Open Dialogues in social networks: professional identity and transdisciplinary collaboration

Aim The aim of this article is to explore the challenges connected to the transformation and emergence of professional identity in transdisciplinary multi-agency network meetings and the use of Open Dialogue. Introduction The empirical findings have been taken from a clinical project in southern Norway concerning multi-agency network meetings with persons between 14 and 25 years of age. The project explores how these meetings are perceived by professionals working in various sectors. Methodology Data was collected through three interviews conducted with two focus groups, the first comprising health care professionals and the second professionals from the social and educational sectors. Content analysis was used to create categories through condensation and interpretation. The two main categories that emerged were ‘professional role’ and ‘teamwork’. These were analysed and compared according to the two first meeting in the two focus groups. Results and discussion The results indicate different levels of motivation and understanding regarding role transformation processes. The realization of transdisciplinary collaboration is dependent upon the professionals' mutual reliance. The professionals' participation is affected by stereotypes and differences in their sense of belonging to a certain network, and thus their identity transformation seems to be strongly affected. To encourage the use of integrated solutions in mental health care, the professionals' preference for teamwork, the importance of familiarity with each other and knowledge of cultural barriers should be addressed.


Introduction
The aim of this article is to explore the challenges connected to the transformation and emergence of professional identity in transdisciplinary integrated care. The World Health Organization has recently declared that mental health is a major societal challenge. About 25% of the population will have mental health problems in its lifetime. Moreover, when suffering from mental illness, many of the afflicted may also experience social difficulties, such as unemployment, discrimination or problems within their social network [1].
Although the concept of integration is a widely used term in the health and social care discourse, the concept has no universal definition. Likewise, we have no predominant model of health system integration or integrated care [2,3]. However, structure, process and outcome of integration represent imperative elements in integrated care [4,5]. The structure of integration relates to the organizational and administrative structure. The process of integration relates to factors, such as the quality of the relationships developed between the actors, and the suitability of the help that is performed [4,5]. The outcome of integration relates to patient satisfaction as well as collaborative skills among the professionals [4][5][6]. Moreover, to enhance integrated health and social care, the patient's total life situation should be taken into account [7]. This means the broader social context within which the person is living and the medical and psychological situation [8][9][10].
However, a high number of research papers focusing on integrated care in terms of multi-agency collaboration point out several difficulties associated with organizational, professional and contextual issues, such as role interpretation, communication, discipline conflicts and leadership [3,[11][12][13][14][15]. Hence, people suffering from complex illnesses and with multiple problems or at risk of developing severe multiple problems may especially be placed in a difficult situation. Due to fragmentation and collaborative difficulties in the helping system for those who require help from multiple agencies, we need to explore approaches that go beyond disciplinespecific traditions. Service delivery may be improved through genuine involvement of the help seeker and the private network. Through negotiation and decision-making that is collectively enhanced among professionals, the person seeking help and the private network, approaches and solutions that transcend traditional boundaries may be the outcome. Hence, by integrating a great variety of voices into one working unit, there may be a greater chance of achieving success. To enhance the use of holistic approaches, however, highly interactive modes of collaboration may be needed.
Interdisciplinary teams, which are representing the most common mode of interactive team work, can be described in different ways [16]. It may be defined so as to analyze and harmonize different disciplines into a coordinated whole [16] as well as having 'shared goals' or 'common methodologies' [16, p. 356]. Professionals working in an interdisciplinary way may aspire to "surrender some aspects of their own disciplinary role, but still maintain a discipline-specific base" [16, p. 356]. An example of this is case conferences during which the members gather together to discuss their individual assessments and develop a joint service plan. Transdisciplinary teams however may be considered as an "interdisciplinary team whose members have developed sufficient trust and mutual confidence to transcend disciplinary boundaries and adopt a more holistic approach" [16, p. 357]. In order to stimulate the emergence of new knowledge, the transdisciplinary team strongly emphasizes a great variation in information sources involving both professionals and non-professionals [16]. Compared to traditional interdisciplinary approaches, transdisciplinarity is regarded as more "context sensitive, eclectic, transient and inventive" [17, p. 850]. Furthermore, in transdisciplinary teams, professional roles may be strongly affected by the requirement of role release and role expansion. The term 'role release' means "accepting that others can do what the specialist was trained specifically to do", while the term 'role expansion' means "allowing that one's job can include more than what one was specifically trained to do" [16, p. 355]. This in turn means that through their focus on flexibility, trust and mutual reliance, transdisciplinary teams underscore factors that are considered to be success factors for cooperation in general [14]. However, by focusing on the great variation in information sources and the flexibility in the professional roles, transdisciplinary teams may improve the possibility of enhancing creative and holistic solutions. On the other hand, the inclusion of the various voices may demonstrate the complexity and advancement of transdisciplinarity. Conversely, transdisciplinarity aims to highlight new and intimate processes of integration. Hence, due to its potential, the effects of transdisciplinary collaboration should be further explored to gain knowledge on the processes and outcome of integration.
Integrated care in terms of complex collaboration may have different side effects for those involved [3]. One result of intimate teamwork carried out in different communities of practice is that the professionals' identity may be challenged [18]. This could occur because the professional identity is so closely connected to knowledge and experience [19]. Moreover, in constantly shifting communities of practice, the way we carry out our work and our professional role may be valued differently [18]. Hence, integrated care in the sense of increased intimate teamwork performed in the presence of the help seekers and the private network and adapted to their daily environment may alter the professionals' frame of reference when it comes to identity. Conversely, aiming to provide tailor-made, fully contextual and comprehensive integrated help may present great challenges to the traditional system of professions and push the tensions between professionals to the edge.
A Norwegian clinical pilot project, entitled Project Joint Development, implemented social network intervention in the form of Open Dialogue. 1 The aim was to provide tailor-made assistance for individuals from 14 to 25 years of age suffering from mental health problems. The aim of Open Dialogue is to emphasize an organisational integration structure by involving professionals from a number of agencies. The intention is that the professionals meet and carry out their work on an equal basis and in the presence of the help seeker and the private network. The process of integration is provided by a treatment approach where dialogue and interaction are key elements. By placing the help seeker and the private network in key positions, the aim is to achieve genuine changes [20]. Due to the inclusion of all these voices, successful outcome of approaches, such as Open Dialogue may require adjustments by the professionals if success is to be achieved. Thus, we want to examine how social network approaches, such as Open Dialogue can provide an approach to problems that have been refractory to integrated care in terms of multi-agency and multi-professional work. Hence, the aim of this paper is to explore challenges to professional identity in multi-agency network meetings, focusing on the way attitudes towards multi-agency practice are embedded in traditions of specialization in the sense of professional knowledge and mutual interaction. 2 More specifically, we will look into how professional identity is related to: the development of professional roles in multi-• • agency network meetings the development of transdisciplinarity in multi-• • agency network meetings.

