist by two researchers [23]. Any discrepancies were resolved through discussion. Study quality was not used in decisions about eligibility but is reported and incorporated into the meta-synthesis.

Data were analysed using thematic synthesis [24]. In the first stage, preliminary codes were developed, focusing on themes relevant to understanding service user views about what constitutes good care. To do this, all relevant material was coded as examples of poor practice or care also provide information, implicitly, of good care. Two researchers inductively line-by-line coded 10 articles each and a third researcher independently second coded 50% of these. Codes were then compared and discussed between researchers until an initial set of codes was developed. The remaining articles were then divided between the three researchers for coding. New codes were added as necessary. In stage 2, an initial thematic framework was developed. Through discussion, a team of five researchers explored similarities and differences between the codes, and individual codes were split or merged as necessary. This team comprised academic and lived experience researchers. Codes were then grouped and arranged into a hierarchy to create a framework of descriptive themes. This was an iterative process involving meetings and discussion by email, and checking the framework against the original data. In the third stage, analytic themes were generated, and the framework was finalised by the research team. Towards the end of this process, the analysis was discussed with the project working group to guide interpretation of the final results. The working group was made up of 29 members with academic, lived experience, and clinical backgrounds.

a) The need for a long-term perspective on treatment journeys

Many studies emphasised the need for a long-term perspective on treatment, supporting gradual improvement over several or many years. Service users tended to report that, although treatment benefits could accrue over time, difficulties in managing emotions and in relationships and daily living also fluctuated, so that progress made was rarely linear.

b) The need for individualised and holistic care

Service users in many studies emphasised the importance of individualised care and of availability of different types of help, rather than clinicians adopting a one-size-fits-all approach. This was reported to be a problem when clinicians focused too much on diagnosis or relied too heavily on delivering recommended and highly standardised therapies, such as DBT. The importance of taking a holistic view of needs, and of focusing on personal goals and aspirations was recurrently reported. In some cases, service users reportedly felt that therapies which were the main treatments offered did not address past traumas or the problems they were struggling with in their daily lives, and instead focused almost exclusively on specific topics such as self-harm or relationships. This could leave service users frustrated that they could not address other issues that were equally or more important to them, or that help in managing the social difficulties and the challenges of everyday life was unavailable. It was also important to service users that therapy helped them adapt skills learnt in treatment to their own personal situations.

c) Large variations in accessibility and quality of mental health services

Many of the included papers focused on specialist “personality disorder” services, and many positive experiences were described of these. Services were recurrently reported to be most helpful when they were accessible and easy to understand, when staff were knowledgeable and warm, where service users were involved in their own care, such as their Care Plan Approach meetings, and where they had good access to high quality services that could offer treatment options well suited to their needs. However, there were also many accounts of complicated journeys through services and of large variations in access to and quality of care, with accounts of good care from generic mental health services being much less common.

d) The centrality of therapeutic relationships

Good client-clinician relationships were described in many sources as being at the centre of good care. Positive qualities for clinicians included being warm, trustworthy, honest, open, accepting, non-judgemental, and interested in their job and in the service user as a person. It was important to service users that they felt supported, valued, understood, listened to, and cared about. Where this was absent, service users felt that they could not be honest with their clinician, that the clinician would be unable to help them, or that their treatment would be poorly tailored to their needs. Positive qualities were described in many papers as being more frequent among clinicians working in specialist services, potentially as a result of good training and understanding of CEN.

