Cross-national comparative mixed-methods case study of recovery-focused mental health care planning and co-ordination: Collaborative Care Planning Project (COCAPP)

Phase 1: literature and policy review and synthesis

Phase 2: case studies

In Phase 2, we conducted six in-depth case-study investigations34 across six contrasting NHS trust/health board case-study sites in England (n = 4) and Wales (n = 2) (meso-level) employing mixed quantitative and qualitative methods. Then, in each site, access was secured to a single Community Mental Health Team (CMHT) from which up to six service users, their care co-ordinators and informal carers were sampled as embedded micro-level case studies. 31 Qualitative data were generated related to care planning and co-ordination processes in each (Figure 2).

FIGURE 2.

Diagram of study plan.

Preparation of the data

Data from the questionnaires were entered into SPSS package, version 21 (Armonk, NY, USA). The data were checked and cleaned by a second researcher prior to statistical analysis. The distribution of the questionnaire data was assessed for normality by exploring the data graphically. Comprehensive sensitivity analyses were completed in order to determine what parameters to use when dealing with missing data. The service-user version of the RSA scale questionnaire in particular had a moderate number of missing data and therefore the parameters that were used for calculating the subscales were based on 50% completion levels.

Exploring the data

Descriptive statistics were calculated for the three questionnaires (the RSA scale, STAR-P and ES). The total scores, subscale means and SDs were derived to produce a ‘recovery profile’ for each site. Where appropriate, these scores were compared against reference values (STAR-P and ES) or the participant groups (the RSA scale). Some further detailed analysis at a descriptive level was completed on the primary outcome scale (the RSA scale) to aid with the triangulation of the qualitative and quantitative data. This was completed at an individual item level on the scale by ranking the mean responses for each question to determine where the most agreement was for the participant groups. The top five items were selected from the questionnaire and presented as a recovery profile for the site.

Inferential statistics

Several one-way analyses of variance (ANOVAs) were conducted to compare differences between the six sites on the RSA scale, STAR-P and ES measures. Subsequent Tukey post-hoc tests were conducted to ascertain which measures differed between which locations. A series of one-way analyses of covariance (ANCOVAs) were completed to adjust the analyses for potential confounders. The variables that were chosen for service users were: age; gender; ethnicity; relationship status and time in mental health services. The variables that were chosen for staff were: age, gender, ethnicity, time working in mental health services and time as a care co-ordinator. The criteria for adjusted analysis between the ANOVA and ANCOVA were the p-value from the omnibus test, the adjusted means and the p-value from the post-hoc test. If the p-value from the omnibus test for the ANCOVAs were not substantively different from the ANOVAs then no further post-hoc analyses were completed.

Correlations

Correlations were carried out to identify if there was a relationship between the outcome measures and to determine if there were relationships among the patients on recovery-oriented focus, empowerment and the quality of therapeutic relationship. Three Pearson’s correlations were completed on the mean total scores for the measures RSA scale and STAR-P, RSA scale and ES, and STAR-P and ES for all participants and by individual site. Cohen’s effect sizes were used to describe the data (small r = 0.10, medium r = 0.30, and large r = 0.50).

For all analyses the significance level was set at a level of 0.05.

Introduction

It has long been recognised that patients with a variety of chronic and complex health conditions often require long-term care from different health and social care professionals working across community and hospital settings. However, those living with multiple health and social care needs often experience a highly fragmented service, leading to suboptimal care experiences, outcomes and costs. 54,55 Many countries have developed strategies to enable better co-ordination of care; however, evidence suggests that these have often not achieved their objectives. 56

The aim of this literature review is to give an account of care co-ordination and care planning in CMH settings. The question underpinning the review is: ‘What interventions have proved more or less effective in promoting personalised recovery-oriented care co-ordination for CMH service users?’. The focus of the review is care co-ordination in UK CMH contexts, but some research from non-UK settings will also be discussed.

Background and context setting

High-profile failures in mental health care in the UK during the 1980s led to an unprecedented evaluation of care co-ordination between hospital and community services to ensure better quality care. The most immediate trigger was the recommendations of the Spokes Inquiry57 into the killing of a hospital social worker by a psychiatric patient. At the time of the inquiry, care for people with severe mental illness in the community had been described as haphazard and unco-ordinated,58 a view reinforced within the Spokes report,57 which described how, prior to the killing, the patient had been able to ‘drop out of sight’ of mental health services whenever she was discharged from hospital. This, allied to fears about high levels of mental illness among the homeless and in the criminal justice system,59 led to the introduction of a raft of strategies and policies to improve the organisation and delivery of mental health care in the late 1980s to early 1990s.

However, the deterioration in standards of mental health care evidenced in the 1980s can be seen as a culmination of public policy decisions over previous decades. This included a reduction in inpatient mental health beds in the UK from 150,000 in the early 1950s to approximately 50,000 in the early 1990s,55 and the medicalisation and professionalisation of mental disorder, which were key themes of the Royal Commission on the Law Relating to Mental Illness and Mental Deficiency. The subsequent Percy Commission Report60 concluded that mental disorder should be regarded ‘in much the same way as physical illness and disability’ (paragraph 5) and that mental hospitals should be run as much as possible like hospitals for physical disorders. In addition, the 1959 Mental Health Act61 separated health and social care for people who did not need inpatient treatment by handing over responsibility for social care to LAs and councils.

The shifts described above created ideal incubating conditions for subsequent failures in care co-ordination that became increasingly evident in the 1980s. Such was the disarray that some commentators at the time described CMH care in England as an ‘unwieldy dinosaur with its health and social care brains working independently’. 62^{(p. 2)}

In the context of an accelerating, policy-driven shift away from hospital care, the CPA was introduced in England in 1991 and in Wales in 2004 to provide case management and to give shape and coherence to the delivery of CMH services. Case management is a method of working designed to ensure that service users are provided with services that are co-ordinated, effective and efficient. However, the CPA was not developed with a particular model of case management in mind and therefore lacked a single or coherent underpinning philosophy of care. 58 As a result, the introduction of the CPA was very much shaped by the local context, and differences in local approaches were tolerated provided that the fundamental features of the CPA were implemented.

The fundamental features included systematic assessments of health and social care needs; the provision and regular review of a written care plan; close monitoring and co-ordination by a named key worker; the involvement of users and carers in planning and provision of care; and inter-professional and inter-agency collaboration. A CPA register was also established to record details of those cared for under the CPA.

The role of the care co-ordinator and effective teamwork were identified at the time of the CPA’s inception as key to successful implementation,63 given the previous repeated failures of agencies and professionals to communicate and successfully deliver co-ordinated care.

Search strategy

Initial literature searches were undertaken using the following key words and terms: ‘mental health’, ‘care planning’, ‘care co-ordination’ (and ‘co-ordination’), ‘collaborative care’, ‘recovery’, ‘recovery focus(ed)’, personali*. JE ran a preliminary search from which a random sample of articles was assessed by BH and MC, to identify relevant papers and possible additional search terms/phrases. Further discussions were carried out by BH, AJ and JE on the modification of the search strategy.

As a result, additional key words/search terms were included: ‘mental illness’, ‘care collaboration’, ‘patient care planning’ and ‘person-centred care’. We also included proximity indicators (such as ADJ or N- as appropriate of each database), truncation ($) and wildcard (*) symbols as well as Boolean commands (AND and OR) where appropriate. Key search terms were searched by their subject (MeSH headings) and by keyword. The following databases were searched: Applied Social Sciences Index and Abstracts (ASSIA), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Allied and Complementary Medicine Database, EMBASE, The Cochrane Library, MEDLINE, PsycINFO, Education Resources Information Center, British Humanities Index, Scopus, Social Care Online and Web of Science. The search was limited to the period 1990 to date and included articles in the English language only. This search was rerun on the databases ASSIA and CINAHL and the search strategy was verified by a health and social care librarian working for the Information Sciences service at Cardiff University.

During a meeting on the 20 November 2012 with the LEAG and the PAG, it was suggested that the term ‘user experience’ be included in the search strategy. Further discussions among the PAG advised that the addition of terms such as ‘recovery’ or ‘recovery focused’ to the search strategy would narrow the focus too much and that research covering these topics in CMH settings should be captured by using the existing search terms. The issue of rejecting papers where research was considered low quality was weighed against an interest in a broad representation of approaches and views. However, some studies were excluded on the grounds of insufficient detail about the research process undertaken within each study. 64,65

Following removal of duplicate cases, 811 references were retrieved and entered into an EndNote version X7 (Thomson Reuters, New York, NY, USA) library. These references were then screened by BH and MC to identify key papers for the meta-narrative synthesis focusing on papers about care planning and co-ordination in mental health in the community. The papers were labelled as Y (Yes), N (No) and M (Maybe). In the end, there were 45 papers labelled Y, 617 labelled N, and 94 labelled M. There was an agreement among the team on the papers excluded. A further snowball search on the web and using Google Scholar (Google, Mountain View, CA, USA) produced 81 references. From this there were none labelled Y, 69 labelled N and 12 labelled M. A final review of the M papers and back-chaining revealed a further three papers that were added to the original 45, giving a final total of 48.

Meta-narrative review

Meta-narrative review looks historically at how particular research traditions have unfolded over time and shaped the kind of questions being asked and the methods used to answer them. As outlined by Greenhalgh and Wong, a research tradition is a ‘series of linked studies, each building on what has gone before and taking place within a coherent paradigm (that is, within a shared set of assumptions and preferred methodological approach shared by a group of scientists)’. 66^{(p. 4)}

Given the wider social, political and historical context outlined above it is unsurprising that the ‘unfolding plot of the research’67^{(p. 423)} from the UK is dominated by researchers’ efforts to understand whether or not the CPA has led to changes in management, service organisation and delivery (Tradition 1). Although our searches were limited to the period 1990 onwards (when the CPA came in), there is little history of research into care planning and co-ordination within CMH settings prior to this date. There are no pre- and post-evaluation studies of the introduction of the CPA, comparing changes, if any, in care planning and co-ordination in CMH settings.

Instead, researchers initially framed their studies within the findings of public inquiries or governmental reviews into difficulties and failures within CMH settings. It seems reasonable to conclude, therefore, that the CPA resulted in changes both to working practices within clinical CMH settings, and also to research priorities and practices within the UK, creating a hitherto unseen tradition of research within mental health services with a focus on care planning and co-ordination.

Another narrative and tradition of research focuses on service-users’ and carers’ experiences of CMH provision (Tradition 2). Although its emergence may not be surprising given that one of the primary aims of the CPA was to increase the involvement of service users and carers in care planning and provision, this focus was rare in the context of the early to mid-1990s. Another notable feature of this tradition is the emphasis on involving service users and carers in the design and execution of research projects68–71 at a time where both government policy promoting service user involvement and the ‘service user movement’ were in their infancy. 72,73 It is within this tradition that the current study most comfortably sits.

Service-user involvement within this tradition includes involvement in some aspects of conventional research projects (such as question setting or data collection) as well as collaborative research where service users work on most if not all aspects of projects as co-investigators alongside academic researchers. There is less evidence of there being user-controlled research in which service users set the agenda, design and conduct the research. 73

Finally, several studies have sought to determine whether interventions have improved the functioning and performance of the CPA (Tradition 3) in terms of improved care co-ordination and care planning. This tradition of research focuses on processes and outcomes of care, and the prevailing language positions the CPA and CMH work as being driven by requirements to demonstrate organisational efficiency. For example, some studies frame mental health work as requiring standardisation owing to its complex nature and the mental health workforce as having a deficit of knowledge as regards effective care co-ordination and planning.

The three research traditions identified are summarised in Table 2.

Striking in the review was the increase in non-UK research into care planning and co-ordination in community settings with only two studies appearing before 2005, at which point the number increases and surpasses UK research output. These non-UK studies were also aligned with the three research traditions described above but are not included here due to space restrictions.

Tradition 1: community mental health care co-ordination and planning, the care programme approach and the organisation, management and delivery of services

The first tradition of research consists of a series of studies primarily from the 1990s that seek to understand the impact of the CPA on the organisation and delivery of CMH services. This tradition of research draws on a range of methodological approaches (see Appendix 4). The CPA is often discussed as a uniform approach, yet it is noticeable that the earliest study of the CPA68 undertaken in three English health districts demonstrates clear variation in CPA interpretation and implementation. Adherence to the principles of care programming within each district’s community care plans ranged from minimal mention of care programmes to the wholesale adoption of the concept into mental health planning and service evaluation. Later published studies also draw attention to the degree of variation in the implementation of the CPA between different trusts within the same health authority and within the same trust. 16,74–78

Variable implementation of the CPA may well contribute to inconsistency in research findings within and across sites. For example, Schneider’s68 study of three health authorities describes how staff perceived the CPA as offering ways of working that were both creative and flexible as well as rigid and lacking flexibility. Similarly, more rigorous documentation and better care planning are reported in some studies,68,79,80 whereas others report a lack of coverage of psychosocial aspects16,77 or risk assessments not completed fully or jointly. 76 Claims that the CPA led to improved continuity of care,68 better team-working across professions16,68,79 and overall effectiveness of care68,70,81,82 also need to be balanced with findings that most general practitioners (GPs) had not heard of the CPA,83 that the role and function of the CPA had not been explained properly to staff,16,70,83 leading to managers not knowing which patients were on the CPA. 81

Uncertainty also exists in terms of the relationship between the CPA and inpatient bed occupancy rates, with Tyrer et al. ’s62 claims of increased admission to hospital following the introduction of the CPA being contradicted elsewhere. 82 The use of hospitalisation as an outcome measure has been criticised as inadequate for assessing programme success in this population, given that rates of hospital admission and length of inpatient stay can be influenced by other factors, such as local service configuration and bed availability. 84

Such uncertainty and variability raise crucial questions regarding the use and effectiveness of the CPA in practice, especially where researchers attempt to make judgements on programme outcome or cause–effect comparisons across more than one research site. The lack of clarity regarding the exact nature of the CPA in practice reflects a similar vagueness globally about the concept of care co-ordination. For example, a review of research found over 40 heterogeneous definitions and models of care co-ordination to be in existence. 85 The lack of clarity and homogeneity has also been identified as an explanatory factor for the failure of many strategies that seek to improve care co-ordination. 56

Sociopolitical factors partly explain the large variations in the way organisations adapted or assimilated the CPA’s requirements. For example, the CPA was introduced with no or limited additional resources,68,83,86 at a time when health and social care spending was being cut and the vast majority of resources for mental health services were allocated to inpatient care rather than day and community services. 87 As a result, little or no staff training was provided, no particular philosophy of care emerged to underpin CPA implementation and employees were left to manage a change in process as best they could. 58

In an era of increased managerialism in the NHS, the introduction of the CPA was experienced by practitioners as a ‘top-down’ policy imposition that failed to build on the existing knowledge, skills and abilities of the workforce, and resulted in tensions over clinical values and cultures. Staff perceived the CPA as leading to more work16,68,82,83 and a sense of being overworked16,68 especially owing to increasing levels of bureaucracy. 16,68,71 Staff also reported that the additional burden of work that accompanied the implementation and day-to-day operation of the CPA led to increased time being spent away from patients. 16,68,83

The unintended consequence of a lack of detailed, national policy implementation strategy for the CPA was a considerable variation in the delivery of services, as well as variation in staff and patient experiences both within and across health authorities. A review of CPA implementation links the ‘paradoxical effect’ (p. 24) of burgeoning levels of local bureaucracy to the national policy decision not to be overly prescriptive about CPA documentation. 1

Overall, this tradition of research has been firmly focused on developing a better understanding of how the introduction of the CPA influenced the organisation and delivery of nascent and arguably inadequate CMH services. Researchers frame the CPA as a policy intervention that attempted to reverse several decades of deteriorating mental health services characterised by, among other things, inadequate co-ordination and organisation of care within and across different professional groups. There is little research that seeks to delineate or explore developments in multiprofessional and multiagency working in the wake of CPA implementation; studies undertaken in this tradition perpetuate the status quo of multiprofessional working at that time, failing to mention or merely hinting at different ways of working within and across teams. However, Simpson’s16 study describes some of the effects of team working in CMHTs, such as finding that care co-ordination was enhanced within teams when members demonstrated respect for co-workers.