Developing a professional identity
Personal identity concerns the question of 'Who am I' and theories concerning identity have been shifting throughout history. It might be understood in terms of individual identity versus dependency of the collective or a presupposed identity versus the individual's ability for reflexivity and identity as social constructionism versus essentialism [21].
Individual characteristics, such as having extraordinary talents or firm beliefs, may have a great impact on the formation of an individual's professional identity. Thus, professional identity is always dependent upon personal identity. Etienne Wenger stresses the dependency on the collective in developing a professional identity. Because we always negotiate meaning with social experiences, identity is created from a combination of both social and individual aspects [18].
Furthermore, Wenger claims that our identity is shaped through participation within and across community memberships, a notion involving mutual engagement, accountability to an enterprise and negotiability of a repertoire. Thus, in order to maintain identity, the work of reconciliation is of great significance for professionals who move between different communities of practice [18].
Developing a professional identity involves identification and negotiability, and the work of identification may be described in terms of inclusion and exclusion, stereotypes, paradigmatic trajectories and trust. Negotiability may be described as listening to other perspectives, seeking control and sharing responsibilities. The ability to take responsibility for meanings within a particular community involves the possibility to negotiate [18].
Wenger discusses the shaping of identity as a mix of participation and non-participation in relationships and activities founded on various degrees of identification and negotiation. Modes of non-participation include peripheriality and marginality, the former term (peripheriality) meaning participation involving less intensity, for instance professionals who are only superficially involved in a case. The second term (marginality) means that certain professionals may experience ignorance concerning ideas [18].
Identities of participation or non-participation may also arise through engagement, imagination and alignment. The first term (engagement) implies joint practice, and occurs when people have their ideas adopted by others. Next, imagination goes beyond engagement in practice (i.e. trade union subjects). Imagination in minor communities may involve participating through stories about local conditions. Identities of non-participation through imagination may emerge because of prejudice through stereotypes, while identities through alignment may occur through commitment or ignorance of professional approaches [18].