Problematic qualities for clinicians included being poorly informed, misinformed, or perpetuating stigmatising attitudes and therapeutic nihilism (an inappropriately pessimistic view regarding the potential benefits of treatment) [72] about CEN. Some sources identified underlying problems as lack of training in working with people with CEN, poor empathy, and understanding, or a perception of people with CEN as “difficult”. There were accounts of clinicians who seemed uninterested in people with CEN, who rushed through their appointments, or were dismissive, unsympathetic, or insensitive. Other negative characteristics included being overly strict, authoritarian, critical, superior, cold, or aloof. Sources often identified such experiences as most frequent in primary care, psychiatric outpatient settings or generic secondary mental health teams. When severe, service users described these experiences as traumatising. Further difficulties reported in generic settings included a lack of consistent relationships with the same professionals and a sense that clinicians had no clear therapeutic plans.

e) Impacts of "personality disorder” diagnosis

Whilst our main aim was to understand service user views on what constitutes good care, many papers also included discussions of the pros and cons of receiving a diagnosis of “personality disorder", including its impacts on the care they received from services. Negative consequences appeared especially prominent outside specialist “personality disorder” services. These included being excluded from services and treatments because of the diagnosis and being met with stigmatising or stereotypical attitudes amongst clinicians and society. There were reports that once labelled in this way, service users were no longer seen as unwell or distressed but as “difficult”, and that some symptoms that they experienced, including psychotic symptoms, were no longer considered genuine. For some, the contested and uncertain nature of the diagnosis made it more difficult to feel in control of their condition because there were so many myths, misinformation, and derogatory attitudes, including amongst their clinicians. Such views include that "personality disorders” are untreatable, that self-harming and other behaviours are merely manipulations to gain attention, and that service users with the diagnosis are liars, attention-seeking, unreasonable or difficult, manipulative, and take resources from other patients. Such attitudes only compounded service users’ feelings of isolation, marginalisation, abandonment, or rejection. Others felt the diagnosis pathologised the impacts of the abuses they had experienced throughout their lives, resulting further trauma and a sense of victimhood.

However, some positive effects on mental health care associated with a receiving a diagnosis of “personality disorder” were also described, especially regarding access to treatment and improved self-understanding. Some papers described service users finding that the diagnosis helped them to reflect on their own feelings and responses, and to engage in treatment. This was especially true if the diagnosis seemed to fit their experiences, and if it was contextualised with helpful information about the condition and treatment options. Where it was accompanied by access to potential helpful therapy, there were reports of the diagnosis offering a sense of validation and relief.

How service users were told about their diagnosis seemed to influence how they subsequently felt about it. Being given the diagnosis by a clinician who understood the condition, who had time for discussion, and who was optimistic about the effectiveness of treatment and the likelihood of recovery was more likely to result in a positive experience. Attempts by clinicians to avoid or sidestep a diagnosis of "personality disorder" were seen as counterproductive by some, inadvertently indicating clinicians’ negative attitudes about the condition and, possibly, invalidating service users’ hopes and understanding of themselves.

Main findings and implications

We found a substantial literature (47 papers) published since 2003, from which a generally consistent set of themes regarding experiences of care emerged. Some overall points regarding implications for achieving good practice can be drawn based on these. Firstly, reports of good practice and helpful treatment, as far as available, seemed to be largely confined to periods of care by services specialising in care for people with a “personality disorder” diagnosis, such as the multidisciplinary teams established for this purpose in parts of England [2]. However, specialist care was often hard to access and time-limited in a way that does not fit with the long journey towards “recovery” described by service users with CEN. Service user accounts across many papers suggest that care pathways are needed that take into account the long timescales involved in living with CEN, and the many set-backs often experienced. Holistic support from empathetic professionals with a good understanding of CEN is needed even during periods when service users are not engaged in intensive therapies. Transitions between stages of care need to be smooth and well-understood by all. In the care of conditions such as psychosis and bipolar, models such as early intervention and assertive outreach services and recovery teams have been developed to meet a range of service user needs over a long timescale. Development and implementation of such models for people with CEN has been much more limited even though this group likewise have long-term and fluctuating needs in many areas of their lives.

