Introduction
The involvement of service users, including people with personality disorder (PD), is receiving increasing support from mental health policy makers and service planners in the United Kingdom (Department of Health 1999, 2009). Similarly, research commissioners are beginning to require researchers to engage with service users and to involve them in the various stages of a study (Public Health Agency 2010). However, we lack good understanding about the nature and extent to which service users are involved in activities such as the practice, planning, organization and delivery of services relating specifically to the field of PD services. We know even less again about the effects of service user involvement in PD services.
Involving people who receive support from mental health services in their care occurs in various capacities, structures and remits (Diamond et al. Reference Diamond, Parkin, Morris, Bettinis and Bettesworth2003; Simpson & House, 2003). A frequently used starting point for consideration of levels of user involvement is Arnstein’s Ladder of Citizen Participation (Arnstein Reference Arnstein1969) and levels or rungs of consultation, collaboration and consumer (or user) control adapted and applied to describe and evaluate involvement in research and services (Hanley et al. Reference Hanley, Bradburn, Barnes, Evans, Goodare, Kelson, Kent, Oliver, Thomas and Wallcraft2004; Strømme et al. Reference Strømme, Tinderholdt, Marit, Sandy and Oxman2006). Consultation is described as asking for views and using these views to inform decision making; collaboration, as active, on-going partnership; and user control, as positioning the locus of power and decision making with service users.
The last decade has seen important changes in PD policy and service provision [e.g. Personality Disorder: no longer a diagnosis of exclusion, Department of Health (2003b); and The Personality Disorder Capabilities Framework, Department of Health (2003a)]. NICE guidelines for the treatment of borderline PD and antisocial PD were published in January 2009 (National Institute for Health and Clinical Excellence, 2009ab). The research in this area has highlighted that service users with a PD diagnosis have had negative experiences of mental health services (Horn et al. Reference Horn, Johnstone and Brooke2007; Rogers & Dunne Reference Rogers and Dunne2011).
There has been little analysis of public involvement in health and social care research (Staley Reference Staley2009). A number of systematic reviews have focused on the effects of user involvement (Crawford Reference Crawford2002; Strømme et al. Reference Strømme, Tinderholdt, Marit, Sandy and Oxman2006). However, none have considered specifically the area of PD. This scoping review considers two arguably difficult conceptual areas ‘service user involvement’ and ‘PD’ in order to consider potential learning for policy and practice.
Methods
A scoping review methodology (Arksey & O’Malley Reference Arksey and O’Malley2005) was employed for a number of reasons. First, service user involvement may be considered to be a complex intervention and the review methodology needed to be broader in scope than the (more specific, narrow) conventional systematic review in order to capture and examine this aspect. Second, an early investigative search indicated a lack of peer-reviewed work. Finally, a scoping methodology is appropriate when the parameters of a topic are unclear and the intention of a study is to ‘map’ the field and identify research gaps.
Stage 1: identification of the research question
The research question revolved around and was informed by the literature about user involvement in PD services. (Crawford Reference Crawford2002; Arksey Reference Arksey2003).
Stage 2: identification of relevant studies
In keeping with a scoping review we adopted a search strategy, which involved searching different sources. Electronic databases and references of papers in the review were searched; a number of clinicians and researchers in the field were contacted; and an internet search was conducted.
Search strategies for electronic databases were developed with an information specialist. Initial searches of three electronic databases (Medline, CINHAL and PsychINFO) were undertaken in order to inform the scoping search. Existing reviews of service user involvement were examined and their search strategies and terms were incorporated into the scoping methodology (Crawford Reference Crawford2002; Strømme et al. Reference Strømme, Tinderholdt, Marit, Sandy and Oxman2006). The scoping search strategy combined ‘PD’ in combination with other keyword terms that are used to describe ‘service user involvement’. An example of a representative search is included in Fig. 1.
Search strategy for Medline (1950 to current).
Titles and abstracts were reviewed in the initial exercise and identified papers were reviewed in full against the search criteria for inclusion. Identified literature was also reviewed for further key terms which could be included in a final search.
The final search was conducted using the keywords piloted in the initial search as no further key terms were identified. The keyword search was run in seven further databases (ASSIA, Ingenta, SAGE, Science Direct, Social Sciences Citation Index, Social Sciences Abstracts and Sociological Abstracts) to identify relevant studies and writings in the social sciences. Papers were reviewed in the same way as those identified by the initial search of the initial three databases.
The method used for the internet search was to employ a strategy of searching four main websites as an initial step (Centre for Mental Health 2011; Emergence n.d. Involve 2011; Personality disorder n.d.). These websites were chosen as they are prominent mental health, PD and service user websites. Links identified on the websites were followed.