The combination of three decades of profound changes in the policy definition of mental health and illness, together with the broader context of large-scale change towards managerialism within the health and social care sector during the late 1980s and early 1990s, leaves the impression that the CPA was destined to fail before it had been launched.

Tradition 2: service users’ and carers’ experiences of community mental health care co-ordination and planning and their involvement in research

Until the 1990s there were few attempts to involve service users in the planning and delivery of their care and treatment. As Perkins and Repper88 point out, service users were excluded from service planning meetings and were only involved via demonstrating their symptoms and hearing the doctor’s prescription. Implementation of the CPA was fairly novel in responding to service user and carer demands for greater involvement in the care planning process, both of which began to be more explicitly promoted as indicators of good practice. 80

Alongside these developments, a view also emerged that service user involvement in research could make an important contribution to the empowerment of mental health service users. For example, service users have long argued that dominant research approaches to mental illness can perpetuate patients’ inequality and disempowerment. 89 Consequently, the coproduction of new knowledge and the transformation of the terms and concepts used by mental health researchers have been promoted as a potentially influential means to achieve broader social and political change. 90

A tradition of undertaking research that involves service users and carers is rooted in the origins of research into care co-ordination and care planning in CMH work (see Appendix 5). Early studies not only focused on service user and carer experiences of CMH but also involved service users and carers as collaborators in the design and implementation of research studies. For example, Carpenter and Sbaraini80 involved service users and a carer in a PAG and in formulating a questionnaire to explore users’ perceptions of the extent to which the CPA empowered user and/or carer involvement. Rose et al. 71 not only involved service users in designing a data collection instrument (semistructured interview schedule) but also extended service user involvement to data collection, with 12 service users being trained to undertake interviews alongside researchers.

Interestingly, Rose et al. 71 describe how, on many occasions, the project team were warned about the potential negative consequences of service users’ involvement in research. In particular, care co-ordinators worried that users were unable to sustain confidentiality about other users. No such difficulties emerged during the study, leading the authors to conclude that ‘users can be successfully interviewed by other users who have been trained to do this. We argue that, in fact, the user interviewers elicit more open and honest responses than professionals do.’71^{(p. 29)} Subsequent studies in this tradition of service-user involvement in research have since reached similar conclusions. 91–94 The narrative within this research tradition positions service users and carers as having much ‘insider knowledge’ to share and that, as a result, questions are framed and studies conducted in ways that are most relevant to users of CMH services.

A moot question is whether or not this makes any difference in terms of the quality and relevance of data collection and interpretation. For example, Rose et al. 95 conclude that the literature is ambivalent about whether or not service-user researchers obtain different quantitative data to conventional mental health researchers (e.g. Rose91 suggesting differences but Rose et al. 96 finding none). However, there is evidence that service-user researchers both collect and interpret qualitative data differently from conventional researchers and in a way that is more in tune with the priorities of service users themselves. 94,97

It is also important to note that most studies of service users’ and carers’ experiences did not involve them in study design and implementation. 70,80,81,86,98–102 However, a commitment to learning more about service users’ and carers’ experiences resonates throughout these studies. This research tradition has also contributed significantly to the debate about whether or not policy changes that were meant to embed principles aimed at increasing service user and carer involvement in CMH practices actually resulted in the desired changes to carers’ and service users’ experiences.

Overall, this corpus of research demonstrates that the introduction of the CPA has mostly failed to deliver on the promises of increasing service-user understanding70,71,86,91 and involvement71,75,77,86,91,92,94,98,101 in care planning and care co-ordination. Research also suggests that other fundamental aspirations of the CPA have similarly had a limited effect on the actual practices of CMH workers. For example, studies show that service users were not aware of, or not allocated, key workers or care co-ordinators70,71,81,92 and had not seen or were not in possession of a care plan or CPA documentation. 70,71,81,86,92,98,101,103

However, the variation in approaches to CPA implementation noted earlier is also apparent when reviewing research findings in this tradition. For example, a small number of studies running counter to the findings listed above have suggested that service users were well informed about their care and treatment plans92,98,103 leading to service users having more trust in staff68,80 and influence and choice in their care. 68,80,103 However, it is worth noting that a considerable weight of evidence suggests the lack of desired impact made by the CPA on service-users’ experiences.

A similar picture emerges in research on carers’ experiences. Studies demonstrate that carers lacked information about the CPA70,100,103 and that carers felt frustrated and isolated70 by not having their views sought or taken seriously. 70,75,86 These findings are not surprising in light of evidence that suggests health authorities often had no formal policy for carer involvement, instead relying upon ad hoc arrangements. 70,75,100 On a more positive note, the CPA had the effect of generating more contact with carers in some areas,68,80 which resulted in greater levels of carer satisfaction. 68,100

None of the studies reports overt organisational and professional strategies of resistance that served to suppress the involvement of service users and carers. Instead, indicators of professional or medical dominance are more subtle but arguably just as profound in diminishing the opportunities for user or carer involvement. For example, Newbigging et al. 77 describe how Independent Mental Health Advocates (IMHA), who play an important part in representing and advocating for service users’ best interests in CPA meetings, are frequently not invited to CPA meetings through apparent lapses in effective communication and diary planning by care teams.

Similarly, when IMHAs were invited, the CPA meetings were poorly organised and often overrun, with the result that IMHAs had to leave meetings before they had finished owing to other commitments. Foucault neatly captures the fact that subtlety is often an overlooked essential for the effective operation of power, stating that ‘Power is tolerable only on condition that it masks a substantial part of itself. Its success is proportional to its ability to hide its own mechanisms’. 104^{(p. 86)}

The CPA has been described as encouraging a focus on service users’ problems rather than their strengths71,92 and as a way of working that prioritises a preoccupation with illness,16,82,94 which is indicative of an approach to caring that creates patient dependency on practitioners through ‘pathologising’ individuals. It is also an approach that means illness comes to define the totality of the person. The nature of the CPA has arguably reinforced a reluctance in some practitioners to move away from an illness model towards a more person-centred and participative mode of mental health practice.

Very few studies explored whether aspects of workplace culture or organisation were prevalent in resisting, frustrating or promoting policy objectives to increase service-user and carer involvement. How far underlying cultural change has kept pace with the more obvious structural reforms in CMH care remains a matter of debate 20 years after the introduction of the CPA. A number of important questions remain about the conditions responsible for promoting or suppressing service-user and carer involvement, the relationship between hierarchies and power structures within CMH workplaces, and the most conducive ways of navigating these to ensure greater participation.

Although those who introduced the CPA deserve praise for encouraging and promoting the development of service-user and carer involvement, this research tradition demonstrates that taking action to involve users requires a willingness to change attitudes and practice, not merely the introduction of policy or best practice imperatives. The danger otherwise is that, without a genuine commitment, calls for greater involvement merely become ‘an exercise in rhetoric’ that leaves existing power relations between professionals and service users/carers untouched.

Many in this research tradition have pioneered approaches to service-user and carer involvement. As a result, service users and carers have contributed greatly to changing how mental ill health is conceptualised and have aided in the production of new knowledge, which has led to a better understanding of methods for improving the lives and advancing the rights of people with mental health problems. It has challenged a model of mental illness as simply deficit and pathology of a psychiatric diagnosis as a ‘master status’105 that swamps any other aspects of the person.

Tradition 3: interventions to improve the care programme approach

The demonstration of clinical effectiveness of health-care interventions has had an increasing role in the UK and devolved governments’ health strategy since the mid-1990s. 106 At the same time, national policy at the macro-level has focused on the CPA as a means of systematically assessing the health and social care needs of those in greatest need, leading to the development of an individual care plan, co-ordinated by a keyworker. As already discussed, however, very different versions of the CPA were implemented at the level of service delivery within and across regions of England. A tradition of research therefore emerged during the late 1990s that reflected both the perceived need to demonstrate and improve clinical effectiveness, as well as capturing the early difficulties and variation in the implementation of the CPA (see Appendix 6).

As already discussed, the CPA led to diverse, often time-consuming, bureaucratic practices that meant staff spent less time with patients. 16,68,83 CPA documentation is framed by researchers and clinicians as problematic, since poor documentation increases the risk of vital patient and treatment information being missing when planning and co-ordinating care. Consequently, the third research tradition describes how the elimination of variation in documentation processes can lead to improved performance and reliability in care planning and outcome measurement.

Macpherson et al. 106 provide an example of a study that encompasses issues related to variation and effectiveness; stating that standardised documentation for individual care planning should be combined with outcome measurement, to give a meaningful measure of the effectiveness of care. To better understand the relationship between care planning and patient outcomes, a formalised space was introduced to the CPA documentation for establishing and standardising treatment goals. Goals were set within a care planning assessment and review meeting and agreed with patients (n = 139), relatives, professionals and advocates prior to the meeting’s conclusion. One year later a clinical review meeting found that 68% of goals were fully and 11% partially achieved (43% of goals were partially or not achieved). Goals targeting drug treatment of psychiatric syndromes were most likely to be fully successful (84%), whereas approaches to self-care skills, side effects, physical/medical problems and family relationships were moderately successful.

Recent UK mental health policy has emphasised the need for services to adopt a ‘recovery orientation’ to improve service users’ experience of care, social inclusion and recovery. An attempt to standardise outcome measurements for recovery was undertaken by Killaspy et al. 102 who assessed a measure of recovery, the Mental Health Recovery Star (MHRS) for acceptability, reliability and convergent validity. Recovery was defined by the authors as ‘a personal and dynamic process of adjustment and growth following the development of a mental health problem’ (p. 65).

Although the MHRS was relatively quick and easy to use and had good test–retest reliability, inter-rater reliability was inadequate. Furthermore, convergent validity suggests that MHRS assesses social function more than recovery, leading to the conclusion that it cannot be recommended as a routine clinical outcome tool but may facilitate collaborative care planning. Interestingly Gould’s94 study reviewed in Tradition 2 also suggested that the MHRS operated more as a measure of social functioning than recovery. Others saw the MHRS as too complicated, putting too much pressure on service users with long-term problems to find employment and being a rigid tool.

Both Lockwood and Marshall107 and Marshall et al. 108 researched interventions aimed at standardising patients’ needs assessments; the first of these studies was a pilot study, which led to the second study. The intervention in both studies is relatively convoluted and complex, initially involving a research nurse undertaking a baseline assessment of the patient to identify needs. Data from these assessments were then entered into computer software to determine which of the practitioner-identified needs or problems required action. When needs required action, the research nurse and consultant psychiatrist considered a list of pre-defined interventions provided by the software, before deciding whether or not the patient was likely to benefit from any of the interventions. The research nurse then provided the keyworker with a report of the needs identified, which were then used to guide care plan discussions between the keyworker and the patient.

In the pilot study significant reductions were seen in ‘unmet’ needs and the level of anxious/depressive symptoms, suggesting that needs feedback improved the quality of nursing assessment and care planning within the CPA. However, the follow-up study108 found that standardised needs assessment did not substantially enhance care planning. The process of using an independent registered nurse, who was not a keyworker or a member of the clinical team to undertake the needs assessment, allied to having to search a computer database for interventions associated to needs, would have made transferability of these findings into clinical practice difficult, as the trial intervention differed greatly from usual practices.

The CUES-U tool (Carers’ and Users’ Expectations of Service – User version) tested by Blenkiron et al. 99 consisted of a 17-item service user outcome scale in booklet form developed by academics, clinicians and service users. CUES-U was described as an ‘important tool’ (p. 334) because it focuses on issues of quality of life and satisfaction that mental health service users (rather than professionals) identified as priorities. Service users (n = 86) completed the CUES-U booklet before returning and discussing the contents of the booklet with care co-ordinators, who then recorded changes made to care plans as a result of receiving the CUES-U feedback.

The CUES-U mediated discussion led to a change in clinical care for 49% of respondents. Care co-ordinators rated CUES-U as a good use of their time in 64% of cases. A large proportion (84%) of service users were satisfied with the level of control and consultation they received; 87% were satisfied with their relationships with mental health workers and > 70% were satisfied with levels of information and advice and access to services. However, a significant limitation of high levels of satisfaction was that patients knew that their care co-ordinator would see their replies, although many did write negative comments in free text boxes. The authors concluded that CUES-U can be an effective and practicable tool for increasing users’ involvement in their care and for service benchmarking.

A move from paper documentation towards electronic CPA (eCPA) records was proposed by Howells and Thompsell. 109 The eCPA consisted of a computer-based CPA system for care planning and documentation – using a Microsoft Word template – designed to improve the quality of information in CPA care plans in a CMH team in London. Completed eCPA care plans were e-mailed to the acute ward, the hospital’s emergency clinic and any other agencies. The CPA manager received the original signed copy, a copy was filed in the case notes and the GP, the patient and carer were given a copy. The eCPA was welcomed by staff with a take-up rate of almost 100%. Patients welcomed the legibility and detail of the forms and expressed no concerns about the change to the eCPA. Care plans were longer and more detailed, being no longer constrained by fixed-size boxes on paper forms. Care plans were also adjusted more frequently by CMHT staff, who did not have to completely rewrite the forms by hand. As a result, the plans better reflected the changing needs of patients.

To summarise, attempts to improve needs assessment and/or care planning processes by changing documentation featured in all of the interventions, either through amending features of existing documentation106,109 or introducing new assessment and planning documentation and processes. 99,102,107,108 Some of the interventions were described as leading to more collaboration with service users during care planning,99,102,106 whereas Howells and Thompsell109 reported greater collaboration within CMH teams. However, Lockwood and Marshall107 and Marshall et al. 108 reported marginal or no improvements in needs assessment and care planning during a pilot study and trial of an intervention that bore little resemblance to usual clinical practice.

Interestingly, those who developed the interventions via an inclusive, multidisciplinary and/or service-user involvement approach reported most success,99,102,106,109 although the quality of research in this tradition is generally weak with no long-term studies of the effects of these interventions on improving care planning and patient outcomes.