Open Dialogue as transdisciplinary work
Social network intervention and network therapy originated in the US in the mid 1960s [22]. The approaches move towards an embracing of the private network (family/friends, etc.) to varying degrees, and can be used to solve problems of both a practical and emotional nature. In the Nordic countries network intervention has been used in relation to a variety of problems [20,23]. Klefbeck and Ogden [24] have focused on network intervention regarding children in crisis. Other kinds of network intervention, such as Multisystemic Therapy, Family Counselling Meetings [24] and Anticipation Dialogues [25] have been developed.
In the Nordic countries research has been completed on social network intervention in the context of network meetings. The research concerns patients with psychosis, dual diagnosis as well as rheumatoid arthritis. The results indicate that network meetings have a significant impact on the patient's mental health [26][27][28][29]. An evaluation concerning organizational perspectives of network meetings revealed difficulties concerning professional collaboration. These difficulties were associated with professional roles, vague organizational structures and unfamiliarity with team partners in which a sense of insecurity emerged [30].
In an attempt to find new solutions for mental health care for people from 14 to 25 years of age, a pilot project entitled Project Joint Development was initiated.
The project aimed to provide help for those people in an early stage of mental illness as well as provide those with more severe problems the opportunity to take a more active part in their own treatment.
Project Joint Development applied a procedural intervention model based on network meetings constructed from 'Open Dialogue' [20,31]. The professionals were strongly encouraged to cross the borders separating professions and agencies. This was to be accomplished by creating a team for every single case consisting of a minimum of two professionals with education and positions relevant to the specific case. 3 The project's main ideas included the following: Organizing and continuity. The first person contacted is responsible for organizing the transdisciplinary team for the first meeting with the social network. In cases where network meetings are the primary intervention, the language and reflection should contribute to making the person seeking help more aware of his or her own resources. If the primary approach is individual treatment, network meetings will represent continuity among the persons seeking help and the private network. Toleration of uncertainty during the process. Instead • • of aiming for rapid solutions to the problem, the aim is to increase the ability to tolerate the time when no response is available. The generation of dialogue is the primary aim of the • • joint meetings to increase everyone's understanding of the problematic situation [20].
In the Open Dialogue approach, a slowly developing dialogue within the network meetings should be attained to create joint understandings and joint solutions among all persons present [20]. This means listening very carefully to the help seeker as opposed to providing prepared plans and ready-made answers.
Due to the evolving dialogues in the network meetings, the professionals' positions may change. By focusing on the voices of the help seekers, the professionals can move toward a more and more personal participation, in the sense that they may increasingly "adapt themselves to the present moment" [32, p. 485], and to the particular context. Thus, rather than being a technique, Open Dialogue represents a basic attitude involving increased transparency and disclosure of the professionals [9,10,32]. Due to this more personalized relationship among the network members, emotions and dialogues may be increasingly shared [32]. Therefore, through the focus on the help seekers' voices and the dialogue founded on their statements, the private network may be sufficiently inspired to create and maintain its own dialogues and solutions [20].
Project Joint Development was initiated and anchored in the department for drug abuse and psychiatry at the local medical hospital. Two municipalities in the hospital's catchment area were invited to participate. All agencies related to mental health care for people 14-25 years of age within the two municipalities, the county services and the relevant departments at the local hospital were included (Table 1). Professionals representing the psychiatric services were allowed to join three network meetings in relevant cases without referrals (founded on their supervision duties). Help seekers who were regarded as suited for network intervention by the professionals were invited to network meetings when they themselves approached agencies involved in the project. Professionals were also encouraged to propose network meetings to persons if this could possibly improve their lives (i.e. teachers who were concerned about particular pupils). The situation was defined as a crisis situation when the help seeker approached the helpers, where the first meeting should be arranged within 24 hours if necessary. 4 Did not satisfy inclusion criteria n=14 (i.e. supervisors, project leaders) Health care sector n=12 Social and educational sector n=2 Given that the project required collaboration from a whole range of professions and agencies, the composition of the groups should reflect the breadth of Project Joint Development [33]. Thus, as the point of departure was to ensure that professionals working in such important agencies as the educational sector and psychiatric services were included in the focus groups, the professionals were selected according to their agency affiliation. 7,8 Furthermore, the groups should have an equal number of participants from the two municipalities as well as the four supervision groups linked to the educational programme. Moreover, the professionals in the groups should be ordinary participants in Project Joint Development (i.e. not supervisors, project leaders etc.) and have participated from the project's outset. 9,10 Twenty-six persons fulfilled the inclusion criteria of the focus groups (Table 1). Based on a purposive sample [33,34] guided by the criteria for the group composition, 12 persons were invited to join the focus groups. One focus group included six professionals working in the health care sector (Health Care Group), whilst the other six professionals were employed in the social and educational sector (Social and Educational Group). Most of the subjects included in the Project Joint Development were suffering from difficulties relating to their social network, such as family or friends/colleagues. Many of them were in need of psychiatric treatment due to various levels of depression and anxiety. 5 At least 20 were suffering from multiple problems. Thirty-one were known to the specialist services prior to the first network meetings. Twenty-one persons were in such shape that no referral to the specialist services was intended when the first network meeting was arranged whilst referral for eight persons were sent to the specialist services.
In order to develop the new practice, a two-year training programme was conducted in which 40 professionals participated (Table 1). This programme consisted of 75 hours of lectures and 73 hours of supervision. The lectures focused on ethics, dialogues, common understanding and processes.