Stage 3: study selection
Criteria were devised iteratively following familiarization with the literature and then applied to the identified literature in order to ‘map’ the field. This approach differs from the methodology used for a systematic review where specific a priori inclusion and exclusion criteria are devised (Arksey & O’Malley Reference Arksey and O’Malley2005). Scoping reviews tend to be non-systematic in nature and tend to focus on the breadth of coverage of the literature conducted on a topic rather than the depth (Rumril et al. Reference Rumril, Fitzgerald and Merchant2010). The approach was necessary in order to scope the as yet unchartered field of user involvement in PD services.
We included papers which described service user involvement in planning, monitoring, developing or commissioning services; teaching or delivering services; development of guidelines or patient information material; research (other than as subjects) and advocacy and peer support. Papers were only included if they also addressed PD services. Papers were excluded if they described service user involvement in individual treatment choices. Arguably, a service user’s involvement in their own treatment choices has an effect on services and their development and, generally, this aspect is recognized as an important part of mental health treatment. We focused on the area of service user involvement where the service user voluntarily or in a paid capacity agrees to offer their skills and experience in order to influence services for others with the same condition – we did not include patient involvement in individual treatment choices as this is a separate albeit-related aspect and requires individual systematic attention in its own right. No restriction was placed on type of study for research-based reports of service user involvement. The review also included non-research-based reports if they described service user involvement in terms of the categories noted above. Papers were restricted to English language but were not restricted by date. Seventeen papers met the inclusion criteria for this review (Fig. 2).
Flowchart showing selection process based on scoping methodology (Arksey & O’Malley Reference Arksey and O’Malley2005).
Stage 4: charting the data
The technique adopted in this review was a ‘descriptive-analytical’ method, which involved applying a common analytical framework to all the primary research reports and collecting standard information on each study. An adapted version of the ‘PICO’ formula was used to chart the data (Centre for Reviews and Dissemination 2009) None of the studies featured a comparison group and so consideration was given to population, intervention and outcome in devising the domains for study.
We divided papers into research-based (empirical) studies and papers which were articles and commentaries exploring relevant themes. The series of data captures and information domains were organized into two tables for subsequent analysis. Table 1. charts the data extracted from the empirical papers and Table 2. charts the data extracted from the articles/commentaries. Data about the ‘poplulation’ was captured through a description of the methodologies employed for empirical papers and for both types of paper the context of service user involvement within the literature. ‘Intervention’ data was captured in terms of the form of service user involvement and the level of involvement as well as whether service users or professionals initiated the involvement. This data capture was informed by the approach described in the systematic reviews by Strømme et al. (Reference Strømme, Tinderholdt, Marit, Sandy and Oxman2006) and Crawford (Reference Crawford2002). Charting in brief the main findings for empirical papers (Table 1) and a brief summary of discussion for commentaries (Table 2) captured the nature of outcomes or effects. Scoping reviews tend not to evaluate the quality of studies in the review (Rumril et al. Reference Rumril, Fitzgerald and Merchant2010).
Data that were extracted and charted from empirical studies
SUI, service user involvement; PD, personality disorder; TC, therapeutic community; MCN, managed clinical network; SUR, service user researcher.
Data that were extracted and charted from commentaries
SUI, service user involvement; TC, therapeutic community; PD, personality disorder; MCN, managed clinical network; SUR, service user researcher.
Stage 5: collating, summarizing and reporting the results
The data was organized and tabulated as described and then analysed further through repeated reading and sorting and finally presented in a narrative form. The iterative approach of repeated reading and sorting allowed identification of key themes or links. This second level thematic analysis is described in the results.
A consistent, explicated approach to charting the data and presenting the results so that the results may be checked against the original data was employed by the first author and the second author checked work at every stage.
Results
Seventeen papers were identified which included diverse material including research and articles/commentaries on service user involvement in PD services. Table 1 presents information which was extracted from empirical studies: methodology employed; the context of service user involvement; the form and level of user involvement; whether service users or professionals initiated involvement; and a brief outline of the focus of findings. Table 2 presents information extracted from articles and commentaries: context of service user involvement; form and level of user involvement; whether service users or professionals initiated involvement and a brief outline of the discussion.
Thematic analysis
Importance of context
Many studies had associations with the English National PD Development Programme. User involvement began as a parallel process to the main work of the programme. The reported early benefits of user involvement led to further user involvement as the programme developed and user involvement became integral to new projects and developments (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007).