Health policy from 1990: changes to community mental health working in England and Wales

Initial moves towards a different way of organising and delivering CMH services can be located within a brief health and LA circular policy document for England. 112 Produced in the context of an accelerating, policy-driven shift away from hospital care, the circular emphasised the importance of ‘systematic arrangements for deciding whether a patient referred to the specialist psychiatric services can, in the light of available resources and the views of the patient and, where appropriate, his/her carers, realistically be treated in the community’. To this end, the document specified that health and social care needs should be assessed and planned and a ‘keyworker’ appointed to oversee and review the delivery and co-ordination of interagency and interprofessional services.

The circular appeared shortly after the publication of plans for an across-the-board reorganisation of community care via the White Paper Caring for People. Subsequently incorporated within the NHS and Community Care Act,113 this placed a responsibility on LAs to implement ‘care management’. With their parallel systems of care co-ordination, for many years thereafter the health-led CPA, and the social care-led care management, ran in unhelpful parallel.

Uppermost in the minds of policy-makers in the early 1990s was the management of the risks associated with the community care of people with severe, long-term, mental health problems. As a result, ‘supervision registers’ were introduced, this again being an initiative within England only, where they were implemented as an add-on to locally structured CPA arrangements. 114 Supervision registers promoted care planning and co-ordination as a mechanism for risk assessment and its management. However, as with the introduction of the CPA, no additional resources were provided to support those appearing on supervision registers.

Building Bridges115 was a detailed document produced for NHS England in 1995, which promoted the importance of interprofessional CMH teams and the CPA as the ‘cornerstone’ of care. 115^{(p. 45)} A year later guidance on the community care of people with mental health problems in Wales was published,116 which formally introduced into the Welsh mental health system the ideas of health and social care assessment, care planning, review and key-working. Neither this nor subsequent additional guidance117 mentioned this as the introduction of the CPA into Wales.

Meanwhile, the election of New Labour to government in 1997 coincided with a significant raising of the profile of mental health as an area for policy action in England, with new guidance on ‘modernising’ the CPA being introduced towards the end of the decade. 118 This reaffirmed the central place of the CPA in modern systems of care but also highlighted where changes needed to be made, noting professionals’ complaints that the CPA resulted in a significant administrative burden for time-pressed care co-ordinators and service users’ views on patchy and inconsistent experiences of care planning.

In post-devolution Wales, a new national strategy for adult mental health services included a section on care planning and co-ordination and on the imperative for all users to have a written plan of care overseen by a ‘keyworker’. 119 Two years later came the formal introduction, for the first time in Wales, of something directly referred to as the CPA. 2 This talked, very clearly, about the CPA as a vehicle assisting service users towards recovery, supporting empowerment and the embracing of holistic care.

The current situation in England and Wales

In England, a further review and ‘refocusing’ of the CPA was undertaken1 which was strong on minimising bureaucracy and on simultaneously driving up the quality of care planning via a commitment to national-level consistency. The CPA was presented as a values-based process through which care is tailored uniquely to the individual and serves to promote social inclusion and recovery. For the first time a set of approved training materials were produced to support practitioners. 120 In current mental health policy for England, represented most completely in the cross-government No Health Without Mental Health121 and its ancillary documents, the CPA takes something of a back seat as the focus shifts to cross-sector and cross-government action to improve mental health across the board.

Meanwhile, in Wales, care planning and co-ordination have taken a distinct turn in recent years. An influential, critical, review of the CPA122 contributed to the power to make new mental health law devolved to the National Assembly for Wales. This process led to the eventual passing of the Mental Health (Wales) Measure [MH(W)M] and its attendant documents, including a code of practice123 and a raft of training material. Part 2 of the Measure obliges health and LA services to produce CTPs for all users of secondary mental health services. The phrase ‘CTP’ supersedes ‘CPA’, although each is required to be developed by a care co-ordinator working collaboratively with the service user and other providers. CTPs must be in writing, and kept actively under review by care co-ordinators who must now be drawn from a prescribed range of professional groups. Since the passing of the Measure, a new overarching strategy for mental health across the lifespan in Wales has been produced,5 along with an initial plan for delivery. 124

Current policy frameworks across both England and Wales emphasise the importance of mental health services in promoting recovery and tailoring care to the needs of the individual. Where the biggest difference remains is in the context of arrangements for care planning and co-ordination, and in the use of a statutory instrument in Wales within which CMH care is overseen. Current mental health policy in England also heavily features the term ‘personalisation’, with no use of this term appearing within current mental health policy in Wales.

Personalisation, through service users holding personalised budgets, is portrayed in policy documents from England as a means of ‘giving people greater choice and control over their care and treatment’. 121^{(pp. 32–3)} In Wales, an Independent Commission on Social Services rejected this view of personalisation, stating that ‘We believe that the label ‘personalisation’ has become too closely associated with a market-led model of consumer choice’. 6^{(p. 15)} The Commission is clear that the focus instead should be on personalised budgets offering the means of establishing patient and public ‘voice’ as a force for improving services that meets the diverse needs of the population of Wales, rather than promoting individualised consumerist choice as a means of improving services through market forces. Peckham et al. 125 similarly identify that different political ideologies are increasingly apparent in policy rhetoric, objectives and mechanisms introduced within devolved national policy, although the question of how visible such differences are at the level of mental health service organisation and service delivery remains to be seen. Such is the focus of this study.

Recruitment and case-study sites

Artois

Artois Trust provides both primary and secondary mental health services to a population of around 1.6 million people. It covers a large geographical area, which is predominantly rural, and serves six LAs. Owing to the large catchment area, ethnicity and deprivation rates vary. According to the 2011 census, in one area 16% of the population were from a non-white background, whereas elsewhere this figure was only 7%. Equally, deprivation rates vary from high in the more densely populated urban areas to low in the rural communities of affluence. Adult psychiatric admissions are provided in seven hospitals and six rehabilitation units with 149 beds, 23 of which are accessible to the principal recovery team that data were collected from. Community services are provided through 13 CMHTs. Adult psychiatric admissions are provided in seven hospitals and six rehabilitation units with 149 beds, 23 of which are accessible to the principal recovery team that data were collected from.

Recovery Profile from the Recovery Self-Assessment Scale

Below we include a recovery profile for the site based on individual item analysis of the five highest rated recovery-orientated practices. This is based on the mean scores of all of the respondents and is presented from the perspective of the service user and service provider.

The five highest rated items for the service users and care co-ordinators were within the choice, life goals and individually tailored services subscales (Table 8). Other high-rated items by the service users were also within the life goal domain, such as staff knowledge of special interest groups, encouragement to take risks and try new things, staff use of recovery language and staff diversity in terms of culture, ethnicity, lifestyle and interests. In addition to this, strengths were seen in the area of choice such as monitoring personal goals and staff listening to and following choices and preferences. Other highly rated items for care co-ordinators were in the area of life goals such as assisting people to develop career and life goals (including education and employment) and accessing activities and other programmes and services. Other strengths were involving significant others and goals around choice such as listening to preferences and choosing practitioners.

Burgundy

Burgundy LHB serves approximately 500,000 people, of whom 1.6% are from black and minority ethnic (BME) groups. It covers a wide geographical area with a mix of urban and rural communities, some of which are densely populated. Twenty-four per cent of the small areas measured in the largest urban part of the locale, and almost 30% of a second part are in the most deprived 20% of areas in Wales. 128 Mental health services are provided in three hospital sites and one community rehabilitation unit with 85 adult psychiatric beds, from eight CMHTs and from a range of specialist services. The main CMHT that data were collected from has 20 psychiatric beds available to them.

Participant characteristics: care co-ordinators

Thirty-one care co-ordinators from four CMHTs within the locality completed the questionnaires. The majority of respondents had spent > 4 years working in mental health services (83.9%) and two-thirds had spent > 4 years working as a care co-ordinator (67.7%). Further details of the demographic characteristics can be found in Table 15.

Participant characteristics: service users

In total, 500 questionnaires were sent to service users to invite them to take part in the study. We received 72 responses, which gave us a response rate of 14.4%. Further details of the demographic characteristics can be found in Table 16.

Summary scores for the questionnaires

The information presented here will be in the same format as that presented for the previous research sites.

Empowerment Scale

The overall mean score for empowerment was above the midpoint for the instrument (Table 17). Out of a possible score of 4, indicating a higher perceived level of empowerment, the mean ± SD score was 2.56 ± 0.38 this is slightly lower than the reference value. 126 The subscale values were slightly lower than the reference group.

Scale to Assess the Therapeutic Relationship

The mean total score for the STAR-P was 34.09 (10.13), 17.13 (5.79) for ‘positive collaboration’, 8.01 (3.05) for ‘positive clinician input’, and 9.09 (2.80) for ‘non–supportive relationships’ (Table 18). These values are slightly lower than the reference values obtained from the validation paper. 46

Recovery Self-Assessment Scale

Mean scores and standard deviations for the RSA scale are provided in Table 19. Mean scores from service users and care co-ordinators on the subscales fell in the middle to moderate range (2.70/5 to 3.79/5). The difference in mean total RSA scores was only marginal; care co-ordinators had the highest ratings (µ = 3.35, SD = 0.56), followed by service users (µ = 3.18, SD = 0.87). The lowest scoring subscales were ‘Involvement’, ‘Diversity of treatment options’ and ‘Individually tailored services’ for both participant groups. Within these subscales, the scores demonstrate indifference.

Recovery Profile from the Recovery Self-Assessment scale

The five highest rated items for the service users and care co-ordinators were within the choice and life goals subscales (Table 20). Other highly rated items by the service users related to life goals such as ‘staff use a language of recovery and believe that I can recover’. In addition to this, respondents reported that most of the services are provided in their natural environment. Other highly rated items for the care co-ordinators were in the area of life goals, such as using language of recovery, assisting the person to fulfil their goals and aspirations, and with development of career and life goals (including education/employment). In addition to this, other strengths were discussing sexual and spiritual needs and interests, listening to the choices and preferences of participants, and actively involving service users to give back to their community.

Narrative summary from interview data: senior managers and senior practitioners

We conducted interviews with two senior managers and five senior practitioners. The senior practitioners consisted of a mental health nurse, psychiatrist, social worker, psychologist and ward manager.

Local context: Champagne

Local developments in the period immediately preceding and during data generation included the introduction of integrated LHB/LA management within CMHTs and the setting up of assertive outreach and primary mental health care services. Newly integrated managers were described as knowing what needed to be done while simultaneously carrying significant responsibilities, lacking wider influence and having to work to two organisations. Separate IT systems and lack of higher-level management integration were given as examples of the progress still to be made in inter-agency working.

A senior manager (C-SM-001) said how expectations of service users and third sector organisations were high following the introduction of Together for Mental Health5 and now the MH(W)M. The Measure, in particular, was seen as highly significant and as exerting an impact. A senior professional said how the Measure ‘is massive’ (C-SP-003), and at the highest level, a partnership board with LHB and LA representation shared responsibility for the creation and implementation of new policies and procedures following its passing. National aspirations were reflected in local expressions of commitment to recovery, user involvement, personalised care, dignity, respect, safety and outcome-focused interventions. These were tempered (for some) by accounts of a lack of senior management vision, gaps between policy aspiration and practice reality, and everyday pressures for staff. For example, one senior professional expressed the view that, ‘There are service user involvement protocols in place across the board, but I don’t think it’s been grasped and led at a senior management level, middle management level’ (C-SP-001). Other concerns included increases in demand and levels of acuity, high caseloads for community workers, cuts in training and development opportunities, and rising administrative workloads.

Care planning and care co-ordination

Implementation of the MH(W)M had placed care planning and care co-ordination on a statutory footing for ‘relevant patients’ accepted by secondary mental health services, and local policy was being developed outlining the expectations of care co-ordinators. Care co-ordinators used the template all-Wales CTP addressing eight areas of life, but lacked an equivalent national template for the conduct of general, and risk, assessments. The CTP format was positively appraised in providing structure and a framework for recording the service-user journey, and in requiring staff to attend to a wide range of needs. It was also described as reductionist and bureaucratic, and as insufficiently weighted towards psychological, occupational and medical domains consistent with the provision of mental health care. A senior professional said how the questions on the CTP forms are:

pretty reductive . . . it would be nice to have a question about, why would you want to do this, because then you can capture a bit of the person’s motivation, and some of their values . . . and also I suppose something about what difference would that make.

C-SP-003

An ongoing challenge, described as a cultural one, was the local determination of what constitutes a ‘good’ care plan.

Only mental health professionals could act as care co-ordinators, and contrasting figures were given for average caseload sizes. The upper range stated for nurses and social workers was 30–40+. Once appointed, care co-ordinators carried significant, formal, responsibilities, including the recognition and management of risk, which was described as sitting at the heart of everything. A senior professional described this as ‘very much at the forefront of what we do . . . safety and risk, rightly or wrongly drives what we do, very much, and a lot of the time’ (C-SP-002). An expectation was that newly created care and treatment plans be reviewed after one month in the first instance, with all care and treatment plans subject to review on at least an annual basis. Widespread support was given to the involvement of users and carers in reviews, but carer involvement varied, reflecting both service users’ wishes and differences in practice among professionals. Here, a senior professional referred to some of the tensions:

I think carers’ needs are always high on, in everyone’s mind . . . we do where appropriate, ask people and support them. But again, our views probably completely differ to that of carers . . . I’m aware that we probably are not delivering the service that they want.

C-SP-001

More integrated records and IT, better cross-agency sharing of information and greater administrative support were all seen as improvements needed to support care co-ordinators in their work, along with more people able to fulfil the role. One view was that care and treatment plans could be completed by people with less training.

Welsh Government-commissioned learning materials supporting the implementation of the MH(W)M were available, and an all-Wales approach to training in risk assessment and management was in place. However, senior professionals talked of a lack of care co-ordinator training opportunities, including in the use of recovery principles and in the competencies needed for the role. Where local training was available, participation could be difficult given prior clinical commitments.

Recovery

Recent policy and legislation set an agenda for the provision of care informed by recovery principles, echoed locally in the creation of an inter-agency charter underpinned by commitments to recovery values. Widespread support was found for recovery ideas, and participants spoke of the importance of hope, choice, empowerment and holism, and of people being enabled to live meaningful lives with or without symptoms. Exemplifying this, a senior professional talked of:

providing hope, and delivering a service that actually makes that person feel hopeful that they can recover in the first instance . . . [and] that they will get out of life what they want to get out of life while living with an illness’, and that recovery should be a ‘whole system view.. [and] should come right from the top and . . . be written into every operational policy, into people’s job descriptions . . . written in every policy, every job description.