Methodological approach Participants
During Project Joint Development, 42 help seekers accepted to be screened for mental health problems, five of these did not arrive at the first screening interview. Thus, 37 were screened within a few weeks after 5 Information about mental health problems is mainly given by the professionals who had met the help seekers and is a result of information assigned to very rough categories. Since the outset of the research project was in February 2004, some help seekers were included in the Project Joint Development prior to the outset of the research study. There is limited knowledge concerning the problems of those who were included prior to the outset of the research study (n=16) and for those who declined to participate in the research study (n=23). For those who were included in the observational part of the research study (n=16) we know much about their problems. For those who were included in the research study, but only participated in the screening for mental health problems (n=21), we also have limited knowledge about their reason for seeking help in the Project Joint Development, since the questionnaires concerning their mental health problems not yet have been analysed. However, concerning those who were included in the research study in general, there were more help seekers who already had a record in the specialist services, compared to those who declined to participate in the research study. 6 The focus groups were carried out instead of group interviews with whole networks. 7 Some professionals had more than one kind of education and/or work experience related to the other group of which they were members. 8 Very few MDs participated in the project and few patients were referred by MDs. 9 On one occasion both representatives working in the same agency had other tasks in Project Joint Development. The informant included was chosen because of his relationship with the municipality and agency involved and with regard to the particular supervisory group. they were encouraged to discuss the skills and knowledge they found to be relevant for network meetings. They were encouraged to refer to actual situations and examples [35]. The first focus groups started with a vignette illustrating a typical case for network meetings. The case was followed by questions on how they could act to facilitate a successful network meeting as opposed to a network meeting with poor outcome.
Before the second meeting ALH wrote a summary based on the key findings for each group. The summary also included questions for the professionals to discuss. The summaries generated reflections on the changes in understanding the professionals had undergone since the first meeting. Before the third meeting, each group received a summary including their own quotes and ALH's preliminary interpretations. The possibility the informants had to recognize their former quotes in the focus groups and read the summaries served as a credibility check [34,37]. During the last meeting, the professionals were encouraged to reflect on how the findings could be applied outside Project Joint Development.
To enhance credibility, the transcripts from the first two meetings in each group have been closely examined by ALH, whilst MH and JS made comments on these initial analyses. The analysis mainly evaluated the group process to increase awareness of biases [34]. The analysis revealed that some expressions in the Social and Educational Group should have been more closely followed-up. Moreover, during the first meeting in the Health Care Group, the group may have been led too strictly, where the professionals may have experienced difficulties in expressing their thoughts and opinions.
Approval for the study was given by the Norwegian Data Inspectorate and the Regional Research Medical Ethical Committee. The participants were informed about the studies both orally and in writing, and they also submitted their written consent with regard to their own participation.