Many of the PD services and research projects had a therapeutic background. It is noted within the two evaluations of PD services that services based on therapeutic community (TC) models have an integral service user involvement component (Crawford et al. Reference Crawford, Rutter, Price, Weaver, Josson, Tyrer, Gibson, Gillespie, Faulkner, Ryrie, Dhillon, Bateman, Fonagy, Taylor, Moran, Beckett, Blackwell, Burbridge, Coldham, Gould, Imlack, Parfoot, Sheldon, Sweeney and Taylor2007; Moran et al. Reference Moran, Fortune, Barret, Spence, Rose, Armstrong, Slade, Mudd, Coid, Crawford and Tyrer2008). Service users who experienced treatment within a TC tended to bring the rules and culture of a TC to the development of new services and to act as ‘culture carriers’ (Rawlings Reference Rawlings2003). Haigh et al. (Reference Haigh, Lovell, Lyon and Duggan2007) noted that ‘…projects like Thames Valley Initiative (TVi) and the South West London Service User Network (SUN) take their model of service user involvement from the themes of democratization and empowerment that have been alive for many years in residential TCs such as the Henderson, Cassel and Francis Dixon Lodge’.
Roberts (Reference Roberts2006) and Sundaram (Reference Sundaram2010) highlighted the importance of policy and service development. In particular, Roberts (Reference Roberts2006) noted the importance of social context including social division and structural inequalities with concern that the difficulties of representing a diverse population may be overlooked.
Wellness and health
The wellness or health of users was a common theme with indications that there was a need for users to monitor or evaluate their health and for relevant parties to be aware of the sometimes demanding nature of the service user role. The need to negotiate their own mental health in the context of a demanding piece of work was a serious challenge to user researchers (Faulkner et al. Reference Faulkner, Gillespie, Imlack, Dhillon and Crawford2008). There was an expectation that service users who engaged in involvement-like activities met ‘wellness criteria’ (Balls Reference Balls2010).
The development of wellness criteria and a code of conduct are tools designed to place service users (or ex-service users) at the centre of considerations. According to wellness criteria, service users should be facilitated to assume responsibility for doing whatever is needed in order to safeguard their health (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007). Focusing on ‘wellness’ rather than ‘illness’ was intentional in order that, for example, taking time out from responsibilities could be seen as a positive step that preserved wellness and allowed a service user to return to work without the negative associations of emotional or mental ill health. Also important was the need to match the type of involvement or participation to the wellness of an individual service user (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007; Jones & Stafford Reference Jones and Stafford2007). Significant difficulties may arise among service users when one of them has a crisis (e.g. attempted suicide) and appropriate supports are required to take account of this type of event (Rawlings Reference Rawlings2003).
Support for services users during involvement activities
Support for service users was the second common theme to emerge. A number of papers reported that a great deal of support was needed for projects involving service users including administrative and emotional support. Some user researchers may go through a crisis during a research project and sufficient numbers should be trained to take account of the difficulties and to enable users to return to the project after a crisis (Ramon et al. Reference Ramon, Castillo and Morant2001). The support required in service user research projects was greater than anticipated; and providing and resourcing sustained support for user-researchers and project co-ordinators of studies was recommended (Ramon et al. Reference Ramon, Castillo and Morant2001; Rawlings Reference Rawlings2003; Barlow et al. Reference Barlow, Miller and Norton2007; Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007; Balls Reference Balls2010). The Service User Research Group England guidance was utilized along with a flexible team approach to projects that allowed users to cover for each other and to drop in and out of a project as necessary in the context of solid formal and informal peer support (Faulkner et al. Reference Faulkner, Gillespie, Imlack, Dhillon and Crawford2008).
Recruitment of service users
Recruitment procedures appeared to be important contributing factors to successful user involvement. In some instances difficulty was experienced in recruiting sufficient numbers of users (Crawford et al. Reference Crawford, Rutter, Price, Weaver, Josson, Tyrer, Gibson, Gillespie, Faulkner, Ryrie, Dhillon, Bateman, Fonagy, Taylor, Moran, Beckett, Blackwell, Burbridge, Coldham, Gould, Imlack, Parfoot, Sheldon, Sweeney and Taylor2007; Moran et al. Reference Moran, Fortune, Barret, Spence, Rose, Armstrong, Slade, Mudd, Coid, Crawford and Tyrer2008). Difficulty involving a sufficient number of users was experienced when recruitment was informal and based on recommendation from clinicians (Barlown et al. Reference Barlow, Miller and Norton2007). Some professionals initiated user involvement and evolved selection criteria for future service users (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007; Balls Reference Balls2010). Service user networks were also used for recruitment (Ramon et al. Reference Ramon, Castillo and Morant2001; Faulkner et al. Reference Faulkner, Gillespie, Imlack, Dhillon and Crawford2008) and Gillard et al. (Reference Gillard, Tumer, Neffgen, Griggs and Demetriou2010) involved recruitment through a service user and carer-led organization.
