Recent government guidelines (National Institute for Mental Health in England, 2003a ; National Institute for Clinical Excellence, 2004) have highlighted communication problems between healthcare staff and people with personality disorder. The guidelines suggest that targeted staff training is needed to remedy this undesirable situation. The Personality Disorder Capabilities Framework has made explicit the capabilities required when working with people with this disorder (National Institute for Mental Health in England, 2003b ). It states that professional training is necessary to develop the capability of ‘establishing and maintaining long-term therapeutic relationships with service users and supporting other staff in maintaining therapeutic relationships (National Institute for Mental Health in England, 2003b ). Although the content of relevant training was not specified, guidance states that ‘service users who contributed to the development of this strategy were keen to stress the value of user involvement in training (National Institute for Mental Health in England, 2003a ). As a logical extension, the study reported here sought to: obtain the views of people with personality disorder on communication with healthcare staff; determine whether use could be made of any of the views obtained in developing a training tool(s).

Method

Four different specialist personality disorder services were approached. These services treat people with differing degrees of personality disorder within the mental health trust in south-west London and range in size from 8 patients in treatment to approximately 100. The treatment modalities used were psychodynamic psychotherapy, dialectical behaviour therapy, in-patient therapeutic community treatment and community-based keyworker case management with psychological therapy and psychiatric input.

The professional leads were requested to seek volunteers with personality disorder to meet as a group with the investigators (K.B./S.M.) for approximately an hour ‘to talk about issues concerning professional training’.

Group meetings were held at the premises of each personality disorder service and a full explanation of their purpose was given, at the start of each meeting. The rooms used for the meetings were familiar to the participants. On one occasion refreshments were provided. Such aspects were designed to put participants at their ease to facilitate their authentic participation.

A modified nominal group technique was employed (Reference Jones and HunterJones & Hunter, 1995). This consisted of structured iterative discussions that explored the negative and positive experiences of the interactions with professionals and how things could be improved. Investigators made contemporaneous field notes that were checked with participants for accuracy at several points during the meeting, at the end and again at the follow-up meeting. These notes were read to the participants by the facilitator and corrections made if necessary until agreement was reached. The meetings lasted between 60 and 90 min.

Thematic analysis was carried out on the field notes by the investigators. At a second meeting of the group the themes identified were checked with the group and altered to improve the extent to which service users’ perspectives were captured. Finally the group members ranked these themes in terms of priority during the second meeting. This process was repeated with each of the groups of service users from the relevant local specialist personality disorder services.

The ranked themes were then used by the trust's personality disorder working party. The latter comprised a multidisciplinary team, with further service user membership, that had been set up in the wake of locally acknowledged difficulties with treating this group successfully. The identified themes thus formed the basis for generating trust guidelines on how to interact with people with personality disorder. This process involved incorporating unaltered, as much as possible, the service users’ themes, adding guidelines to meet other trust requirements and altering the language or wording where unavoidable.

We were advised by the local health authority ethics committee that ethical approval was not required, but we took measures to conduct the study in an ethical manner.

Results

There were 13 patients from three of the four personality disorder services who were able to participate actively in the group discussions. Reasons for exclusion offered by the personality disorder service leads who were approached, were that certain service users were ‘too ill’ to tolerate this work or that they did not meet the single inclusion criteria of having been informed of their diagnosis. (The total number of patients excluded on this basis by their service leads was not known.) One of the personality disorder services was unable to recruit any volunteers for the project for the reasons given above and because of a lack of uptake by those approached. Membership of the groups changed between the initial meeting and the follow-up meeting. One participant left their group within 5 min of the meeting. However, an initial and follow-up meeting was held for each of the three participating services.

The raw data that were collected highlighted relevant issues, from which clear themes were identified. The latter were developed into the service user guidelines. These were further refined by the trust personality disorder working party into 11 trust guidelines on how staff should interact with those with personality disorder (see data supplement to online version of this paper). Examples of this process are given in Table 1.