Analysis
The focus group interviews consist of 198 pages of transcripts (first and second meeting). The subsequent analysis is based on content analysis, focusing on explicit and latent underlying content [37,38]. Content analysis places important categories for the material in the centre of the analysis by using a stepby-step approach [39]. Moreover, as we were aiming for an explorative study, content analysis could provide us with the overview we needed. Multistage focus groups make it possible to observe more in depth the emergence of a developing professional identity. However, multistage focus groups also rep-present at the first two meetings. 11 At the third meeting four persons were present, including one by telephone conference call. Out of the six persons in the Social and Educational Group, five were present in the first focus groups, whereas four persons were present 12 at the second meeting. Only two professionals in the Social and Educational Group participated in the final meeting.
When the first focus groups were first established, three members had no practical experience, whilst the others had participated in from three to more than 25 network meetings. The professionals in the Health Care Group had more experience from network meetings compared to the Social and Educational Group. In both groups, there were two or more members with experience from more than 25 network meetings when the second meeting was held.
The age range (2004) was 33-58 (mean=46), including four men. The number of years employed in the current position ranged from 3 to 25 (mean=8) (2 missing). Five persons (2 missing) had previously worked in agencies with relevant tasks concerning individuals with mental health problems.

Procedure
The two focus groups met three times, the first encounters taking place in 2004 and 2005. They met once again in 2007 in order to validate the former findings through discussing central topics more closely. ALH was the leader of the first two meetings in each group, whilst the final meetings were led by ALH and MH. Each focus group lasted for 2-2.5 hours, taking place in the child and adolescent clinic, and these were all audiotaped. The first and second meetings in both focus groups have been transcribed verbatim by ALH. The transcriptions include breaks, expressions, such as laughter and sighing and the informants' interruption of each other [35]. During the interviews, ALH wrote notes and made short verbal summaries of how their expressions were understood. During the last meetings, ALH and MH undertook some open reflection [36] with the aim of encouraging the professionals to confirm or correct their understandings. Immediately after the focus group, ALH made audiotaped summaries of how the focus groups had functioned, focusing on the conversation and group dynamics [34].
In order to grasp the professionals' reflections concerning their professional identity in network meetings, 11 A student was present in the second meeting in the Health Care group. The student was only observing and did not participate in the discussions. 12 Two persons had moved, whilst another one who had been absent during the first meeting was now present.  13 Quotas concerning professional role include their present understanding of their professional role, including difficulties that emerged. Quotas related to teamwork concern their understanding of collaboration in network meetings in which mutual interaction is highlighted.
The categories and codes were created by ALH. JS and MH read the preliminary analysis. During the development of the final analysis, they also read reports in which the category system was presented and which the meaning units were assigned to categories. MH's and JS's feedback were used to modify the category system and the assigned meaning units, where the aim was to identify the most suitable meaning units, categories and codes [37]. When the categories and codes were finally clarified, MH read and confirmed the analytical themes. To further ensure credibility, MH examined the data to explore if any items had been systematically or randomly excluded or if irrelevant items had been included [37].
Each focus group's quotas are referred to as either HCG (Health Care Group) or SEG (Social and Educational Group). The particular meeting is referred to as first, second (or third) meeting (m) and the given informant by a number (SEG/1m/1). Results from the third meeting are included if they are of great significance with regard to the first and second meetings.

Results
The findings reported here (see also Table 2) emphasize the group discussions concerning professional roles and teamwork in network meetings. In order to illustrate the professionals' opinions, some quotas are included. The supplementary text represents a summary of the discussion connected to each category. resent the risk of losing the continuity of the core representatives due to the number of meetings that were held [34].
The first step of the analysis was to read through the focus groups' discussions to obtain an overall understanding. From this first reading, about 45 topics were identified. During the subsequent analysis, we identified the most important themes that emerged from the discussions. After having identified the main theme, "Professional Identity, Professional Role and Transdisciplinary Collaboration in Network Meetings", we created two categories; 'Professional Role' and 'Teamwork' ( Table 2). Nineteen subcategories and 19 codes were linked to these categories before the subcategories and codes were merged into eight codes during the final analysis.
However, the discussions in the groups developed rather differently, e.g. professionals in the Health Care Group focused on their sense of insecurity. To highlight these differences and thus identify differences between the groups from one meeting to the next, we identified sequences of discussion linked to each category during the final analysis. This identification was based on both the interview guide (e.g., asking about professional roles), and what the participants themselves brought up [33] (e.g., the impact of mutual reliance). We analysed each sequence closely in order to identify the core message, creating one code for each group and meeting consisting of a minimum of four transcript sequences reflecting each category, i.e. 'Role release across stereotypes' and 'The impact of mutual reliance' [37].
The codes reflect the core messages in each category. As there were great variations in each group concerning the density of quotas in each sequence, the codes may represent various numbers of quotas, which in turn may represent answers to actual questions from the researcher or other informants, be a minor part of a discussion, or concern the first meeting being a part of the case discussion.
To some extent the categories overlap one another, and hence the results could be analysed in connection