C-SP-001

Anecdotal evidence was said to exist of shifts away from paternalism and from exclusively biomedical approaches. A senior manager (C-SM-001) also talked of necessary cultural and professional change to fully realise this values-based aspiration, and that rebranding existing ways of working would not be enough. Recognition of organisational commitments to recovery and of greater user involvement was tempered by concerns that recovery practice was developing at an uneven rate, with teams moving at a different pace. One view was that ‘warm words’ (C-SP-003) had been grafted onto a medical model, and another was that training in recovery and in the use of recovery tools needed to take place. Recovery plans and strengths approaches barely featured in participants’ talk, and the use of care and treatment plans for this purpose was hampered by the CTP template wording, which lacked a clear recovery focus. The idea that cultural change was needed was also extended to people using services, as well as to those providing them.

Personalisation

Participants saw personalisation as referring to care that was individually tailored, and (for some) as care that additionally gave people the freedom to lead or develop services for themselves in potentially idiosyncratic ways. A senior manager commented that, ‘It means making sure that we’re striving to make sure that the person’s at the centre of their case’ (C-SM-001) and a senior professional talked of, ‘the person leading, and having control over what happens to them within the services’ (C-SP-001). Reference was made to the importance of acknowledging potential differences in view between people using, and people providing, services. Tension was seen as a possibility when wanting to adopt a personalised approach with people who were very unwell and not wanting contact. Risk was seen as a challenge to personalisation. Another concern was that personalised care could raise expectations for help in areas in which health and social care services were ill-suited to respond.

Care and treatment plans were thought to be helpful in the promotion of personalised care, and personalisation as an ideal was seen to fit with recovery as part of the overall aim of the service. One participant (C-SP-001) said this aim was already being achieved. This was not a uniform view, however, and others were clear that personalisation was not happening: ‘Personalisation is at the moment a concept but it hasn’t really been put in practice’ (C-SP-002). Differences were found by background, with senior social care professionals demonstrating greater awareness of personalisation than senior health-care professionals. Professionals were also not sure if LHB policies existed in this area, and training opportunities did not feature. Moves towards direct payments were described, but participants recognised that the use of personal budgets was not well developed. It was suggested that one way that more personalised practices could be developed was to establish a forum where staff and service user experts might share and learn from each other’s experiences.

Suggestions for improvements

Participants spoke of the importance of having vision, and of being able to offer choice, to instil hope and to involve people in their own recovery. Having the time to provide high-quality care was important, but people spoke of a demanding system that inhibited practitioners’ best intentions. Closer integration between health and social care agencies was proposed with a view to harmonising procedures, along with interprofessional and interagency training to support recovery-focused services and personalisation. Brave decisions were asked for to support the shift to personal budgets, and managers were asked to renew their focus on leading the incorporation of recovery ideals into practice. Professionals spoke of wanting encouragement for more positive risk-taking and of risk trumping the focus on recovery. The experience of under-resourcing was described as dispiriting, and of care plans pointing to services that are not available. Specific new initiatives, which service users were said to have asked for, included investment in peer support and resource centres. Change across the whole system was seen as applying to third-sector organisations that with statutory services might need to change their ways of working.

Narrative summary of interview data: service users, carers and care co-ordinators

We conducted interviews with six service users, four carers and five care co-ordinators in Champagne.

Care planning and co-ordination

On the whole, care planning was felt to be useful, especially as a means of monitoring personal progress: ‘[knowing] how far I’ve come from where I was and how far I’m going to go’ (C-SU-001). It was, however, acknowledged that this usefulness was also contingent upon other factors, such as the implementation of regular updates, the scheduling of structured activities and a consistent level of regular personal contact from staff. Carers, in particular, were appreciative of the CMHT support that accompanies the care-planning process. Although most service user respondents had not received copies of their care plans, and some had not seen a care plan at all, their potential usefulness was nevertheless recognised. The one respondent who did have a copy described referring to it on a regular basis, and feeling a sense of ownership in relation to it. Importantly, service users also emphasised that, for some, it may be difficult to admit that mental health care is required and that, in such situations, help may not be welcomed. With the exception of one service user, respondents understood their care plans fairly well. The inclusion of emergency contact details was also considered important by service users. There was some suggestion from service users that electronic formats might be useful. This possibility was not mentioned by carers, and care co-ordinators preferred written (paper) formats, with one respondent noting that there have been changes to the language used to describe care-planning work.

For care co-ordinators, care plans are considered useful as long as sufficient time is available to work on them properly; the need for more time was a shared concern within this group. The function of care plans as providing a valuable timeframe within which they could work was recognised, although some respondents were unclear whether or not service users had copies of current plans. For these respondents, care planning was associated with collaboration, risk management and recovery, although there were mixed feelings about the importance of regularly referring to the documents. One respondent considered care planning to be a helpful way of monitoring service users’ progression, whereas another felt that it was more important to have a good knowledge of exactly what is happening at any given time with a service user’s care. One respondent conceptualised the care plan as ‘a bit of a contract’ (C-CC-004), facilitating the development of an agreement with service users about the content of their plan. The variability of service-user engagement was felt to be challenging in some cases, and an example was given of one service user who hides their care plan so that it cannot be easily located or read.

Overall, care co-ordinators were keen to emphasise that the usefulness and effectiveness of care planning is dependent upon the specific individuals involved. Interestingly, although one respondent suggested that the care-planning process could work in the absence of care plan documents, another felt that care plans were crucial unless service users could be seen on a daily basis. It was also mentioned that care plans should be conceptualised as an ongoing, continual process, rather than as a task that ought to be completed and ‘signed off’. Another hindrance to effective care planning included the ‘cumbersome’ (C-CC-002) IT systems currently in use.

Most service users were satisfied with their level of involvement and felt that they were adequately listened to. One respondent without a care plan expressed a wish for greater involvement and noted that the provision of care plans without the appropriate follow-up process was unhelpful. Carers did not describe a high level of involvement themselves – although some did attend meetings – and this did not seem to be a major area of concern. For some, a purposeful choice was made not to be involved, such that service users were given full scope to speak for themselves: ‘I wanted him to speak his mind alone and then I would come in afterwards’ (C-CA-003). Care co-ordinators, meanwhile, felt that they encouraged involvement from service users and carers wherever possible: ‘I think I can speak for everybody in the team where it’s patient centred’ (C-CC-005). Respondents emphasised that informal involvement – ‘on a walk and talk basis’ (C-CC-002) – ought to be recognised as valuable. From a care co-ordinator perspective, then, ‘it’s very much [the service user’s] . . . as much as possible we put the onus on the patient’ (C-CC-005).

Mixed feelings about the consideration of safety and risk were expressed by respondents. Most service users agreed that these aspects had been adequately addressed. It was however acknowledged that such issues were not always explicitly discussed and that even where they were, discussions usually centred primarily on the identification of emergency contact information. Some service users were uncertain whether risk assessments were included in their CTP, and one respondent mentioned that they were still able to ‘get away with’ self-harming [C-SU-004], and another described feeling neglected since their discharge from hospital: ‘[at home] nobody’s checked up on me or anything’ [C-SU-005]. There were no serious concerns expressed among carers, who were on the whole fairly satisfied with the management and consideration of risk and safety, but did feel that perhaps these matters could be discussed more thoroughly with CMHT staff.

For care co-ordinators, however, thorough risk assessment was acknowledged to be ‘paramount’ [C-CC-004] to the care planning process, and it was suggested that it can sometimes be difficult to establish the necessary balance between safety and positive risk taking, especially given the shift among practitioners away from a risk averse culture and towards embracing positive risk taking more generally. Respondents described the importance of knowing service users’ individual limitations in this latter respect, and noted the importance of considering possible risk to both self and others. It was suggested that the awareness of risk was often an intuitive matter: for one respondent, it was a continual focus – ‘it’s in the back of your mind constantly, I guess’ [C-CC-004]. Another respondent described their approach as ‘very safety conscious’ [C-CC-005], and also explained that current working practices are underpinned by a ‘very low tolerance of any sort of verbal abuse or . . . physical intimidation’ [C-CC-005].

Other than one service user respondent who recounted a traumatic experience of having been poorly treated at a local Accident and Emergency department, this group did not express strong feelings about possible improvements to care planning. Carers, meanwhile, although generally satisfied with their current situations, suggested that better communication between different services would be helpful, and felt that some services were inadequately funded, resulting in a lack of available staff and limited contact. Better access to respite care was also cited as a possible improvement, as well as increased help from out-of-hours services, more face-to-face contact with staff, and a greater number of structured activities for service users.

Care co-ordinators were similarly concerned about the current lack of resources in mental health services and felt that, as a consequence of this, existing work necessarily involves more crisis management than ought to be the case. It was also noted that medical practitioners, in particular, have very little time available for the CTP process, and that perhaps it would be better assigned to staff in other professions. More training for the drawing up of care plans was deemed necessary, and it was suggested that peer support workers would be useful for service users in helping them to write their care plans themselves. The duplication of information observed in care plan formats was considered unnecessary, and it was emphasised that generic templates do not work effectively: ‘one size fits no-one’ [C-CC-002].

Recovery

On the whole, service users were less certain than carers or care co-ordinators in terms of defining recovery, with some respondents more knowledgeable than others. Responses were very mixed: two respondents associated recovery with the notion of ‘getting better’, being symptom free and remaining out of hospital. Others acknowledged that, although they will not make a full recovery, they are keen to be symptom free without having to take medication. The ability to engage in structured activities (e.g. work, hobbies) was indicative for some service users of a shift towards recovery, and some respondents felt that a healthy mind, a positive outlook and the ability to function socially were key objectives.

Some service users had been introduced by their care co-ordinators to various day centres and other local facilities or helped to access suitable courses. Some took up these opportunities but others tried them briefly but felt unable to continue owing to the inhibiting nature of their mental condition:

I’ve been told about [NAME OF LOCAL CENTRE] . . . I’ve been there a few times, but I don’t particularly like it there, and there’s one in [PLACE NAME] . . . I went there for a very short time, a period . . . It was all right, but I think you know your limits, like. If you feel a bit paranoid or panicky you just withdraw, so that’s why I didn’t bother going again see, yeah.

C-SU-001

Another service user said that she had not really been given information about support groups but her co-ordinator had ‘got me weekly art therapy for up to two years. . . and I start a 12 week Mindfulness course in a couple of weeks’ time’ (C-SU-002).

Most carers, meanwhile, seemed to have a good understanding of recovery, acknowledging that the recovery process is necessarily a wholly individual one, that the primary objective is to manage life within the boundaries of mental illness, and that this will necessarily mean living a different kind of life than previously experienced pre-illness. These views were mirrored in care co-ordinators’ responses, which were similarly knowledgeable and also focused upon the significance of individualisation. The re-establishment of a meaningful day-to-day life was considered important here, and the process of empowerment – building upon service users’ strengths and abilities – was recognised to be valuable for recovery purposes. One carer said that the service user had been helped to undertake ‘a lot of activities’ (C-CA-003) and proceeded to list positive thinking, cognitive–behavioural therapy, table tennis, a gym and a local centre that provided numerous activities and support. Service users agreed strongly that their own strengths and achievements were recognised. Some association of recovery with a drive towards discharge targets was made, but this was not a dominant tendency.

There was a general uncertainty among respondents over whether or not care plans were helpful for recovery, and no strong opinions were voiced in relation to this. Some service users, in particular, felt that the structured goals incorporated into their care plans had been useful, whereas others valued their implicit function as a ‘safety net’ alongside other support networks, for example, friends and family, as well as structural support such as benefit payments and financial assistance. Other respondents suggested that a more ‘active’ care plan would be beneficial in striving for recovery. Carers’ responses on this point were equally diverse; medication was felt to be important in some cases, and the value of psychological therapies – with a desire for greater availability – was also mentioned. Support from the appropriate professional at the right time was described as being crucial to the recovery process. Less emphasis was given by carers to the identification of personal strengths, and one respondent was unsure whether or not these would actually be recorded in the care plan document upon their recognition. None of the service users had a recovery plan, and although one carer assumed that such a document had been drawn up, they were unclear about this. Although there was a general awareness of the existence of additional help and support networks, carers indicated that such resources were often difficult to find, and that more guidance was required in this respect.

Most care co-ordinators felt that they did already work in a recovery-focused way, and there were suggestions here about the ways in which CTP provides a good structure for visualising the recovery process. Limited resources and risk adversity were cited as hindrances to the delivery of recovery-oriented care, as were the demands of detailed documentation. There was a sense that as a consequence, work often involves a great deal of ‘firefighting’ (i.e. dealing with emergencies rather than advance planning). It was noted by some respondents that longer-term service users might find the drive towards recovery upsetting at times, and also that the implicit emphasis on discharge can be challenging for staff.

Personalisation

Most service user and carer respondents did not have a clear understanding of the term ‘personalisation’. Several had never heard it at all, and the concept had never been explained to them. Care co-ordinators, meanwhile, were all at least familiar with the term, but many were unable to provide a clear definition, and felt the concept to be an ambiguous one. As a consequence of this ambiguity, two respondents described feeling uncomfortable about using the term at all. For some care co-ordinator respondents, personalisation was associated with choice, bespoke services and the unique tailoring of care provisions, but this was not universally recognised. It was also acknowledged that, although the concept of personalisation may be useful in principle, a lack of relevant services and support systems renders it less valuable in practice.

For service users, there were mixed feelings about the extent to which care was currently tailored to their own individual needs. Some felt that this was the case, whereas others were less convinced. Similarly diverse feelings were expressed in terms of whether or not respondents felt in charge of their care; this was not the case for all. There was also some sense among service users that they preferred to defer to the knowledge of carers and professionals when making key decisions. Carers were similarly ambivalent; some respondents felt that services were personalised, whereas others cited a need for more comprehensive help and support. There was also some acknowledgement of the fact that service users’ expectations for personalised care cannot always be met, and that some services (especially psychological therapy) are simply not available.

Care co-ordinators maintained that their approaches are as personalised as possible in the face of budget constraints, time limitations and demanding workloads. It was widely acknowledged that there is often a disparity between the concept of personalisation as an ideal and the reality of the care that can actually be delivered. For one respondent, the flexibility and creativity encouraged by a personalised approach were seen as positive qualities, although many others recognised that lack of service-user engagement, and the reluctance of some to embrace choice in their care planning, were significant hindrances. It is notable that, although personal budgets were mentioned by care co-ordinators, neither carers nor service users referred to these.

Embedded case-study comparisons

Detailed case-study comparisons were made for six cases consisting of three triads of service user, carer and care co-ordinator; two of service user and care co-ordinator only; and one of service user and carer only. Tables were drawn up for each of the themes used for our framework analysis; an example is presented in Appendix 9.

Dauphine

Dauphine Trust provides mental health and community services to a population of around 750,000 people. It covers an extremely densely populated urban area, which is very multicultural. Figures from the 2011 census showed that this site’s catchment area had one of the most ethnically diverse communities in the UK. For example, one area had 45% Asian (Indian, Bangladeshi, Pakistani, Chinese and other Asian), 27% white British/white other, 19% black African/Caribbean and the remaining 9% split between mixed, Arab and other backgrounds. According to the 2010 English Indices of Deprivation (EID), a high percentage of the areas covered by this site are among the top 10% of the country’s most deprived boroughs. Inpatient mental health services are provided from three hospital sites with 297 acute inpatient beds, and community services from 10 CMHTs. For this study, the CMHT that data were collected from had 14 psychiatric beds accessible to them.

