“People just don't understand their role in it.” Collaboration and coordination of care for service users with complex and severe mental health problems

Abstract Purpose To explore professionals' and service users' experiences and perceptions of interprofessional collaboration and coordination for service users with complex and severe mental health issues. Design and Methods A qualitative study involving semi‐structured interviews of professionals and individual interviews of service users. Data were analyzed by thematic analysis. Findings Participants described challenges and suggested improvements concerning Distribution of roles, responsibilities, and tasks; Communication; and Knowledge and attitudes. Practice Implications Mental health nurses and other professional helpers should have a particular focus on common aims, clear division of roles, planning and timing of interventions, and communication with other professionals and service users.

provide appropriate care for by nurses and other professional helpers. These service users need coordinated, continuous care from a range of practitioners in multiple health and social care services.
They often have cooccuring substance abuse, are overrepresented in police interactions and compulsory treatment 18,19 and account for a majority of spendings in health and social care. 20 Mental health nurses and other professional groups spend much of their time and resources on this group of service users. Optimal collaboration and coordination for and with this group are particularly important, as it is a group in need of continuous and consistent care and it is particularly vulnerable to breaks in relational and informational continuity. 21 However, there is a gap in the evidence pertaining to collaboration and coordination with and for these service users with complex and severe mental health issues.

| Aim
This multiperspective study aimed to explore the collaboration and coordination among service providers relevant for service users' recovery. Further, the study aimed to inform future improvement of

| Study design and setting
The study was a qualitative multiperspective study conducted in the geographical area of a Local Health Trust in Norway. In Norway, the responsibility of providing primary health services lies with the communities and specialist mental health and addiction services lie with the health trusts. In the communities, community mental health nurses, mental health community teams, supported housing, day centers, and home-based services are available.
General practitioners (GPs) and primary care emergency services in the communities serve as gatekeepers to specialist services. The police are involved in bringing in service users considered a risk to themselves or others to their GP or the primary care emergency center. GPs there may further refer service users to community mental health centers (CMHCs; secondary care) or to hospitals (tertiary care) for specialist outpatient or inpatient treatment.
Outreach teams at the CMHCs are responsible for service users with complex and severe mental health and/or addiction issues who are particularly hard to reach or challenging to relate to. Further relevant institutions are child protection services and the Norwegian Labour and Welfare Administration ("NAV"). Each service has an electronic journal system with information that is only available to the specific service. Access to information is strictly regulated by legislation.

| Participants
Six individual interviews with service users with complex and severe mental health problems were performed. They were all using the outreach team at a CMHC responsible for mental health specialist services to a population of approximately 34,000. The outreach team served 40 service users with severe mental health conditions, very difficult life situations, and behaviors that were perceived by professionals as particularly challenging. The service users were invited to participate in the study via their primary contact at the outreach team. They were interviewed about their experiences as service users. The semi-structured interview guide included questions about collaboration and coordination in terms of how the service users experienced that the professionals understood and responded to their needs, how professionals worked together, and how collaboration and coordination could be improved. Five interviews were conducted in the participant's home and one in a car at a parking lot.
Nine homogeneous group interviews with mental health nurses and other professional helpers from relevant services were conducted. The semi-structured interview guide included questions about how participants experienced and perceived collaboration and coordination for service users with complex and severe mental health problems, and how collaboration and coordination for these service users could be improved. To increase the internal validity of the study, professionals were asked to describe service users they perceived as particularly challenging to relate to and provide appropriate care for before they were asked about their experiences and perceptions of collaboration and coordination for and with these service users.
Interviews lasted approximately one to one and a half hours; they were audiotaped and transcribed verbatim.