Our interpretation:
Members of both groups seek role expansion, aiming to increase the help seeker's activity. Some members of the Social and Educational Group have conflicting interests, searching for clarity about organizational issues and hence interdisciplinarity. They are all still fighting against stereotypes.

The development between the meetings
Their motivation about identity alteration moving towards transdisciplinarity is illustrated through role release and role expansion at the first meeting. The second meeting indicates a change as the professionals within each group had rather different experiences.
Thus, as discrepancies about role transformation

Professional role First meeting in the focus groups
The professionals in the Health Care Group highlight the difference between therapy and network meetings, stating that: "It is [a] conversation with therapeutic effect, but we're not doing therapy" (HCG/1m/2) and "I'm not going to be an expert, and people should never experience me as one" (HCG/1m/1). They discussed professional terms to be used in network meetings like 'collaborative partner' as opposed to 'therapist'.
The members of the Social and Educational Group argued more about their roles. Some of them were eager to facilitate a non-prescriptive behaviour by focusing on the help seeker's personal opinion concerning his or her life situation. They encouraged the individual to make his/her own choices: "You've got this food platter, and here are the different dishes you can choose" (SEG/1m/4). They also denoted behavioural difficulties: "It's easier to go into the role of helper rather than being passive" (SEG/1m/1). Others emphasized their ability to provide supervision and advice.
They all reported difficulties throughout the role development process. This was mainly due to the stereotypical approach to professional roles introduced by professionals unfamiliar with transdisciplinary network meetings, the individual seeking help and the private network. Additionally, professionals in the Health Care Group claimed that practical issues were often brought up by these same professionals: "They can be prescriptive and much focused on implementing solutions" (HCG/1m/5). The Social and Educational Group confirmed this by referring to the pressure they felt when either the individual seeking help or their family called for immediate help: "Now it was a question of what we had to offer. They really demanded that I come through on this" (SEG/1m/1).

Our interpretation:
Professionals in the Health Care Group search for role release by reducing the impact of therapeutic skills and altering terms denoting their position. Also for those without any previous experience in social network interventions this was possible. On the contrary, members of the Social and Educational Group emphasized communication guided by the help seeker and role expansion. The stereotypes that become apparent represent difficulties transforming their professional roles according to transdisciplinarity. Nonetheless, participating in network meetings created new possibilities for these participants to relax their professional borders. Hence, their motivation and potential for altering their professional identity according to transdisciplinary collaboration may be strongly affected.

Second meeting in the focus groups
The professionals in the Health Care Group discussed the impact of familiarity with each other as being important according to team formations. During the project, mutual reliance had emerged among the professionals included in the project: "Before it was like the teachers and child protection workers and school psychologists sat together, but now we're much more persons sitting together and I feel in a way this is my group, or our group" (HCG/2m/2). But still, difficulties emerged due to the multi-agency approach: "Maybe some aren't that comfortable with that role, allowing oneself to be vulnerable with your own thoughts and feelings" (HCG/2m/6). In the third meeting some professionals claimed that the most important factor concerning mutual confidence was depending on their partners' belief in Open Dialogue.
The professionals in the Social and Educational Group pointed to difficulties caused by different interpretations of the situation: "If someone from the social sector says it happens, then maybe there's someone who has more psychiatry that says, No, we have to wait with this, this is the past, present or future" (SEG/2m/6). They also discussed how exercise of authority affected the collaboration: "If network meetings don't manage to keep the kids within acceptable boundaries, then it's easier to initiate other measures" (SEG/2m/1). These expressions were replicated during the third meeting, when the Social and Educational Group discussed mental health care in terms of the law: "Everyone knows where government has placed the responsibility" (SEG/3m/6).
The professionals in the Social and Educational Group, when asked directly, denied that a sense of insecurity had emerged during practice. However, they said: "It's kind of good to know what one can expect from the others. And then there's the personal. What attitudes does he have?" (SEG/2m/1) and "It's really important who. It's very personal " (SEG/2m/5).