Participant characteristics: care co-ordinators

Twenty-eight care co-ordinators from three CMHTs within the locality completed the questionnaires. The majority of respondents had spent > 4 years working in mental health services (92.8%) and had spent > 4 years working as a care co-ordinator (85.7%). Further details of the demographic characteristics can be found in Table 27.

Participant characteristics: service users

In total, 477 questionnaires were sent to service users to invite them to take part in the study. We received 92 responses, which gave us a response rate of 19.2%. Further details of the demographic characteristics can be found in Table 28.

Summary scores for the questionnaires

The information presented here will be in the same format as that presented for the previous research sites.

Empowerment Scale

The overall mean score for empowerment was above the midpoint for the instrument (Table 29). Out of a possible score of 4, indicating a higher perceived level of empowerment, the mean ± SD score was 2.62 ± 0.44; this is slightly lower than the reference value. 126 The subscale values were all lower than the reference group.

Scale to Assess the Therapeutic Relationship

The mean total score for the STAR-P was 36.07 (9.03), 18.62 (4.92) for ‘positive collaboration’, 8.46 (2.75) for ‘positive clinician input’, and 9.14 (2.87) for ‘non-supportive relationships’. These values fall slightly below the reference values obtained from the validation paper; however, one of the subscales, non-supportive relationships, is very close (Table 30). 46

Recovery Self-Assessment Scale

Mean scores and SDs for the RSA scale are provided in Table 31. Mean scores from service users and care co-ordinators on the subscales fell in the middle to high range (2.70/5 to 4.04/5). For the mean RSA total score, the care co-ordinators score higher (µ = 3.57, SD = 0.56) than the service users (µ = 3.12, SD = 0.94). There is some variability between responses for the service users and the care co-ordinators. For the service users, the lowest scoring subscales were ‘Involvement’, and ‘Diversity of treatment options’. Within these subscales the scores demonstrate indifference. The highest rated subscale score is for the ‘Choice’ subscale; in particular the care co-ordinators scored highly on this, indicating that this is seen as an area of priority by the clinicians.

Recovery Profile from the Recovery Self-Assessment Scale

The five highest rated items for the service users and care co-ordinators were within the choice and life goals subscales (Table 32). Other items rated highly by the service users were that staff members listened to them and followed their choices and preferences, used a language of recovery and encouraged them to take risks and try new things. Other highly rated items for the care co-ordinators were in the area of choice, such as listening to choices and preferences, choosing practitioners and monitoring progress on goals. In addition to this, other strengths were related to life goals; staff agreed that their primary role is to help people in recovery to be knowledgeable about specialist interest groups, use a language of recovery, develop career and life goals, and fulfil aspirations. Staff also strongly agreed with the importance of their role in helping people to become involved with activities to connect with communities, and in developing personal leisure interests and hobbies.

Narrative summary of interview data: senior managers and senior practitioners

We conducted interviews with three senior managers and five senior practitioners. The senior practitioners consisted of two mental health nurses, a psychiatrist, a social worker and an OT.

Local context: Languedoc

Local developments in the period immediately prior to data generation included the restructuring of community services and changes to the interprofessional composition of teams. Assertive outreach and early intervention, once provided through separate teams, had both been brought back into CMHTs. Social workers had been withdrawn from teams in order to concentrate exclusively on eligibility assessments, personalisation and personal budgets. A senior manager explained how (as a consequence) care co-ordinators are generally:

a registered nurse, a doctor, or a qualified OT. Very rarely you would find a social worker, because under the Section 75 agreement then they’re no longer expected to be a care co-ordinator.

L-SM-001

Budget cuts had led to the loss of up to five support worker posts, and the workload of those remaining was said to be increasing. In this context of retraction and reorganisation, staff morale was described as low.

Mixed views were expressed on values and goals: participants talked of recovery, individualised care, respect, involvement and having a community focus. They also talked of making a difference and the ‘6Cs’ in the post-Francis report129 era. The ‘6Cs’ consist of care, compassion, courage, communication, commitment and competence and are the central components of the Compassion in Practice paper as articulated by the Chief Nursing Officer. 130 Where one saw continuous improvement, others spoke of managers concerned with reputation, targets and protecting their jobs, and of national policies and aspirations having little impact on practice locally. A senior manager said how, at a management level, ‘the values and principles would all be very much around meeting targets, protecting the reputation of the trust . . . I don’t want to sound too cynical, but it’s difficult not to’ (L-SM-002). The trust was variously described as ‘working within a business model’ (L-SP-005), and as being paternalistic and medically dominated. Other views were that quality varied and that some people received a service which they would not if they lived elsewhere.

Care planning and care co-ordination

The CPA was used to organise care for people with serious mental illnesses and complex needs, and attracted a variety of views. It was described as a structure for making sure that needs were assessed and a recovery-focused programme of care offered and reviewed, overseen by a care co-ordinator who had a good relationship with the service user. For example, a senior manager said how the CPA was:

to ensure that people are offered a needs-based recovery focused programme of care that meets their individual needs, their family’s needs. And that spans all different types of care that they might need.

L-SM-001

A senior professional said how the CPA has a:

number of features, these would include a named care co-ordinator, having regular reviews and having a written care plan, which is normally shared with all the people, all the agencies, who are delivering care to this person.

L-SP-001

The CPA was also talked of as a monitoring system to reduce the risk of serious incidents, and as being different things to different people. Distinctions were drawn between what the CPA was and what it could be, with patchy implementation across the locale.

In the trust, a general structured assessment tool was used, with the specific assessment of risk said to be more important than other areas (e.g. crisis and contingency planning). A senior manager said how:

the thing that we score best on in . . . audits is the risk assessment, simply because people think that that’s more important . . . and to some extent they’re probably right.

L-SM-002

Information technology systems across the trust and LA were not compatible, and the trust’s package was said to lack functionality, forcing staff also to use paper records. Staff were unable to print documents for people in their own homes, and one senior professional spoke of a need to make the CPA less service-centred and more like the user-owned approach found in maternity care. PbR and clustering attracted different views. One view, from a senior manager (L-SM-001), was that it helped the CPA with costing and quality. Another, from a professional, was that the clustering tool increased financial pressures, reduced contact time between service users and practitioners and used out-of-date language: ‘words fail me on that one, how they’ve used this tool that’s about . . . 25, 30 years old, the wording in it is appalling’ (L-SP-001).

Local authority participants saw the CPA as something done by health staff within the trust. The CPA was not something that social workers were involved with, as following local arrangements they were not expected to fulfil the role of care co-ordinator. Caseloads were said to be rising but different estimates were given of caseload numbers, with an upper figure of 40 cited for nurses. They, plus OTs and psychiatrists, were the professionals most likely to fulfil the care co-ordinator role. Becoming a care co-ordinator reflected the service user’s presentation and assessment, but also local capacity. Some described the CPA as medically dominated, leading to poor teamwork and poor reviews, in which people’s voices were not heard. A senior professional said how the CPA is ‘weighted very heavily [in favour of doctors]’, and how ‘everybody in those CPA meetings should have the opportunity to have their point of view. They’re not meetings where the doctors give their orders out and say what they want to have happen’ (L-SP-002).

Reviews were usually 6-monthly and organised around the diaries of psychiatrists, who also determined the precise frequency of meetings and (according to one participant) vetoed the involvement of some people. Meetings sometimes happened in people’s homes, and participants talked of experimenting with video chat technology as a way of involving service users. Carers and people working outside of statutory services were usually invited, and one participant expressed a view that carers should be at reviews even where the service user did not want this. Others talked of tensions around confidentiality and consent, with carers’ assessments routinely offered.

Training involved a mandatory day (which was either every 1 or 3 years) addressing the CPA, clinical risk, diversity and electronic systems. One senior professional talked of having no specific care co-ordination training, and one saw care co-ordination as an expected and everyday part of the work of a nurse.

Recovery

Broad understandings were given of a recovery approach. This was described as highly personalised and individual, and about valuing strengths, independence and goals. A senior manager emphasised the change from traditional services this demanded ‘when everybody started to talk about recovery, people thought it’s something that we do anyway . . .I don’t think that there’s always been appreciation of the depth of it’, and how ‘people in some respects underestimated the level of investment that you have to make as an organisation and as a practitioner to work in a true recovery focused way’ (L-SM-001).

Evidence of recovery ideals in action cited by participants included co-produced care plans, some use of WRAPs and plans for a Recovery College. Resistance to recovery was also discussed. Participants talked of conservative medical practices, partly in response to work pressures and worries over risk and blame. At the highest level, managers were said to lack ‘passion’ for changing established ways of working, and at least some practitioners and service users were said to want to continue working in established ways. One senior manager said that:

The principles that are behind the recovery model are the same principles that are behind a lot of stuff that AMHPs [approved mental health professionals] and social workers and people would do anyway, and most decent nurses and OTs.

L-SM-002

In the case of tools (e.g. the Recovery Star), concerns over additional paperwork were given as reasons for their underuse. A view given by some participants was that if service users have capacity and decline therapies or other interventions then there is little else that can be offered.

Personalisation

There was a general awareness of personalisation being about people taking more control over their lives. Service users setting their own outcomes was cited as a way of helping this, but the process for achieving this could be relatively superficial (e.g. asking users to provide signatures). A senior professional view was that what people did was personalised because ‘I think it is by its very nature, in what we do’, and that ‘care planning is very personalised in itself, specific to each person’ (L-SP-005). Participants also pointed to the large numbers of people who the trust could not be sure had received copies of their care plan. Another view was that personal budgets brought about greater personalisation. The set up of these formed a large part of the work of social workers under local arrangements, who were said to have specific targets to meet. Varying estimates were given on the uptake of personal budgets, and requests for service users to make contributions to their own care were given as reasons why some services were turned down. Personal budgets were also seen as difficult to deliver, being surrounded by administrative processes described by one professional as ‘absolutely dire [because they turn a] simple process . . . into an absolute mass of paperwork [. . .]’ (L-SP-002).

Suggestions for improvements

Leadership was seen as important, along with cultural change. This was said to require a move away from a predominantly medical, paternalistic approach towards one which was more oriented towards recovery, personalised care and service user involvement. Training for staff on personalisation and recovery was said to be needed, along with helping service users to also better understand these concepts. Less bureaucracy was asked for. Challenges included staff inertia and budget cuts, and staff feeling that jobs were insecure.

Narrative summary of interview data: service users, carers and care co-ordinators

We conducted interviews with six service users, two carers and six care co-ordinators in Languedoc.

Care planning/co-ordination

There was a general awareness among service users and carers of care being planned and co-ordinated. For most, care plans were felt to be useful as a means of sharing information, recording key details and keeping track of progress, as well as for recording important contacts (e.g. emergency numbers). For service users, care plans themselves were deemed to constitute only a small part of the care planning process, with more priority and significance attributed to relationships and medication, and were thus felt to be relatively unimportant. Carers agreed that regular CPN visits were especially valuable. It was also noted that care plans sometimes fail to meet all relevant needs and may not cover specific elements such as contingency arrangements.

There was a similar emphasis among care co-ordinators upon the usefulness of care plans in terms of ‘keeping track’ – not only of service user individuals, but also of public safety. Efficient assessment and individualised care planning with a view to meeting personal needs were also considered to be of central importance. Interestingly, the information sharing practices foregrounded by service users and carers were, for care co-ordinators, associated with overwhelming quantities of paperwork and time pressures, which can result in rushed work. It was also acknowledged by this group of respondents that some service users see little value in care plans and are sometimes disinterested in them. Observations were made, too, about the difficulties sometimes experienced in finding ‘common ground’ between care co-ordinators and service users, which can be a hindrance to good care planning.

Levels of engagement with care plans varied widely among service users, although this was apparently less of a focal point for carers, who raised few concerns. Some service users had been given care plans whereas others had not; some took little notice of them, whereas others found the scrutiny of their engagement with their own personal details uncomfortable: ‘like being under a microscope’ (L-SU-003). It was generally felt that care plans are primarily important and useful for staff – something that staff are obliged to do. Service users and carers were keenly aware of this and agreed that the main function of the document was to facilitate effective intra-staff communication and to demonstrate that staff are able to maintain a clear picture of what is happening at any given time. Despite care co-ordinators’ insistence that involvement is the most important aspect of care planning, and that service users should own their care plans, they nevertheless recognised that this is not always the case in practice. One care co-ordinator said, ‘I think if it isn’t well understood it can just become a paper exercise and also what sometimes makes it not work is the pressures of the job, which a lot of the times is about, well meeting targets and it becomes more like a tick box sometimes’ (D-CC-002). Reflecting service users’ views, one respondent described the care plan document as being owned by care co-ordinators and written for service users, rather than truly belonging to them. Care co-ordinators also suggested that the level of involvement likely to be experienced by service users was partly dependent upon the extent to which their needs and expectations were seen as ‘realistic’ (L-CC-002). In such circumstances, there was said to be scope for significant input.

Service users described mixed experiences of involvement, along a continuum from fully involved, to good collaboration, through to no input at all. One example was given in which a service user was given scope to write their own care plan, but then staff added additional or missing details on their behalf. Some respondents observed that there was a general shift towards more collaborative work, although this was clearly not a universal experience. One respondent reported that their existing care plan had changed very little in the last decade. Although the paper format was preferred overall, most service users admitted that once the care plan is written, they do not refer to it again, and it was often seen as a piece of paper that they did not take particularly seriously. Similarly, carers used care plans primarily as an aide memoire, as a summary of meetings, or around review times, and did not feel that care plans functioned as ongoing ‘active’ documents:

You can give all the copies [of a care plan] out in the world, you know, but if it doesn’t mean anything in reality. You can write anything on a piece of paper but if nothing is getting done practically, it doesn’t mean anything.

L-CA-002

There was some preference among carers for more comprehensive care plans, and the notion of a portfolio of consecutive care plans (as a means of following progression more closely and incorporating updates more clearly) was also mooted. Care co-ordinators were also of the opinion that care plans are more a function of meetings than they are an active day-to-day document. Recognition was made of the fact that, as a consequence of this, omissions or oversights tend to be identified and addressed only when reviews are carried out.

Care co-ordinators raised concerns, too, about the existing format and structure of care plans, with most of the opinion that these could be shorter and simpler. The time-consuming nature of care plan documentation was mentioned:

Paperwork for instance takes probably up to like 75% of the care co-ordinator’s time . . . you probably see somebody here for an hour and [the paperwork] . . . ends up taking probably half of your day and that then is not useful at all . . . [because] that time you could have been using to see other people.