Table 1. Examples of raw data and how they were incorporated into guidelines

Raw data	Service user guidelines	Trust guideline
‘ … self harmers don't harm to piss them [staff] off’	Patients should be treated with respect	A respectful and polite courteous professional manner is best at all times, avoiding personal disclosure. It is unhelpful to expose patients to disagreements within and between the team
	• with introductions, shaking hands • they should not be judged or blamed • boundaries should be respected • what patients report should be acknowledged as meaningful
‘ if you can't work [because you are ill] they think you are lazy.’
‘ … the way they talk to you, saying “Oh! Not you again”.’
‘ We should’nt be made to feel like a child with the staff as the parents with problems themselves…’
‘ … tensions between professional groups should’nt affect your treatment, they should agree between them how to treat you.’
‘ Time is important. They should give long enough appointments… they should take a period of time over a number of appointments before deciding a treatment plan… They should not make you feel rushed… Shouldn't look at watches.’	The value of time should be appreciated	If a diagnosis of personality disorder is made an interactive discussion with the patient is desirable. This will need a reasonable amount of time and an appropriate setting (private, free from interruption, etc). An acknowledgement that a diagnosis can feel stigmatising is important
	• • appointments should not be rushed • • assessment should take place over an appropriate length of time
	Being told the diagnosis is important
‘ Using the word “disorder” isn't very helpful…’	• to relieve guilt • explain previous difficulties • the diagnosis should be explained in an encouraging rather then negative way
‘ The way the diagnosis is given is important — it should be tailored to the individual, as a way of explaining the difficulties you have.’
‘ it's hard to trust new people…’	Patients should receive consistency and continuity from the team	Where possible continuity of contact between patient and individual professionals is desirable. Where this is not practical, notice of staff time off or changes of staff is helpful
‘ I don't want to keep telling people what my problems are… it's important that they know me.’	• by seeing the same person within the team • by being able to have contact with a professional who knows about the patient outside appointment times

Discussion

We report a successful attempt to involve service users in a trust's development of guidelines for interaction with people with personality disorder. We found that it was possible to conduct a number of structured group meetings and maintain a focus on relevant issues. Initial fears that discussions would prove too demanding for participants who were currently unwell proved unfounded. On the contrary, patients reported the experience of participating in this project as positive and freely gave their time to help us improve interactions.

An acknowledgement that the clinician's contribution to the interaction is not necessarily all positive may be a step towards improving the encounters between people with personality disorder and staff (Reference Norton and McGauleyNorton & McGauley, 1999). We found that involving service users in the process of feeding back their experiences of service provision helped to shed a light on aspects that clinicians might consider changing.

Furthermore, in our opinion, participants’ comments correspond with complaints clinicians raise when treating people with personality disorder, that members of teams may disagree about diagnosis and treatment and that resources are inadequate.

However, the process of involving service users raised several issues. Gaining their participation was problematic. Despite targeting services with approximately 150 identified people with personality disorder, only 13 service users took up the invitation to participate. Those taking part were not therefore necessarily representative of all service users with personality disorder. It is possible that those with grievances to air might have been more willing to take part; hence their views might be unusually negative. If so, this could support an argument that specific guidelines are only required by this minority of patients.

It might be argued that a larger sample of service users, including those who are not currently receiving treatment or are receiving it from a non-specialist personality disorder service, might improve the representativeness of the views. The fact that professional leads of the services were in a position to choose or otherwise influence those service users invited to participate could also have introduced bias. The changing membership of the groups possibly impaired the group's ability to foster a trusting attitude towards the investigator and a firm belief that service users’ views could influence professionals’ training. A more consistent membership would almost certainly have led to a more thorough evaluation of the draft guidelines produced and might have led to a more collaborative, superior but different product.

The above limitations could be overcome by widening the source of recruitment (e.g. to voluntary agencies), service users being involved in facilitating the groups and using advocacy for those who might struggle to communicate in the groups (Reference Simpson and HouseSimpson & House, 2003). However, the recruitment and maintenance of the small groups of service users with personality disorder took a considerable amount of effort by the investigators and professional leads. The nature of personality disorder means that people may struggle to commit to and develop working relationships themselves. Larger numbers in each group might have made the task for the group and the investigator unwieldy.

Given the foregoing, and considering that our data are congruent with those reported elsewhere (National Institute for Mental Health in England, 2003a , b ; National Institute for Clinical Excellence, 2004), this argues in favour of their validity, and guidelines for personality disorder have been produced for use in south-west London. It will therefore be possible to determine further the validity of our data by evaluating their clinical usefulness. This is our next step. If, after appropriate training, the guidelines prove to be ineffective in improving the user satisfaction with specialist services, we shall need to return to address the methodological limitations. If the guidelines prove effective, however, we shall consider how to introduce such training to a larger group of staff.

We conclude that our study shows that, with a degree of effort and persistence on the part of professionals, people with personality disorder can be involved to provide a distinctive perspective in pursuing the goal of improving the quality of their services.

Declaration of interest

None.