Secondly, good relationships and skilled support from clinicians who convey hope regarding long—term improvement in CEN are seen as central throughout pathways through the mental health care system. Service users tend to value highly clinicians who have the right skills to create safe spaces in individual and group treatment and manage exploration of challenging topics such as self-harm and trauma. Across the included studies, clinicians with the necessary skills and values seemed to be mainly found in specialist services, with at times appalling descriptions of lack of understanding and hopefulness, and stigmatising attitudes and behaviour elsewhere in primary and secondary mental health care. Addressing this appears central to achieving good practice, whether by designing pathways so that people with CEN normally receive care throughout from people with some specialist understanding of their condition, and/or by large-scale programmes to improve attitudes towards and understanding of CEN across the healthcare system. Stepped care models, in which some service users receive CEN interventions within generic services, are often advocated as a way of meeting needs of people with CEN across the care system. The success of such models is likely to require attitudes to CEN in generic services to be much more positive than has tended to be described in the current review [73]. Given the centrality of therapeutic relationships in recovery, service users also often advocated for a choice of therapist, but this seemed to be offered relatively rarely. Peer support appears, from the studies, reviewed to be an area with considerable scope for innovation: the mutual acceptance and understanding and sense of belonging available from peers is experienced as very helpful and validating, especially as loneliness appears to be a core difficulty in CEN. There were few accounts of harnessing this potential beyond therapeutic groups, although it seems to be a significant potential component of good practice [74].

Thirdly, as with other longer-term mental health conditions, care needs to meets a range of psychological, social and physical needs. Yet, service users reported that support from specialist services often focused mostly on self-harm and emotional regulation, with people who did not feel ready to focus on these issues or who had other care priorities sometimes excluded from care. Thus, we suggest that achieving good practice should involve designing holistic services that offer not only specific therapies (which are often highly structured and focused) but also support people with social and practical difficulties, looking after physical health, managing substance use, and with managing relationships and reducing loneliness.

Finally, being given a “personality disorder” diagnosis can have profound effects on all aspects of service users’ experiences with mental health care services. On the one hand, the diagnosis was sometimes described as helpful in contextualizing distressing symptoms, especially if it was communicated in a sensitive manner, and could allow service users access to specialist care. However, it was also clear that stigmatising attitudes held by clinicians could be detrimental to a service user’s sense of self-worth and ultimately impede their recovery. Such stigma was most common in generic mental health services and primary care, and could lead to pessimism amongst clinicians about the prospects of recovery and consequently to service users being denied access to care and treatment. These findings indicate a need for improved training for clinicians outside specialist ‘personality disorder’ services.

Throughout the included studies, there appeared to be frequent mismatch between service users’ clear assessments of their needs for long-term engagement with holistic care delivered by empathetic clinicians with a realistic but hopeful understanding of CEN, and the service contacts they experienced. This reinforces the need to include people with lived experience of CEN and of using services, as well as their families and friends, in the development and assessment of services and care pathways. The perspectives of clinicians, investigated in an accompanying review [75], as well as those of family and friends, also need to be understood to ensure plans to improve care are feasible, and to develop approaches to reducing stigma and improving understanding and attitudes. Furthermore, we note also that trials of therapeutic interventions to date have tended to focus on testing effects of specific therapies over relatively short durations [76]. The results of this study make clear that service user views of good practice tend more to focus on access to a broad and individualised treatment that lasts over the long term.

Limitations

The process of concisely synthesising findings across many qualitative studies from different dates and countries inevitably leads to some loss of nuance and simplification of findings, while allowing cross-validation between studies regarding themes that are recurrent in different populations. There was considerable heterogeneity regarding participant characteristics, treatment type, and methods, but neither the reporting of data in most papers nor our approach of looking for commonalities between papers allowed us to identify differences by groups. Papers varied in inclusion criteria, with some samples primarily of people with “emotionally unstable” or “borderline personality disorder” diagnosis, others of mixed samples. However, even the latter seemed primarily focused on the difficulties of emotional regulation and impulsive behaviour that may lead to a “borderline personality disorder” diagnosis. Generalisability to their experiences of people with other “personality disorder” diagnoses is thus limited. Further, we planned to include studies that did not describe their sample as having a ‘personality disorder’ but were nevertheless considered by a senior psychiatrist in the research team to fit the diagnosis. We considered this to be important for the reasons described in our introduction and inclusion criteria. However, as shown in Table 2, almost all of the included papers specified that the included samples had received a ‘personality disorder’ diagnosis, self-identified as having the diagnosis, or were service users of specialist "personality disorder" services. This makes our work more clearly congruent with other research in this area, but does mean that our process may not have captured papers investigating a similar population in which the diagnosis was not explicit.