L-CA-002

It was suggested that any information included solely for the purpose of meeting targets – rather than as a means of enhancing service users’ experiences – ought to be removed from the document altogether. On the whole, service users seemed to understand their care plans, although it was mentioned that a clearer explanation of some headings might be beneficial. Care co-ordinators agreed that a simplification of existing paperwork would benefit all involved, and would also make documentation more accessible for all readers. Feelings were mixed about the usefulness of alternative (e.g. electronic) formats; this was seen to be dependent upon the skills and abilities of individual service users, but not thought to be suitable for all. Care co-ordinators were ambivalent about the IT systems used in their work. Assessments were described as comprehensive (despite being a little repetitive), and respondents complained that they had limited access to two of the systems used which, in turn, are not mutually compatible.

There was noticeable uncertainty among service users over whether or not issues around safety and risk had been assessed, discussed or incorporated into their care plans, and these did not seem to be major areas of concern here. Only one respondent described having a very helpful crisis and contingency plan and felt confident about using this as and when needed. Concerns were also raised about the sensitivity of risk-based discussions in a home visit environment where there are children present. Carers were more confident that safety and risk had been adequately considered by care co-ordinators, although they acknowledged that this was not always reflected in the care plans themselves. One respondent also noted that the existing risk assessment does not include provision for admission to hospital, or any contingency plan, despite care co-ordinators’ emphasis upon the importance of such inclusions. In contrast, risk assessment was felt to be a priority for care co-ordinators, and risk was understood to be more important from a practitioner’s perspective in general. ‘They’re always paramount . . . [my risk] radar is always very high up, yeah, and I always take risk very seriously’ (L-CC-002). Risk is assessed on an ongoing, continual basis, and care co-ordinators described a willingness to prioritise these, sometimes to the detriment of care plans or well-being plans. The main concerns raised in this context were in relation to blame, and the consequences of this for staff. Care co-ordinators also acknowledged the difficulties inherent in sharing the full details of risk assessments with service users, who would struggle with knowing exactly what information was held by the staff.

For service users and carers in particular, the relationship with their care co-ordinators was felt to be fundamentally important. One service user said:

[CPN is] supposed to see you every fortnight but . . . Sometimes they’re not in for weeks so you might not see anyone for a month or six weeks . . . My care co-ordinator is part time anyway, so she’s very, very busy all the time, and when she does come to see you, she’s literally, sits there, and she’s not even in my house ten minutes . . . I can’t tell her how I’m feeling or, and get into an in depth conversation that, in couple of minutes. You’re not programmed to talk like that. You have to feel, you know what I mean?

L-SU-004

Although care co-ordinators seemed to be less emphatic about this, they were very concerned about having to compromise time spent with service users in the face of ever increasing administrative tasks. There was also some discussion of a harsh compromise between demands for standardisation on one hand, and individualised care planning processes on the other.

The fluidity of care planning processes, frequent changes of care co-ordinator, and interruptions in continuity more widely, were cited as challenging by all three categories of respondent, with most emphasising the disruption that this creates. Similarly, some respondents felt insecure about the possibility of having support systems taken away as a consequence of structural and organisational changes; this was felt to have a negative impact on their sense of control over their care planning. Many felt that such issues could be better managed, ensuring that contact and support are actively maintained at all times, rather than having to be pursued by service users and carers during periods of change. Carers also recounted problems with continuity and high care co-ordinator turnover, and some expressed a wish for greater recognition of their role in relation to – and the scope of their relationships with – service users. Concerns were raised about feelings of isolation for some service users under community care. Better communication between care co-ordinators and GPs was also desired, and this was mirrored in care co-ordinators’ suggestions for better connections with physical health services via the use of well-being plans. There was some doubt expressed over whether or not GPs actually read the care plans that they are sent.

Recovery

Definitions of recovery varied across and within all categories of respondent, with one service user stating that the concept had not been discussed with them at all. There was, however, general acknowledgement of the complexities involved. Some service users rejected the notion of recovery altogether, although this was typically where it was understood in relation to the idea of a medical ‘cure’. Most respondents associated recovery with managing illnesses and symptoms, and (re-)establishing a satisfactory quality of life:

What I understand now, it’s for managing my illness. Recovery also involves knowing when you’re at risk, finding support, or, it’s hard to deal with it. And being able to access professionals who will be able to help me deal with new things.

L-SU-001

Although service users and care co-ordinators mentioned hope as a significant factor – it was also suggested that this would be a preferable term – this was not discussed by carers, who were in turn more focused upon the possibility of regaining independence, and who emphasised that only gradual progress ought to be expected. The ‘journey’ metaphor was mentioned only by service users, but this was not the dominant definition. Care co-ordinators also mentioned that difficulties may arise in situations where practitioners’ views of achievable progress or achievements sit in conflict with service users’ personal views of the same.

Service users at this site were unsure about the impact of recovery on care planning, and some did not make a clear connection between the notion of recovery and strategies of coping and self-management. Their relationships with care co-ordinators and/or support workers were instead felt to be very important in terms of recovery. For care co-ordinators, the recovery-oriented approach has resulted in a greater focus in care plans upon specific goals, as well as more of a drive towards referral and discharge; this was not seen to be a positive development. In general, care co-ordinators described working in a recovery-focused way; there were mixed feelings on whether or not procedures have changed as a consequence of this, although some respondents did feel that care planning had become less prescriptive and more collaborative overall. One care co-ordinator tried to explain the desire to provide a personal, responsive approach:

The definition of recovery from A to B is different. So, and as such you have to have the resources to tailor the care you are providing to meet that very aspiration of that very particular person . . . it’s an issue you have to keep on reminding yourself [about].

L-CC-006

There was some suggestion among care co-ordinators that the concept of recovery was perhaps for the benefit of organisational goals, as well as a sense that the intention for new clients is to move them on through services and reduce their level(s) of dependence.

Not all service users use recovery plans. Some respondents had not heard of WRAPs, whereas others had used them at various points, even if not continuously. Only one respondent felt that the Recovery Star had been helpful with care planning. Similar inconsistencies emerged in carer responses, although several did feel that WRAPs would, in principle, be useful. Carers also recognised the period immediately following hospital discharge to be especially challenging in relation to the formulation of recovery plans. There were also mixed responses among care co-ordinators; some indicated that WRAPs were used only occasionally, whereas others maintained that service users were encouraged to develop them.

All care co-ordinators emphasised the importance of taking a strengths-based approach to care planning, and this was mirrored in the responses of most respondents across the three categories. Carers did, however, note that the recognition of strengths and abilities does not necessarily mean that these will be recorded in care plan documents per se.

Among the criticisms made by service users of care received were the side effects of medication, which have resulted in weight gain and a loss of creativity as a consequence of adhering to the ‘chemical cosh’ (L-SU-001). Some also expressed sadness at being unable to regain the life that they had lived pre-illness. Carers were especially concerned about support around medication, noting that amendments to this can be traumatic, such that service users sometimes remain on unsuitable prescriptions, fearful of the possible consequences of any changes. For care co-ordinators, meanwhile, long waiting lists (especially for psychological therapies) and inconsistent levels of service-user engagement were cited as hindrances to the delivery of recovery-focused care. A perceived lack of organisational support for recovery, and an overemphasis on meeting targets, was seen as similarly challenging in this respect.

Most service users and some carers and care co-ordinators spoke about help being provided to access local voluntary services, day centres and activities and there seemed to be quite a lot available. One service user attended an art group and pottery class and said:

We get a lot of support there through other service users and that from that [which is helpful because] you get your distraction, but there’s also, from whatever you’re doing but there’s also people, other people there that you can keep in contact with . . . You just keep going, you just meet people, you start talking and I’ve still got friends from when I was on the day ward in 2009.

L-SU-005

There was some confusion among carers and care co-ordinators about what was available, and the need for an updated directory was mentioned: ‘Yeah we could do with a directory really of everything that’s around but that’s updated and I mean in real time, really, but things are changing that much, what’s here today might not be here next month’ (L-CC-006).

Personalisation

There was a significant degree of ambiguity around the term, with no unequivocal definition provided, although almost all respondents linked this to the notion of individualised care plans, rather than generic ones. This was the case even for care co-ordinators, some of whom – despite having received dedicated training – were not familiar with the term and others of whom had no clear sense of its meaning. Both service users and care co-ordinators also drew links between personalisation and ownership or ‘control’ of care planning, although this was not mentioned by carers. Most respondents across the three categories described their existing care as adequately personalised, with care co-ordinators in particular maintaining that service users’ needs are heard, and that that their contributions are recognised. There is much emphasis among care co-ordinators upon the likelihood of service users having ‘unrealistic’ expectations, which may not be matched by, for example, the availability of therapies or hospital beds. There was also some mention of the workload involved:

I suppose often care co-ordinators, myself included, sometimes we get slightly panicked when people talks about personalisation because we think, oh shit, God that’s going to generate a lot of work. And therein lies the problem that’s fundamentally wrong with the system.

L-CC-004

Embedded case-study comparisons

Detailed case-study comparisons were made for six cases consisting of two triads of service user, carer and care co-ordinator; and four of service user and care co-ordinator only. Tables were drawn up for each of the themes used for our framework analysis; an example is presented in Appendix 9.

Provence

Provence Trust provides both community and inpatient mental health services to a population of around 1.5 million people. The catchment area is predominantly rural with some urban localities, within which are provided specialist services. According to the 2011 census, > 90% of the population are from White British backgrounds. The EID reported that some of the areas within the catchment are affluent and among the least deprived 10% of areas in the country. Adult inpatient services are provided from six hospital sites, with approximately 30 CMHTs and 225 adult psychiatric beds, of which 34 beds are allocated to the principal CMHT that data were collected from.

Service users

One-way ANOVAs of all subscales were conducted and subsequent Tukey’s post-hoc tests demonstrated some areas of significant differences between research sites on the STAR-P measure of therapeutic relationships. There were significant differences across sites in the mean total STAR-P score [F(5,429) = 3.45; p-value = 0.005], the positive collaboration subscale [F(5,426) = 3.75; p-value = 0.002] and the positive clinician input subscale [F(5,431) = 2.80; p-value = 0.017]. There are no global differences across the sites for the RSA scale and ES. Table 39 shows the mean item scores, alongside the parameters of significance.

Burgundy in particular performs well on the STAR-P scale (Figure 4). If Burgundy is used as a reference site, other research sites may be considered in relation to this. Languedoc also performs well on the perceived therapeutic relationship; however, Provence and Dauphine do not perform as well on this scale. Tukey’s post-hoc tests revealed that service users in Burgundy (38.49 ± 8.55) score higher on the mean total STAR-P than those in Provence (32.33 ± 11.91; p-value = 0.003). Burgundy also performs particularly well in comparison to perceptions of therapeutic relationships in Dauphine; however, this does not reach statistical significance (33.53 ± 9.23; p-value = 0.056).

FIGURE 4.

Mean positive collaboration subscale score for service users ± 95% confidence interval. Scoring range for the scale from 0–24. **p-value < 0.01.

Subscales of the Scale to Assess the Therapeutic Relationship

Tukey post-hoc tests revealed that service users in Burgundy (19.81 ± 6.57) score higher on the ‘Positive Collaboration’ subscale than those in Provence (16.15 ± 6.57; p-value = 0.002) and Champagne (17.13 ± 5.79; p-value = 0.062); however, this does not reach significance. There were no significant differences between Burgundy and the other sites on this subscale. Figure 5 shows the mean positive collaboration subscale with Burgundy as a reference site to the other sites.

FIGURE 5.

Mean total STAR-P score for service users ±95% confidence interval. Scoring range for the scale from 0–48. **p-value = 0.01.

Further Tukey post-hoc tests revealed that responses on the ‘Positive Clinician Input’ subscale are different across the research sites. The reference site, Burgundy (9.46 ± 2.45), performs significantly better than Provence (7.83 ± 3.49; p-value = 0.012) and Champagne (8.01 ± 3.05; p-value = 0.45). Burgundy and Champagne are the two Welsh research sites, which suggests that the differences within this subscale may be appearing at a local level rather than as a result of macro-level policy and the implementation of CTP and processes surrounding it (Figure 6).

FIGURE 6.

Mean positive clinician input subscale score for service users ± SEM. Scoring range for the scale from 0–12. **p-value < 0.01.

Care co-ordinators

One-way ANOVAs were conducted for the mean RSA total score and the five RSA subscales for the care co-ordinators (Table 40). There was a significant difference between the research sites in the ‘Choice’ subscale [F(5,195) = 3.40; p-value = 0.006]. There were no significant differences found in the mean total RSA score and the other four subscales; however, the ‘Diversity of treatment options’ subscale was approaching significance [F(5,195) = 2.10; p-value = 0.068].

TABLE 40 - Summary score statistics for the care-co-ordinator responses to the RSA scale, STAR-P and ES

**p-value < 0.01.

SEM, standard error of the mean.

RSA scale	One-way ANOVA parameters	Artois	Burgundy	Champagne	Dauphine	Languedoc	Provence
		Mean (SEM)	Mean (SEM)	Mean (SEM)	Mean (SEM)	Mean (SEM)	Mean (SEM)
Life goals	F(5,195) = 0.71; p = 0.617	3.68 (0.12)	3.73 (0.11)	3.79 (0.09)	3.54 (0.15)	3.82 (0.11)	3.70 (0.09)
Involvement	F(5,195) = 0.98; p = 0.429	3.01 (0.13)	2.91 (0.11)	2.92 (0.13)	2.99 (0.15)	3.23 (0.12)	2.87 (0.10)
Diversity of treatment options	F(5,195) = 2.10; p = 0.068	2.96 (0.14)	3.23 (0.13)	2.94 (0.13)	2.98 (0.16)	3.24 (0.12)	2.74 (0.09)
Choice	F(5,195) = 3.40; p = 0.006**	3.76 (0.10)	3.92 (0.11)	3.70 (0.11)	3.46 (0.10)	4.04 (0.10)	3.58 (0.13)
Individually tailored services	F(5,195) = 1.74; p = 0.126	3.18 (0.13)	3.10 (0.13)	3.11 (0.13)	3.49 (0.15)	3.42 (0.15)	3.42 (0.13)
Mean total score	F(5,195) = 0.997; p = 0.421	3.35 (0.11)	3.41 (0.10	3.35 (0.11)	3.31 (0.13)	3.57 (0.11)	3.25 (0.08)

When using Dauphine as the reference site (the site with the lowest score on this subscale), it is apparent that other sites are performing significantly better around the subscale of ‘Choice’ (Figure 7).

FIGURE 7.

Mean item response ‘Choice’ subscale score for care co-ordinators ±95% confidence interval. Scoring range on a scale of 1–5. *p-value < 0.05, **p-value < 0.01.

This scale measures some important perceptions that may have a significant effect on patient outcomes, concordance to care and collaboration with clinicians. Subsequent Tukey’s post hoc revealed that service users in Dauphine (3.46 ± 0.68) score significantly lower on the subscale than Languedoc (4.04 ± 0.50; p-value = 0.008) and Burgundy (3.92 ± 0.65; p-value = 0.041). There were no significant differences between Dauphine and the other sites on this subscale. A summary table for the five highest scoring items on the RSA scale for each site is included in Table 41 for the service users and Table 42 for the care co-ordinators.