Three projects outlined core requirements or capabilities that users were required to possess (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007; Faulkner et al. Reference Faulkner, Gillespie, Imlack, Dhillon and Crawford2008; Balls Reference Balls2010). Capabilities referred generally to skills such as the ability to look beyond individual circumstances to address the broader needs of people with PD. These papers described a requirement for a lived experience or knowledge of the difficulties associated with PD. Service users felt that a ‘readiness’ for training and an individual’s willingness to participate in training activities were important (Balls Reference Balls2010) and Jones & Stafford (Reference Jones and Stafford2007) described the development of a career structure for the employment of experts by experience.
Differences were observed between the employment of service users who had previously experienced treatment in a TC and non-TC service users (Rawlings Reference Rawlings2003). The ex-TC patients were less impulsive, made more constructive decisions about dealing with crises and group dynamics during meetings were more stable. A difference was also observed between early recruitment and later recruitment. Service planning developed more quickly when recruitment occurred later and initial project operational issues had been resolved. Programmes formed more slowly to fit with particular local services and needs when recruitment occurred early in service design and development.
Discussion
Mismatch between policy, practice and research attention
This scoping review of the literature relating to service user involvement in PD services found only 17 papers including non-research-based work despite adopting a broad scoping methodology. There was no reference in the published literature in the review regarding this lack of attention, which is surprising given the increasing emphasis on user involvement, which is advocated in policy documents. In particular, the scoping review did not identify any papers from outside the UK. There were no outcome-focused or economic evaluations of service user involvement
The importance of context
A number of studies showed an association with the English National PD Development Programme. Overall, the literature seemed to indicate a progression from professional, top-down initiated service user involvement to a bottom-up, user-initiated approach. For example, the National PD programme provided an impetus and perhaps resources for user involvement but increasingly service users assumed ownership for involvement. However, it is surprising that only one paper described service user involvement in forensic PD services (Moran et al. Reference Moran, Fortune, Barret, Spence, Rose, Armstrong, Slade, Mudd, Coid, Crawford and Tyrer2008) especially in the light of the focus on Dangerous and Severe Personality Disorder Services as part of the National PD programme.
The development and evolution of user involvement in PD services merits further research in order to improve understanding about factors that foster successful user involvement and greater ownership by users. These factors may include elements of a TC culture for example democracy, empowerment and a flattened hierarchy.
The ‘mechanisms’ of user involvement
In common with the findings of other reviews, the literature is varied in the detail and attention given to descriptions of user involvement as an ‘intervention’ (Crawford Reference Crawford2002; Strømme et al. Reference Strømme, Tinderholdt, Marit, Sandy and Oxman2006). Lack of detailed information created difficulties in terms of assessing the form and level of user involvement and within some papers a wide variety of activities were described but not elaborated upon. Collectively, the literature suggests that the form and level of involvement are important to understanding meaningful user participation.
The effects or outcomes of user involvement
The findings or outcomes of studies were mainly based on processes and related to the delivery of user involvement. Emotional or practical benefits for service users were noted in some studies. Emotional benefits included empowerment, validation and the installation of hope and to a lesser degree changed expectations. Practical benefits included finding pathways back to work, the production of guidelines and the production of research. None of the papers described outcomes or effects in terms of PD services or treatment and changed clinical outcomes. A similar finding has been observed in studies of user involvement in other areas (Crawford Reference Crawford2002; Strømme et al. Reference Strømme, Tinderholdt, Marit, Sandy and Oxman2006; Staley Reference Staley2009).Greater expansion of assessed outcomes may give a better indication of what can be achieved by service user involvement. Outcomes should be compared against stated aims. More importantly, perhaps, is the need to incorporate the concept and measurement of ‘personal recovery’ into the assessment of outcomes (Donnelly et al. Reference Donnelly, Scott, McGilloway, O’Neill, Williams and Slade2011).