| Analysis
A data-driven stepwise procedure in line with thematic analysis was used. 22 The data analysis proceeded as follows: the first author systematically coded all text material and defined preliminary themes. The researchers discussed and agreed upon a common understanding of the semantic and latent constructs underlying the material in the preliminary themes. Based on the common understanding reached during these discussions, the first author made the final categorization of the contents and drafted the BIRINGER ET AL.
| 901 manuscript based on the notes from the discussions. To ensure the internal validity of the study, results were compared with the original transcripts throughout the writing process. Acknowledging that the researchers' own involvement and prior understanding may influence on which knowledge is acquired, 23 reflexivity was practiced throughout the study by holding team meetings and discussing possible interpretations of semantic contents and latent constructs underlying the data. Coding was performed in N'Vivo 12.

| Ethics
Approval for this study was obtained from the Norwegian Social

| RESULTS
The characteristics of public services, professionals, and service users are shown in Table 1. Out of the 36 participants with professional backgrounds, 28 were women and 14 were nurses. All participating service users were male and their age ranged from 21 to 69 years; the mean age was 37 years. All service users had severe problems in a range of areas-including family, social network, work, and housing-and all of them depended on social benefit payments for their daily support.
Professionals provided rich and detailed descriptions of their experiences of collaboration and coordination in and among services as well as of personal and contextual characteristics of service users with complex and severe mental health problems who they perceived as challenging to relate to and provide appropriate care for. They described the situation of service users and their professional helpers as a "vicious circle" (Figure 1). The vicious circle started when the service users were discharged from or left specialist treatment. When the next crisis occurredfor example, their psychotic symptoms worsened, their use of drugs escalated or their next-of-kin became exhausted-the service users visited their GP or the primary care emergency center, or they were picked up by the police, to be admitted to specialist services again. Professionals' responses to service users' crises were described as "back and fort" and "firefighting," with frequent disruptions in information flow and personal relationships between service users and professional helpers.
Further, professionals experienced frustration because the tools or services that could have met the service users' needs sometimes were unavailable or did not exist, and as professionals frequently had to take on the unpleasant gatekeeper role when service users were supposed to get help via other services. In response to the question of what characterized service users they experienced as challenging to relate to and provide appropriate care for, professionals gave the following descriptions: chaotic lifestyle, threatening to take their own life, frightening or threatening others, do not wish to use services, do not keep agreements, disagree with professionals' views of problems, at the limit of their ability to take care of themselves, in need of care and support for many years, and lacking material resources.
In terms of diagnoses, comorbid mental health and addiction problems, relapse of psychosis, unstable personality disorder, comorbid somatic conditions, and "diagnoses that professionals know little about" were mentioned.   In particular, compulsory admissions to specialist mental health services were experienced as a "big problem" by professionals, as they were so troublesome to administer for the GPs and other involved parties. The route into specialist mental health services via the GPs on call at the primary care emergency center was experienced as a time-consuming "detour": I think the most frustrating is that I feel that the primary care emergency center is the wrong place for these patients. I feel that I am the wrong person, and I have to handle them and I am responsible for them. They [the service users] need someone to talk to, they need backup, and they need to be heard…. And there isn't time and space for that at the primary care emergency center.
(P24, GP at the primary care emergency center) Many participants, however, experienced specialist mental health services as unwilling to take in service users in crisis, although no satisfactory solutions for treatment and support of the service users were available in the community. Professionals in specialist mental health services, in contrast, felt they were under pressure due to limited capacity in their services. One consequence of the limited capacity was premature discharges and the "vicious circle" described in Figure 1. Availability of services rather than support needs often determined which service the service users were offered. For instance, nurses in the community home care services frequently had extensive contact with severely ill service users because the nurses were available 24/7.

| Available information about the service users
Participants from the police, acute psychiatric ambulatory team, and GPs described that they sometimes had to "jump into" very challenging and risky situations without having the information necessary to deal with the service users in a professional, proper, and con- Restrictions concerning the sharing of information and communication about the service users across services were frequently mentioned as causes for loss of information, delay in handling issues, extra hassles, and increased risk. Nurses in the community services responsible for following up with service users after discharge from the specialist mental health services were unsure of what to do and how to react when the service users' problems escalated, as they frequently did not have access to updated treatment-and crisis plans. Access to updated information about the service users' needs, their professional and social network, current situation, and updated plans for treatment was deemed necessary to make appropriate decisions.