TABLE 41 - Summary table for the five highest rated items on the RSA scale by service users

Rank	Artois (mean of 3.84–4.38 on Likert scale)	Burgundy (mean of 3.86–4.34 on Likert scale)	Champagne (mean of 3.69–4.65 on Likert scale)	Dauphine (mean of 3.67–4.64 on Likert scale)	Languedoc (mean of 3.67–4.64 on Likert scale)	Provence (mean of 3.49–4.11 on Likert scale)
1	Choice Staff do not use threats, bribes or other forms of coercion to influence my behaviour or choices	Choice Staff do not use threats, bribes or other forms of coercion to influence my behaviour or choices	Choice Staff do not use threats, bribes or other forms of coercion to influence my behaviour or choices	Choice Staff do not use threats, bribes or other forms of coercion to influence my behaviour or choices	Choice Staff do not use threats, bribes or coercion to influence my behaviour or choices	Choice Staff do not use threats, bribes or other forms of coercion to influence my behaviour or choices
2	Choice Most of my services are provided in my natural environment (i.e. home, community, workplace)	Life goals The role of staff is to assist me and other people in recovery with fulfilling my individually defined goals and aspirations	Life goals The role of staff is to assist me and other people in recovery with fulfilling my individually defined goals and aspirations	Choice Most of my services are provided in my natural environment (i.e. home, community, workplace)	Choice Most of my services are provided in my natural environment (i.e. home, community, workplace)	Choice Most of my services are provided in my natural environment (i.e. home, community, workplace)
3	Life goals The role of staff is to assist me and other people in recovery with fulfilling my individually defined goals and aspirations	Choice Most of my services are provided in my natural environment (i.e. home, community, workplace)	Choice Staff listen to and follow my choices and preferences	Life goals The role of staff is to assist me and other people in recovery with fulfilling my individually defined goals and aspirations	Life goals The role of staff is to assist me and other people in recovery with fulfilling my individually defined goals and aspirations	Life goals Staff believe that I can recover and make my own treatment and life choices
4	Individually tailored My service provider makes every effort to involve my significant others and other sources of natural support in the planning of my services, if this is my preference	Life goals Staff use a language of recovery (i.e. hope, high expectations, respect) in everyday conversations	Choice Staff help to monitor the progress I am making towards my personal goals on a regular basis	Life goals Staff believe that I can recover and make my own treatment and life choices	Life goals Staff believe that I can recover and make my own treatment and life choices	Life goals The role of staff is to assist me and other people in recovery with fulfilling my individually defined goals and aspirations
5	Life goals Staff believe that I can recover and make my own treatment and life choices	Choice Staff help to monitor the progress I am making towards my personal goals on a regular basis	Life goals Staff are knowledgeable about special interest groups and activities in the community	Choice Staff listen to and follow my choices and preferences	Choice Staff help to monitor the progress I am making towards my personal goals on a regular basis	Life goals Staff are diverse in terms of culture, ethnicity, lifestyle and interests

TABLE 42 - Summary table for the five highest rate items on the RSA scale by care co-ordinators

Rank	Artois (mean of 3.97–4.22 on Likert scale)	Burgundy (mean of 4.14–4.43 on Likert scale)	Champagne (mean of 4.07–4.31 on Likert scale)	Dauphine (mean of 3.79–4.23 on Likert scale)	Languedoc (mean of 4.21–4.79 on Likert scale)	Provence (mean of 3.91–4.18 on Likert scale)
1	Life goals Staff believe that people can recover and make their own treatment and life choices	Choice Staff do not use threats, bribes or other forms of coercion to influence the behaviour or choices	Choice Staff do not use threats, bribes or other forms of coercion to influence the behaviour or choices	Choice Staff do not use threats, bribes or other forms of coercion to influence the behaviour or choices	Choice Staff do not use threats, bribes or coercion to influence the behaviour or choices	Life goals Staff believe that people can recover and make their own treatment and life choices
2	Choice Staff do not use threats, bribes or other forms of coercion to influence the behaviour or choices	Choice Progress made towards goals (as defined by the person in recovery) is monitored on a regular basis	Life goals Procedures are in place to facilitate referrals to other programmes and services if the agency cannot meet a person’s needs	Life goals Staff are diverse in terms of culture, ethnicity, lifestyle and interests	Choice Most services are provided in a person’s natural environment (i.e. home, community, workplace)	Life goals The role of staff is to assist a person with fulfilling their individually defined goals and aspirations
3	Life goals The role of staff is to assist a person with fulfilling their individually defined goals and aspirations	Life goals The role staff is to assist a person with fulfilling their individually defined goals and aspirations	Life goals Staff believe that people can recover and make their own treatment and life choices	Life goals Staff believe that people can recover and make their own treatment and life choices	Life goals The role of staff is to assist a person with fulfilling their individually defined goals and aspirations	Life goals Staff actively assist people in recovery with the development of career and life goals that go beyond symptom management and stabilisation
4	Individually tailored Every effort is made to involve significant others and other natural supports in the planning of a person’s services, if so desired	Choice Most services are provided in a person’s natural environment (i.e. home, community, workplace)	Life goals Staff are knowledgeable about special interest groups and activities in the community	Life goals Staff routinely assist individuals in the pursuit of their educational and/or employment goals	Life goals Staff believe that people can recover and make their own treatment and life choices	Life goals Staff routinely assist individuals in the pursuit of their educational and/or employment goals
5	Choice Most services are provided in a person’s natural environment (i.e. home, community, workplace)	Life goals Staff believe that people can recover and make their own treatment and life choices	Individually tailored Every effort is made to involve significant others and other natural supports in the planning of a person’s services, if so desired	Life goals Procedures are in place to facilitate referrals to other programmes and services if the agency cannot meet a person’s needs	Diversity of treatment options People in recovery are given the opportunity to discuss their sexual and spiritual needs and interests	Choice Staff do not use threats, bribes, or other forms of coercion to influence the behaviour or choices

Adjusted analysis

The advantage of using unadjusted ANOVAs to examine the data is that this retains the maximum number of participants in the analysis; the disadvantage is that case-mix differences between the sites could potentially bias the findings. In contrast, the strength of using ANCOVAs is that they adjust for the potential confounders but this is achieved at the cost of losing participants in the analysis owing to missing data on the covariates. This reduces the statistical power of the analysis and changes the composite of the sample, which poses a problem for the interpretation. Analysing the data using unadjusted and adjusted analyses offers a check on the robustness of the original (unadjusted) findings.

The ANOVAs reported above are not adjusted for potential confounders (i.e. variables that are potentially associated with the outcome variables and which may be differentially distributed across sites). For service users, three demographic variables (age; gender; ethnicity) and two care-related variables (relationship status; time in mental health services) were identified as potential confounders. These variables were used as covariates in a series on ANCOVAs to determine whether or not they substantively change the findings from the unadjusted analyses. For care co-ordinators, three demographic variables (age, gender and ethnicity) and two measures of clinical experience (time working in mental health services and time as a care co-ordinator) were identified as potential confounders for a further series of ANCOVAs. To determine whether the adjusted findings were different from the original findings, we compared the p-values for the omnibus ANOVA/ANCOVA, the p-values for the post-hoc comparisons, and the unadjusted and adjusted means for each scale and subscale for each site.

Service users

For the service users there were no substantive differences in the adjusted analysis for any of the subscales or the total scores of the RSA scale and the ES. Similarly, there were no substantive differences for two of the three STAR-P subscales or the total STAR-P score. However, for the ‘Positive clinician input’ subscale of the STAR-P there were minor differences in the adjusted means and this resulted in the difference between Burgundy and Champagne becoming non-significant (p-value = 0.285). The observed difference between Burgundy and Provence in the unadjusted analysis remained significant in the ANCOVA (p-value = 0.040).

Overall, the ANCOVA analyses for service users show that, although there were marginal changes in the adjusted means, these changes were not substantive and did not affect the overall pattern of findings. Consistency between the adjusted and unadjusted analyses suggests that the findings are robust.

Care co-ordinators

For the care co-ordinators, two of the five subscales of the RSA scale suggested no substantive differences in the adjusted analysis. However, for the ‘Choice’ subscale, some previously non-significant post-hoc comparisons become significant in the adjusted analyses. Figure 8 shows that the difference between Burgundy and Dauphine and the difference between Languedoc and Dauphine remained significant, but, in addition, Burgundy (4.00 ± 0.12) and Languedoc (4.07 ± 0.14) were now also significantly different from Provence (3.44 ± 0.14) (p-value = 0.034 and p-value = 0.024, respectively). As the graph shows, these new differences emerged because the adjusted mean for Provence was marginally reduced and became closer to the reference site of Dauphine. It should be noted that the changes in the means are marginal and that the overall pattern of findings remains essentially the same.

FIGURE 8.

Adjusted mean item response on Choice subscale score for care co-ordinators ± 95% confidence interval. Scoring range on a scale of 1–5. **p-value < 0.01. a, identifies a significant difference compared with Dauphine; b, identifies a significant difference compared with Provence.

For the total RSA score, changes in the adjusted means resulted in an emergent significant difference between Dauphine (2.99 ± 0.14) and Languedoc (3.65 ± 0.13; p-value = 0.019) (Figure 9a). For the ‘Diversity of treatment options’ subscale the difference between Dauphine (2.84 ± 0.86) and Burgundy (3.24 ± 0.78) became statistically significant (p-value = 0.035) (Figure 9b). For the ‘Life goals’ subscale, the differences between Dauphine (3.21 ± 0.15) and two other sites also became significant: Languedoc (3.92 ± 0.14; p-value = 0.024) and Champagne (3.86 ± 0.13; p-value = 0.029) (Figure 9c).

FIGURE 9.

Adjusted mean scores for care co-ordinators ± 95% confidence interval. (a) RSA total score; (b) diversity of treatment options; and (c) life goals subscale. *p-value < 0.05.

In terms of statistically significant differences between sites, the findings for care co-ordinators changed more than the findings for the service users in the adjusted analyses. Proportionally, the observed number of changes in significance was small compared with the number of post-hoc comparisons conducted, and changes in significance were predicated on relatively small absolute changes in the adjusted means across the various sites and outcome measures. The vast majority of the post-hoc comparisons did not change in terms of statistical significance in the adjusted analyses. The findings from the ANCOVAs therefore are suggestive of potential site differences on particular measures between specific sites, but these findings should be interpreted cautiously as there were no specific a priori hypotheses about site differences, and the loss of participants as a result of missing data on the covariates changes the sample composition.

Correlations between the outcome measures

Pearson’s correlations were completed at the global level with all participants to determine if there were associations between the responses on the questionnaire scales. Table 43 shows that there is a strong positive correlation between the RSA scale and the STAR-P (r = 0.61, N = 409; p-value < 0.001). There is an association between the recovery-oriented focus and ratings of the quality of therapeutic relationships among service users. There is a weak relationship between the RSA scale and ES and a negligible relationship between the STAR-P and the ES.

To explore the pattern of the higher correlation between the RSA scale and STAR-P, subsequent correlations were completed at the subscale level. There were positive associations found between the ‘positive collaboration’ and ‘positive clinician input’ subscales and the five RSA subscales (Table 44). There were small to medium correlations between the ‘non-supportive clinician input’ subscale and the five subscales of the RSA scale. This is expected, as this subscale is a negatively framed subscale and, therefore, it is less likely to correlate with the RSA scale.

Although it appears that these questionnaire scales have a strong positive correlation, the results should be interpreted with caution. There is the possibility that a ceiling effect in the data may have negatively skewed the data. When assessing the correlations by trust/health board it is apparent that there is some variability (Table 45).

The RSA scale and STAR-P demonstrate the strongest and most consistent correlations compared with the overall analyses. There is, however, some variability across sites. All correlations demonstrate at least medium effects, and most are classed as large effect sizes.

With the RSA scale and ES, the sites are demonstrating greater variability, with some sites showing medium effects (Dauphine, Languedoc and Provence) and other sites showing smaller effects (Champagne and Artois); one site shows no relationship (Burgundy). The discontinuity across these sites suggests that some of these scales are behaving differently across the sites; however, further exploration of this is beyond the scope of this report.

The STAR-P and ES show less variability, with all sites showing a small to moderate association.

Local context and developments

Across all six sites there was an acknowledgement that services had been or were still undergoing a significant process of restructuring. The nature and impact of this period of flux appeared to vary, in particular between the two countries but also in relation to the integration of health and social care. Overall, there appeared to be moves towards greater integration of health and social care staff within teams, with colocation of NHS and LA personnel. Staff in Provence in particular stressed the strong shift towards generic mental health roles. Languedoc appeared to be moving in a different direction, with social workers being removed from CMHTs to focus on social care assessments and personal budget applications. This reportedly led to higher workloads within CMHTs and lower morale; this was also the only site in which medical domination of services was cited as an issue. Greater uncertainty about developments seemed to be expressed where more adventurous, wide-ranging restructuring had taken place, such as in Provence, where services were being centralised and teams now worked across age groups. This was accompanied by strongly articulated recovery-focused values, but there were acknowledgements by senior staff that staff morale had been hard hit by the upheaval, and the anticipated gains remained uncertain.

Development of new services in primary (e.g. primary care liaison roles) and secondary (setting up or merging of assertive outreach or crisis teams with CMHTs) care were reported, often alongside concerns over the impact of cuts in budgets, reductions in social care services and the challenges of providing continuity of care and effectively managing risks across an increasing range of third sector providers under the ‘any provider’ legislation in the Health and Social Care Act 2012. 131

In the two Welsh sites, the Mental Health Measure was clearly a strong driver for developments with senior managers and practitioners in both Burgundy and Champagne extolling its virtues, although the impact on everyday practice was less clear. Integration of health board and LA managers and teams was high on the agenda, although integration in the Champagne site appeared to be frustrated by continuing differences at the most senior levels and the absence of shared and accessible IT systems. Similar frustrations with IT were articulated across all sites.

Often strong commitments to the recovery agenda on all sites appeared to be frustrated by seemingly contrary policies (e.g. CTOs) and the related administrative demands; high acuity, high demands on services, high caseloads and the inadequacies and/or inflexibility of IT systems. Moreover, other organisational targets were skewing care co-ordination work, with recovery being interpreted as through-put and discharge, targets for PbR, personal budgets and organisational prioritising of particular types of support all working to standardise rather than personalise care or place recovery at the centre of things.

Care planning and co-ordination

The core aspects of the CPA were simply and straightforwardly expressed by senior managers and practitioners in terms of assessment, care plan, care co-ordinator and regular reviews. However, there appeared to be greater difficulty in articulating the implementation of the policy. There were major challenges in trying to make the CPA meet the very different needs of service users, practitioners, managers and commissioners. It continues to be administratively heavy and is insufficiently oriented towards recovery. There was widespread reporting of and often dissatisfaction with changes to documentation and some lack of clarity over CPA eligibility. In England, there were some concerns that the introduction of clustering and PbR was resulting in care plans being less individualised (Artois, Dauphine, Languedoc).