Our interpretation:
The Health Care Group's example regarding mutual confidence and trust illustrates the importance of becoming familiar with partners in order to achieve transdisciplinary engagement. Nevertheless, some professionals are still being marginalized through questioning their competence. Professionals in the Social and Educational Group demonstrate practical implications because of diverse understandings and thus, the become stronger, the difficulties with respect to identity transformation become apparent.

Teamwork
First meeting in the focus groups Professionals in the Health Care Group experienced an increased sense of professional insecurity through their practice. Professionals working at the primary care level supposed that individuals were referred to the specialist service because of a need for more sophisticated treatment, hence the statement: "I start feeling insecure about my role as contact person and as the person they (help seeker) trust" (HCG/1m/1). Their feelings of insecurity were also related to the mutual reliance between themselves and those who were working in the specialist service: "Does that mean that when you and I sit in a network meeting, you see me as the expert and yourself as the follower?" (HCG/1m/2).
They noticed in a broader sense the challenges of negotiation brought on by the multi-agency perspective: "When composition is multi-disciplinary and there are participants from the school system, I think they're more aware of the distance. At the same time, they're more on the sidelines, because here they talk to the person who's sick. The concepts of health in a way" (HCG/1m/6).
In response to a direct question about the educational programme, the professionals in the Social and Educational group 14 noticed challenges related to differences in agencies as well: "The health care sector in a way only sees its own clan" (SEG/1m/4). They emphasized the difference in the knowledge base between themselves and the Health Care Group: "We're like supposed to have respect for the job we do. That we actually meet with most of the kids" (SEG/1m/1). Contrasting collaboration prior to the onset of the project, one claimed: "It's the traditions we're a part of that determine how things happen. I don't feel like there's been very much change" (SEG/1m/3).

Our interpretation:
Professionals in the Health Care Group demonstrate an increased sense of insecurity linked to mistrust in others. They place professionals in the other group in a marginalized position, pointing to their unfamiliarity with medical terms in which a decrease in eclectiveness emerges. However, the Social and Educational Group place themselves in the periphery, pointing to the health care sector as representing the principal paradigm. These factors may be of major importance concerning their mutual identification and negotiability.
identity change is demonstrated. However, even if the professionals change their roles and attitude, the tradition and discipline-specific dimensions may create difficulties in situations where other professionals and non-professionals maintain their actions according to stereotypes [18]. Moreover, the findings indicate that professionals in primary services place themselves somewhat in a nonparticipative position through claiming to have different areas of competence and insufficient legitimacy. Professionals in the social and educational sectors seem to be placed in a marginalized position due to their lack of medical terminology. However, they also place themselves in the periphery through their complaining about the lack of attention paid to their perspectives. Likewise, professionals in primary health care services question their position through making a distinction between primary and specialist services. These results correspond to a study [11] finding that 21% of professionals in primary care asserted that professionals in the specialized mental health care sector showed a lack of respect with regard to their level of skills and expertise. Moreover, close collaboration calls for social intimacy and social competence [41,42]. Since the emergence of new solutions and creative ideas at first glance may be considered strange and unprofessional, mutual reliance may affect the processes generated in teamwork in the sense that vulnerable professionals may be less creative [42].
The findings from our study point out the importance of the professionals' sense of security, as well as the impact of mutual reliance. Although their mutual reliance increased during the project, their focus on their collaborative partners and their attitudes was maintained. Bearing this in mind, the processes generated in teamwork find the importance of the professionals' personalities when taking part in close collaboration to be a critical factor. Moreover, working in a transdisciplinary way increases the professionals' knowledge about each other, both professionally (how they understand collaboration in terms of their position) and individually (their personal values and beliefs). Consequently, transdisciplinary collaboration may potentially contribute to a stronger culture of transformation, compared to interdisciplinary collaboration.

Conclusion
Through synergetic effects, it follows that transdisciplinary social network intervention may also improve results in other cases involving the same professionals. This may occur through the generation of more flexible solutions for the help seekers based on increased levels of reciprocal confidence among the professionals. Moreover, the focus on person centredness followed by a change in the helpers' position may in turn affect the stereotypes associated with professionals.
need of respect for eclecticism in order to achieve collaborative processes which will lead towards transdisciplinarity.