| 905
Ad hoc contacts paved the way for more flexible solutions tailored to the need of the service users. Professionals valued ad hoc contacts with other professionals for exchange of advice and information. Such exhange made professionals feel that the decisions they made were appropriate and risks reduced.

| Meeting places and systems for communication
Professionals described the existence of "brick walls" between services. They complained that opening hours of most services were limited and reaching professionals by phone was difficult. They pointed to the need for effective systems for rapid exchange of information between services.
Some service users had established interprofessional support groups that regularly met, with or without the service user present.
Several service users found the group meetings useful, as they ex-

| Professionals' knowledge and attitudes
According to participants, the professionals' background, training

| DISCUSSION
In this multiperspective study involving relevant services and service users, mental health nurses and other professionals pointed to mutual respect, common goals, information sharing, and knowledge of other available services as important aspects of collaboration and coordination. Information transfer shared understanding and working atmosphere are associated with handoff quality. 24 As a comprehensive approach including a range of public services, including the police and child protection services, and first-hand experiences were taken, this study most likely has high socio-ecological validity and transferability.  26 and models of care with a high degree of interprofessional collaboration. 27 Many international studies point to a gap between policy aspirations for care coordination and personalized care planning, and current practice. 28 As for the organization of public mental health services, a central question raised by the present study was whether the same professionals should be responsible for a service user's care across inpatient and outpatient settings (continuity of care) or whether there should be separate teams (specialization). 29 The current reforms in Europe are inconsistent with regards to the question of which to favor, although the current study and existing evidence suggest better outcomes and stakeholder preferences for continuity of care systems. 29 In line with the present study, previous studies point to peer support and inclusion of family members in planning and follow-up as central in collaboration and coordination. 26,30,31 Integration of education of professionals and service user involvement in services may lead to mutual understanding and respect among professional groups and between service users and professionals. Fortunately, some recent developments have been made in Norway. Specialist addiction and mental health services are being colocated and electronic communication systems allowing for requests and transfer of administrative or clinical information between community care, GPs, and specialist services are available.
The study has some limitations that should be recognized.
All researchers had professional backgrounds, and coresearchers with experience as service users were not included to strengthen the service user perspective in the study. Although the researchers were aware that their preconceptions could affect the questions asked and conclusions drawn, intersubjective elements may have influenced data collection and analysis. Information from service users was not very rich or in-depth, and some experiences were schematically described. This may, however, suggest that the service users iterviewed represented the intended group.

| CONCLUSIONS
This multiperspective study provides a valuable comprehensive understanding of collaboration and coordination for and with service users in and across mental health and social care and the police. The study points to a need for involvement of service users and next-ofkin, interprofessional education, supervision, and meeting places and systems for communication to improve collaboration and coordination. Personal planning, care coordinators, and support groups are valuable tools for the improvement of relational continuity and coordinated assistance with housing and work, and mental health care.
Professionals point to a gap between existing services and service users' needs. Structural and legal changes are still necessary to improve care for service users with severe and complex mental health problems.

| Implications for nursing practice
The gap in collaboration and coordination between professionals and services is consistently identified as a major impediment needing effective solutions. Findings from this paper emphasize the value of nurses' and other professionals' efforts to improve communication with professionals, service users, and next-of-kin. Mutual respect, common goals, information sharing, and knowledge of other available services are important aspects of collaboration and coordination.
Collaboration and coordination can significantly improve the quality of service delivery for persons with complex and severe mental health issues, that is, service users who are particularly vulnerable to breaks in the continuity of care.
T A B L E 2 (Continued)

Suggested improvement Source
A mental health nurse from the community mental health services should be available to the primary care emergency center Primary care emergency center Paper exercises help professionals get to know about available services and how and when these should ideally be contacted and work together Acute ambulatory team Child protection services and community mental health services should collaborate more about parents' mental health issues

Child protection services
Mental health services should focus more on the welfare of the children and more often send notifications of concern to the child protection services Child protection services Abbreviation: GP, general practitioner.