In Wales, the Measure was seen as very important, but there was no consensus on its actual impact. Views regarding the CTP were more mixed. In both sites, it was agreed that the All Wales CTP provides the structure or a framework to work within, but it was seen by some as reductionist. In Burgundy, it was claimed that not all psychiatrists used the CTP.

The introduction of electronic CPA documentation appeared to bring some benefits in terms of accessing and sharing information, but it also brought unwieldiness and a greater distancing from service users, as working collaboratively with people on care plans in their homes became more difficult. The solution, paradoxically, may lie in the development of more integrated IT systems and the use of modern, mobile technology.

Care co-ordinator caseloads reportedly ranged from 25 (Artois) to 40 or more (Champagne, Dauphine, Languedoc) or even 50 (Provence), and in most sites these were increasingly attributable to growing demands and tighter staff budgets. The care co-ordinator role was now held by any mental health professional, most often mental health nurses, social workers and OTs, less often psychologists, and rarely psychiatrists. Artois now also employed ‘non-professional’ staff (on Band 4) as care co-ordinators, with some concerns being expressed following incidents.

The care co-ordinator role was seen as a ‘generic’ role, with staff taking on increasingly similar co-ordinator duties, but one that was also complex, carrying significant responsibilities and often tasked with delivering most of the direct care. However, this genericism contained a tension when allied with capacity pressures within teams and attempts to allocate care co-ordinators to service users based on degree of fit. The desire to be responsive and provide a more personalised approach to care co-ordination was often frustrated by the lack of capacity within a team’s or individual’s caseload. Hence, a newly referred service user that might benefit from social worker input as s/he has predominantly social care needs is allocated a mental health nurse as care co-ordinator. The co-ordination tasks are similar but the therapeutic or supportive interventions are shaped, and at times limited, by the disciplinary background of the co-ordinator. There were widespread calls for more training on the CPA and specifically for care co-ordinators.

Efforts to increase the involvement of carers were mentioned occasionally (Dauphine), but more often it was the challenges and continuing tensions around issues such as service users’ consent and confidentiality that were mentioned (Artois, Languedoc), or uncertainty around how to involve people (Burgundy). There were varied experiences of care planning and care co-ordination from the viewpoints of service users, carers and care co-ordinators across all six sites, with no distinct differences identified. At best, care is planned in a structured and collaborative way with clear communication and opportunities for service users to influence and feel some ownership of the process. At worst, service users say that they feel cast aside, not involved, or that their care is planned as an obligatory task and in ways that are confusing and rigid.

Care co-ordinators are sometimes more positive about care being planned in individualised, collaborative ways than are service users and carers. They tend to welcome collaboration, but some also worry over how to address differences in view or believe strongly that some service users, particularly those with longer contact with services, are unwilling or unable to engage in such a process.

Risk was consistently seen as central to the CPA/CTP process by senior staff across all six sites. Similarly, safety and risk were paramount for care co-ordinators, but service users and carers were hardly involved in the assessment of these and many care co-ordinators expressed concerns over sharing their views, or those of their colleagues, around issues of risk. Service users and carers were often unaware of risk assessments being documented. Perhaps as a consequence, some service users and carers are largely unaware of risk management taking place and can feel that risk is not managed sufficiently or proactively. Risk aversion limits positive risk-taking and sits at odds with recovery ideals and practices.

Care plans are useful for staff and are recognised as having to be created; they have value as records of what has been decided and for including contact details and, sometimes, crisis and contingency plans. For a handful of people care plans are like a contract. But, for the majority of people, care plans are not highly valued and are not seen as active documents; many people (care co-ordinators and service users included) do not routinely refer to care plans once they have been created. Many service users do not have care plans, or have received and quickly discarded them. Others file them in a drawer, never to look at them again. Interestingly, care co-ordinators often thought that care plans were difficult to understand, but most service users did not agree; they just thought that they were of limited utility. Some care co-ordinators mentioned the development of first person care plans but service users rarely, if ever, mentioned this. There is clearly a tension between standardisation of documentation and individualised approaches to care planning. There was some, but not much, evidence of care planning addressing service users’ strengths and accomplishments.

As has been reported previously, care co-ordinators find care planning and care co-ordination administratively burdensome and complain that they have insufficient time for face-to-face work with service users and their families. Therapeutic relationships, however, are seen as very important; along with the care co-ordinator’s skills and knowledge of the service user. Service users and carers want to see more of their care co-ordinator and recognise discontinuity (e.g. when care co-ordinators leave or when there is organisational change) as a barrier to good care. There is no clear appetite for electronic care plans: service users and care co-ordinators repeatedly reported that current IT systems (which are often incompatible within sites, and between NHS and LA services) are frequently inflexible and a hindrance to collaborative care planning.

Carers have varied experiences within sites, but many are not engaged with care planning and co-ordination. Some do not see this as problematic, instead seeing it as something that occurs primarily between the service user and their care co-ordinator.

Surprisingly little was said about review meetings, although experiences varied; at best these meetings are collaborative, with good preparation and involvement of the service user and others. Often, care co-ordinators described difficulties arranging multidisciplinary involvement and ad hoc meetings with perhaps just the service user and psychiatrist.

Across all sites, the lack of sufficient staff, resources and time to work more collaboratively with service users and carers was reported by care co-ordinators.

Recovery

Among senior personnel there was some clear articulation of what recovery and recovery-focused values might be, but varying degrees of sign-up to recovery and frequent talk of resistance among some staff. In Languedoc, for example, there was a broad understanding and evidence of a Recovery College, co-produced care plans and some use of recovery tools (e.g. Recovery Star, WRAP) but a reported lack of passion among senior managers, and conservative medical staff content to retain more traditional approaches. Similar resistance among medical staff and some older nursing staff was also reported in Provence, alongside a belief that ‘not all service users will recover’. Nonetheless, managers in Provence also cited the introduction of peer workers and greater use of ‘recovery language’.

In Artois, recovery was seen as a ‘bridge between medical and social models’ and also something ‘we have always been doing’. It was also seen as being about discharging service users reluctant to move on. Barriers to the implementation of a greater focus on recovery in Artois included the need for more staff, more time, improved IT systems and a stronger organisational commitment.

Senior staff in Dauphine acknowledged that the move towards recovery had been fairly recent and had been met by some resistance among some staff, and that the CPA alone ‘was enough’. However, there was a drive to increase the use of personal budgets, and help with obtaining employment was much more commonplace. Some recovery-focused tools were being introduced (Artois, Languedoc), but this was much less clear elsewhere (Burgundy, Champagne, Dauphine, Provence).

In Wales, the Measure had placed recovery high on the agenda for senior staff, with good understandings of recovery seen in Champagne but more mixed understandings of recovery expressed in Burgundy. For some in Champagne, the CTP did not have a clear focus on recovery and it was felt that greater training was needed for both staff and service users to bring about a change in culture.

More training around recovery working and the use of recovery tools, and perhaps a better understanding of the new ways of working to deliver recovery-focused care, were suggested for both staff and service users (Champagne, Dauphine).

As with care planning, service users, carers and care co-ordinators had varied views and experiences in relation to recovery across all sites, with the term itself often having different meanings for different people. Different views between professional groups were also mentioned. For some care co-ordinators it was even seen as unhelpful or deceptive, as it appeared to be more about discharging service users from caseloads. Across sites, many service users used terms commonly found in recovery literature, such as choice, independence, hope, fulfilment; but, for many, recovery is primarily about managing and coping better with their illness. Very few talked about recovery as a ‘journey’.

There were no clear views that care planning helped recovery, unless care plans included practical steps or helped service users to accept or talk about their mental health. Some service users and carers said that strengths were acknowledged. For service users and carers in particular, conversations and relationships, along with families and friends, were identified as being far more important than care plans in promoting recovery.

Even within single sites there is variable use of, experience in, and enthusiasm for recovery tools. At best, WRAPs are seen as very personalised and helpful, but some care co-ordinators worry about the extra work required. Some care co-ordinators say they have always worked in a recovery-focused way, and what hindered them were organisational targets and issues such as adversity to risk, documentation, limited resources and ‘firefighting’. There were also the challenges of negotiating over sometimes competing views between service users and care co-ordinators on achievable progress and managing medication side-effects, which can seriously impair recovery.

Personalisation

Across the case-study sites, personalisation was not understood as consistently as the concept of recovery among senior personnel, and definitions tended to include person-centred care plans and the use of first-person terminology in care plans (Artois); placing the person at the heart of social care (Artois); and the use of personal budgets (Artois, Dauphine) and direct payments (Champagne).

In relation to personal budgets, there was a clear lack of uniformity and this was also the one area where there was a marked difference between the two countries, perhaps reflecting the different policy emphases. Some sites saw the use of personal budgets as a key tool in the move towards a focus on recovery and personalisation of care (e.g. Dauphine, Languedoc, Provence). In both sites in Wales, there was much less emphasis on the use of personal budgets, which were seen very much as part of a creeping ‘marketisation’ of health care and, as a consequence, were subject to more resistance and a more limited move towards the use of direct payments (Champagne).

Even where personal budgets were promoted, it was recognised that they were accompanied by extremely heavy administrative loads, which took practitioners away from face-to-face contact with service users and their families (Artois, Dauphine, Languedoc). There were also often severe delays in actually receiving funds, which impacted on service users and staff (Dauphine). Elsewhere, service users often failed to meet the strict approval criteria against which they were judged, or were reluctant to make a contribution (Languedoc). Senior staff also spoke of there being tensions when service users were unwell and suggested that talk of a more personalised approach could ‘raise expectations’ that could not always be met (due to tight budgets and restricted local services and options) (Champagne).

There was also an articulation of some of the tensions that exist in the supposed move towards greater personalisation, with CTOs (Artois) and clustering cited as a counterpoint (Dauphine). Being able to purchase services from any provider also increased difficulties around sharing information and issues of confidentiality (Dauphine). Uniquely, in Languedoc, under an agreement made under Section 75 of the National Health Services Act 2006 between the LA and the trust, social workers were removed from CMHTs to focus on assessments and processing of direct payments and personal budgets. 132 The aim was to increase the move towards the use of personal budgets by service users to purchase care and support, which was seen as a key tool in the move towards personalisation of services. It appeared that, although the success of this approach was still to be determined, the impact on remaining staff within CMHTs was less than positive, with caseloads and workloads increasing with the loss of the social workers from the teams.

Commissioners of services in Provence also saw personal budgets as key drivers towards greater self-directed support. The Trust had exceeded initial targets, but senior staff spoke of the significant administrative burden. Staff also needed to develop new skills and understanding to negotiate with providers. Concerns had also emerged about the potential increased risk when dealing and communicating with multiple providers following recent serious incident enquiries. Nonetheless, personal budgets were now being used widely, according to senior managers and practitioners.

Among service users, carers and care co-ordinators, personalisation was understood in very different ways across sites. Often it was equated with individualised care, choice, service user-led services, tailoring care, and balancing structure and flexibility. More specifically, it was used to refer to new practices: personal budgets and direct payments. A number of people in all sites had no idea what personalisation is about.

Service users and carers variously described care as being personalised, or as very much not personalised, depending on personal experiences, with no distinct differences between sites emerging. Some service users saw their care plans as very personalised, others most definitely did not. Similar feelings were expressed by carers. Personalisation was constrained by lack of resources and availability of local services, by service users not always being aware of the options available to them, and by service users deferring to professionals. Gaps were observed between the ideal and the reality, with staff reporting high administrative workloads and the complexity of actually accessing and using personal budgets. As with recovery, some care co-ordinators reported that personalisation can raise unrealistic expectations.

It was also apparent that services across sites often allocate care co-ordinators to service users on the basis of caseload space rather than appropriateness, further challenging personalised relationships or more personalised approaches to strengths, interests and goals. In terms of personalisation, care planning and co-ordination it was not clear that personalisation can be achieved consistently using current templates and approaches.

Semistructured interview schedule

Guidance for completing this form

The aim of this template is to record information on recovery and personalisation in the most recently available care plan for the specified recipient of mental health care. Where possible you should answer all questions. Where there is not sufficient information or clarity then please make a note in the notes box. The content of the care plan section is where we expect there to be some difficulty in completing this template and we offer this guidance to help. All reference to the person in the document should be taken to mean the recipient of mental health care.

Inclusion of the person’s views: look for evidence that the person had a say in the care plan. The ways in which the goals are worded might reveal this. For example ‘John would like to stop taking his medication’ indicates that this is John’s view and it also suggests a view that is not usually evident from workers.

Co-production: is there evidence that goals and treatment plans have been jointly formulated or show a shared ownership of goals and outcomes.

Strengths-based assessment: this can be seen in care plans that formulate goals built upon the strengths, skills or talents of the individual. An example might be ‘Jane would like to build on her analytical skills by enrolling on a systems analyst training course’.

Personalisation: is there evidence of choice and control for the person over their care and treatment? Are these based upon the person or just what is available? A care plan solely focused upon medication, side effects, symptoms or risk should not be considered to show personalisation as this ignores other types of support.

Personal budget: these give people control and choice over the purchase of personal social care services to meet their eligible needs. Is there a reference to this or discussion about the use of personal budgets in the care plan?

Recovery-oriented practice: does the care plan indicate a focus on building recovery such as facilitating new relationships, assistance with education or back to work, finance or money, personal care or physical well-being or developing a new sense of purpose?

Person-centred goals: are the care and treatment goals focused on the person’s assessment of their needs?

System-based goals: in contrast to the above are the care and treatment goals solely or majorly focused on system needs such as compliance with treatment.

The person’s views in risk assessment: frequently people are not asked for their view on risk assessments and many will not be aware that they have occurred; where a risk assessment is included are the person’s views on this evident in the document?

The person’s views in risk management plan: is there evidence that the person has agreed to the risk management plan or has been allowed the opportunity to discuss this and provide their perspective?

The person’s views in crisis plan: is there evidence that the person has had a say in the crisis plan, expressed a view on it or indicated their agreement or disagreement with the plan? Please make a note in the notes section to help us understand your response.

The person’s views in relapse plan: if there is a relapse prevention plan has this included the person’s perspectives?

Orientation to social outcomes: does the plan show a focus on social outcomes such as relationships and social networks or support, housing, work or education? If this is partial please make a note in the notes section.

Orientation to medical outcomes: on balance, is the focus on medical outcomes greater or lesser than the focus on personal or social outcomes? Please make a note in the notes section to help us understand this.

Encouragement for self-management: is there evidence that the care plan is encouraging shared responsibility with the person, for instance are there attempts to provide opportunities for the person to create their own plans or goals?

Recognition of personal relationships: is there evidence of a focus on fostering new or maintaining existing relationships?

Advance directives: are there plans for when the person is unable to make known their wishes for treatment? Is this signed and supported by workers?

Notes: please use the notes box beneath questions for adding detail. In the larger notes box it would be useful to give an impression of preponderance in the care plan, is it more focused on recovery and personalisation overall or does this make up only a small proportion of the whole? Include anything else you feel is directly relevant to the care plan that is not otherwise covered.

Collaborative Care Planning Project care plan review template