Wellness, health and support for service users
Wellness and health emerged as a key theme though arguably it has not been given the attention it requires. Health difficulties amongst service users were reported in papers but they were not considered as an outcome or finding of the work. It is important that both positive and negative outcomes are assessed and appraised in order to increase our understanding of user involvement in PD services. Similarly, drop out from user involvement is important but it has not received sufficient attention. Jones & Stafford (Reference Jones and Stafford2007), for example, reported positive results from user involvement without discussion of the poor response to questionnaires and the reasons for dropping out of the study. As noted earlier, wellness criteria have been devised as a tool to help service users safeguard their health (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007). The question of how best to safeguard and respond to wellness and health is linked to the issue of the support for service user involvement, which has tended to be underestimated.
Recruitment
Recruitment was described as difficult in a number of papers whereas other papers discussed recruitment processes in detail. Recruitment practices varied and in two papers an evolving understanding of recruitment was evident with initial selection based on recommendations by clinicians and then later selections were based on developed selection criteria (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007; Balls Reference Balls2010). This evolution would appear to be linked to the progression from professional top-down initiated work to user-led, bottom-up initiated work.
Core requirements or capabilities were described in three papers (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007; Faulkner et al. Reference Faulkner, Gillespie, Imlack, Dhillon and Crawford2008; Balls Reference Balls2010). These papers described a lived experience or knowledge of the difficulties associated with PD. Service users may fulfil different roles depending upon their stage of treatment and wellness (Haigh et al. Reference Haigh, Lovell, Lyon and Duggan2007) and service users who have not experienced treatment yet may be less able to engage in demanding service user roles. Roberts (Reference Roberts2006) argued that it was important all service user voices are heard in the development and planning of services including those who are not actively engaging in treatment.
Limitations of this scoping study
In general, the lack of literature affords only tentative conclusions about the key issues and more research is required to identify the differential impact of key aspects of user involvement. In addition, there was a lack of information about the form and level of involvement. Most papers reported small numbers of users involved in services and for research-based papers there was a lack of detail regarding methodology. It is possible that a wider search may add to the literature, for example, by including additional electronic databases and using terms relating to forensic mental health services as well as contacting a wider range of clinicians, researchers and user involvement bodies to identify further grey literature. However, given the breadth of the search strategy used in the scoping review it is unlikely that any new substantive material would be uncovered. The scoping study could potentially be strengthened by a consultation exercise to inform and validate findings from the main scoping review (Arksey & O’Malley Reference Arksey and O’Malley2005).
Further research
The lack of published research in user involvement in PD services is significant (and there are few suggestions for further research in the published literature). There was insufficient research to identify with confidence the effects of user involvement in PD services or the key elements to successful user involvement. The extent to which service user involvement in this contested area of mental health provision differs from other services requires further consideration and attention. This review adds to the literature by indicating a number of elements of user involvement worthy of further study. It is suggested that an appropriate framework for future research is to consider the context, mechanism and outcome of user involvement. Furthermore, consideration of user involvement as a complex intervention using Medical Research Council (MRC) guidance in designing appropriate research methodology may be helpful. The MRC guidance for developing and evaluating complex interventions (MRC 2008) outlines a development-evaluation-implementation process. This review provides a starting point for the development of interventions, which could be subjected to piloting and feasibility testing before formal evaluation.
The elements and themes that emerged from the review provide hints and insights about the development of user involvement in PD services. For example, a common link between studies and reports – a TC – points to the elements of a TC culture which may lend themselves to effective user involvement (e.g. democracy, empowerment and the flattened hierarchy of the TC). A conceptual analysis is required in order to clarify potential processes or mechanisms (e.g. democratic relationships) and outcomes (e.g. feeling empowered).
Conclusions
This scoping review of the literature on the effects of user involvement in PD services has revealed only a small amount of published work despite policy support for user involvement in health services. Significant gaps in the literature were evident suggesting that investment in research into user involvement in PD services has not followed policy support. There are considerable difficulties in considering meaningful user involvement including lack of exploration of the form and level of involvement. The effects or outcomes of user involvement were mainly based on processes and related to the delivery of user involvement. An understanding of the effects of user involvement and the best means of assessing these will become increasingly important due to the need to justify the resources required for user involvement in a current climate of financial austerity.
While there was insufficient research to identify with confidence the effects of user involvement in PD services or the key elements to successful user involvement, this review points to a number of potentially fruitful areas for further study such as the ‘cultural’ components of user involvement. The MRC guidelines for developing and evaluating complex interventions (MRC 2008) may provide a framework for future research and the results of this scoping review may be used to contribute to the development of an intervention that incorporates a cycle of piloting, modelling, feasibility, testing and appraisal.
Acknowledgements
We would like to acknowledge the expert guidance on database searching provided by Mr Richard Fallis, Subject Librarian, School of Medicine, Dentistry and Biomedical Sciences, Queen’s University Belfast.