To exclude papers written regarding service systems very different from current ones, we only identified papers written since 2003, the year “Personality Disorder: No Longer A Diagnosis Of Exclusion” was published [9], so that potentially important experiences before that date will have been missed. Although eligible for inclusion, we did not find papers written in languages other than English. Most of the papers were from the UK, so that our results, as well as the perspective of the authors, may result in a disproportionate focus on the UK. However, we found that similar experiences tended to be reported in the various included countries as well as across the timespan of our study. However, except for China, all included countries were higher income, and most samples were either largely White or did not report ethnic background. Thus we could not assess the relevance of our findings to these wider global and societal contexts. This a key task for further research, especially given preliminary evidence that attitudes, help-seeking, patterns of difficulties and identity in this clinical group may be shaped in important ways by cultural context [77].

Participants in the included studies were mainly people who were to some extent engaged with services. Thus, the experiences of potentially the most dissatisfied and underserved group, people who are not engaged with any kind of care, are likely to be under-represented. We did not include in our searches the perspectives of family members and friends who support service users: there is evidence that the burden they experience may be substantial, but that they often report exclusion from decision making and support: capturing their perspectives in the qualitative literature will thus also be important in further research [78].

Lived experience commentaries

The importance of providing individualised and holistic care instead of a “one-size-fits-all” approach in community services for people with CEN cannot be emphasised enough. We are not defined by a “Personality Disorder” label, but should be respected and treated as the unique human beings that we are.

It is evident that the stigma of the diagnosis is still insidious—especially amongst staff in generic community mental health services. This is extremely disappointing to see 17 years after “Personality Disorders” were officially declared “No Longer a Diagnosis of Exclusion”, which may be indicative of a culture resistant to change.

Unfortunately, it strongly resonates with some of our own lived experience that having to work with clinicians outside specialist services who demonstrate no real understanding of or empathy and respect for people with CEN often does more (iatrogenic) harm than good. In fact, hardly anything could be more re-traumatising than blatant ‘malignant alienation’, [79] which in any other context would be considered unthinkable.

In order to tackle such entrenched attitudes, we need a culture shift across community services. Mandatory CEN-specific training for clinicians should be co-produced with service users and embed helpful features of specialist services as well as trauma-informed care. However, any learning can only be successfully implemented in practice if it is consistently reinforced through role-modelling.

Overall, this meta-synthesis highlights a desperate need for change in order to provide the right care at the right time in more inclusive mental health services for people with CEN. Parity of esteem between services for CEN and other SMIs—where pathways are much better established and the importance of long-term support is widely recognised–is long overdue. Ultimately, we cannot afford to waste another 17 years without genuine progress towards treating people with CEN with the dignity and respect that they deserve.

Eva Broeckelmann and Jessica Russell

Two decades of research tell us that interventions need to float an individual’s boat. The boats are ideally equipped for a long voyage, sail at their own pace, choose their destination and have kind, skilled staff on board.

This isn’t big, clever or new. New research concurs with research written two decades ago. Why aren’t we fixing those boats?

Evaluations demonstrated a positive change in negative attitudes and stigma amongst staff after attending co-designed and co-delivered KUF (Knowledge & Understanding Framework) training. Funding has since been cut.

Survivor led organisation Emergence CIC developed innovative ways of working that were co-delivered or led by survivors. The lack of inclusion of co-produced work within the review demonstrates a missing literature base. Survivor knowledge has been decimated alongside funding cuts.

The question isn’t about what to do, or how to do it. The question is why aren’t we?

Why keep sabotaging the boats we already know we need?

Tamar Jeynes