The development between the meetings
Mutual confidence was the main topic for the professionals in the Health Care Group at the first meeting, whilst this idea became important to the professionals in the Social and Educational Group during the second meeting. Professionals in the Health Care Group pointed to difficulties in arriving at a common understanding during the first meeting, whilst this is illustrated by the Social and Educational Group during the second meeting. As problems of identification and negotiation existed, identity changes were affected by this factor.

Discussion
Our aim in this paper was to explore the challenges to professional identity in multi-agency network meetings, focusing on the way attitudes towards multi-agency practice are embedded in traditions of specialization in the sense of professional knowledge and mutual interaction and more specifically, how professional identity is related to: the development of professional roles in multi-• • agency network meetings the development of transdisciplinarity in multi-• • agency network meetings To conclude the findings: Professional role: Reconciliation to transdisciplinary roles emerged for some members of both groups. Other members found role release unfeasible and called for traditional and therefore interdisciplinary roles. The role-developing processes were during both meetings strongly affected by the anticipation of stereotypical roles by those who were unfamiliar with network meetings.
Teamwork: Professionals in the Health Care Group were affected by a sense of insecurity towards other members in the network. They underlined the importance of having familiarity with each other in order to increase mutual reliance and thus increase possibilities of transdisciplinarity. Professionals in the Social and Educational Group discussed their position by pointing to lack of complete acceptance by professionals in the Health Care Group. They also argued about the impact of the exercise of authority and responsibilities in terms of laws for transdisciplinary collaboration.
In this study, the findings reveal a reconciliation of professional roles during the project in which an emerging well as positive directions. Moreover, the term 'interdisciplinary' was used in Project Joint Development instead of the term 'transdisciplinarity'. Although the substantial messages concerning collaboration given to the professionals was in terms like holism, transcending, creativity and flexibility, the fact that the professional were not familiar with transdisciplinarity as a key concept may have influenced on the analysis and interpretation.
Moreover, most of the help seekers in this study were suffering from frequent mental health problems and consequently, the professionals have mainly been dealing with difficulties that are fairly common in their daily work. We consider this to be a benefit in terms of credibility. Bearing this in mind, we believe that the findings from Project Joint Development may have relevance in other settings where the aim is to implement and work in a transdisciplinary way involving both professionals and non-professionals in the health, social and educational sector. This is due to the fact that challenges to transdisciplinary collaboration are so general in the sense that intimate teamwork is related to such factors as communication, motivation and enthusiasm [14]. Conversely, projects that aim to implement transdisciplinary collaboration in the sense of social network intervention will to a certain extent gain the benefit and meet the challenges reported in this paper. Bearing this in mind, the increased familiarity between the professionals developed in transdisciplinary multi-agency teamwork may improve the health care system in general [3]. However, the results also illustrate several challenging aspects with respect to the achievement of successful transdisciplinary collaboration. Thus, according to the findings in Project Joint Development, we need to address the following: Emphasize motivation and personal commitment • • [14]. Even though role reconciliation seems to be more challenging for professionals representing peripheral agencies, the inclusion of these professionals is equally important as professionals representing the leading paradigm. Increase the professionals' ability to become • • familiar with each other [30,42]. This means being aware of the importance of creating meeting places (such as training and supervision groups) for the professionals in order to increase confidence and trust. Lack of familiarity between the professionals can hamper flexibility and creativity in transdisciplinary teams and should be addressed adequately in the development of integrated care solutions. Nurture professionals who have a preference for • • teamwork. Through their assessment of teamwork as being especially important, these people may contribute greatly to the genuine integration and expansion of integrated care. Be aware that cultural barriers contribute to delaying • • in the process of integration among professionals and non-professionals in multi-agency work. Nonetheless, reinforce the efforts in order to develop knowledge and practice concerning collaboration involving a variety of different voices.

Methodological considerations
The sample in this study represents a great variation in professionals and agencies and hence it provides us with great diversity in the information. However, this diversity means that the dialogues held in the focus groups may have produced different information because of the different affiliations and attitudes linked to each professional [33]. These differences may have been even more reinforced as the project was initiated by the local hospital and was also managed by the agencies involved. The fact that the focus groups were arranged in the child and adolescent clinic underlined the fact that ALH was working for the hospital. However, when combined with the fact that ALH had also observed much of the previous activity linked to Project Joint Development and therefore was a well-known person to the informants, this may have affected the information, in negative as