Early-intervention services for first-episode psychosis are the result of a service reform introduced in England during the past decade. Their purpose is to enable earlier diagnosis, assertive management, and sustained contact with patients aged 14–35 and experiencing a first presentation of psychosis (1,2). These services are now part of the international health service landscape, with early-intervention services in Canada, Australia, New Zealand, the Netherlands, Italy, Germany, and Scandinavia (3,4). In the United States, a number of pioneering early-intervention services have been established, and numerous research projects are under way (5,6,7).

Early-intervention services in England and elsewhere are encouraged to collaborate across many diverse organizational boundaries, particularly between adult mental health services (AMHS) and child and adolescent mental health services (CAMHS) (8). Collaboration between these agencies is needed because the incidence of psychosis starts to rise sharply in the 15- to 18-year age range and the special needs of this young group cannot be adequately met by the skills of CAMHS or AMHS alone. Given that CAMHS and AMHS are often separately funded and managed, as well as based in different organizations, key tasks cited in policy and practice documents for early-intervention services in relation to developing the interface with CAMHS are "to agree on ways of joint working, referral protocols and follow-up care" and to "develop shared training with CAMHS to extend their expertise on developmental and family issues" (9,10,11).

However, despite this policy emphasis on collaboration, evidence suggests that the implementation of policy and the development of "cross-boundary" services, especially in mental health in the United Kingdom (UK), have been variable. Difficulties in establishing effective partnerships in mental health services include incongruent geographical service boundaries, professional boundary issues, and diverse arrangements for health and social services (12). In the UK, collaboration in the context of services for adolescents has been particularly criticized (13). A recent review of services has identified that specialist services, which include early-intervention services, are still not available to most young people (14). This shortcoming is also reflected in the wider literature as an international problem in both developed and developing countries (15,16). At an international level, funding of health care systems, access, and availability to public mental health systems affect how services develop and work together (17).

The contemporary view of modern mental health services in the UK is that they should be community focused and integrated with other health care, and they should offer a seamless interface between the different partners involved. However, an evaluation of services in development has shown under-resourcing for a comprehensive approach to managing the patient and family, widespread variations in service availability, and fragmented service development in some areas (17,18,19). There is a lack of consensus as to when CAMHS and AMHS should collaborate in patient management, what model this collaboration should be based on and what responsibility should be retained by each service (20,21).

The consequences of poor communication and partnership for young people with symptoms of psychosis are far reaching and may disrupt many aspects of their life, including education, employment, and physical and mental well-being (22). The aims of this study were therefore to explore, in detail, the barriers, facilitators, and types of relationship between CAMHS and early-intervention services and to highlight ways of working and best practices that could improve the care offered to young people with first-episode psychosis.

The settings for this study were 20 primary care trusts (PCTs) and three Strategic Health Authorities (SHAs). Fourteen of the trusts were located within the three SHAs and six were outside these SHAs. The six additional services were included in the study because they represented sites of particular interest either because of their model of early-intervention services or because of specific issues or difficulties in establishing the service. The three SHAs covered a diverse geographical area with varied levels of hardship.

In the UK, primary care services are managed by 152 PCTs. Each PCT covers a separate local area. PCTs receive about 80% of the total National Health Service budget directly from the Department of Health. The role of PCTs is to decide what health services a local community needs and to provide and commission these services. They are also responsible for delivering national health policy at a local level. SHAs are responsible for larger areas of England and incorporate a number of PCTs. The roles of SHAs include strategic planning of health services, monitoring the performance and standards of PCTs within their area, supporting PCTs in implementing national policies into practice, and ensuring that national priorities are integrated into local health service plans.

From these different sites, 97 executives with a responsibility for mental health and 75 leaders and team members from early-intervention services and CAMHS were invited to participate. Four focus groups were held with PCT executives from different managerial levels and two were held with team leaders and team members from early-intervention services and CAMHS.

For each set of interviews and focus groups, topic guides were constructed to reflect the variety of expertise among stakeholders but with common core questions relating to early-intervention service roles and responsibilities, partnership facilitation, planning, finances, commissioning issues, and challenges associated with implementing and establishing early-intervention services. Each interview and focus group (conducted by EE) was fully audiotaped and transcribed, and field notes were written up.

The constant comparative method, guided by the Framework approach, was used to analyze the data (22). This method is based on a grounded-theory approach to data analysis in which theories are generated from the data. Each transcript was read and reread with the field notes and analyzed concurrently with data collection. Disconfirming evidence was sought throughout, and emergent theories were modified in response. Interviews continued until data saturation was achieved and no new themes were emerging. Respondents were sent a copy of their transcript and emerging themes; they were invited to comment, and their views were then incorporated into the analysis.

The 172 individuals in roles identified as key to the study were approached and invited to participate, and a total of 142 agreed to a semistructured interview. Four PCT mental health commissioning positions were vacant at the time of the interviews. Other reasons cited for nonparticipation included ill health, a lack of time, and unfamiliarity with the role. Six focus groups involving 31 participants, including PCT executives and joint commissioners as well as team leaders and team members from early-intervention services and CAMHS services, were held between February 2004 and September 2007. Interviews took place with senior executives, directors, chief executives, children's service directors, mental health leaders, joint commissioners of AMHS and CAMHS, commissioners of children's services, middle management executives, and early-intervention and CAMHS leaders, managers, and team members.

This article discusses the key themes emerging from the data that are directly relevant to the implementation of early-intervention services and the development of collaborative working partnerships between early-intervention services and CAMHS: communication, joint learning strategies, the interface facilitators, and the development of innovative service models.

In general, most individuals, at both the operational and strategic management levels, recognized that poor communication between early-intervention services and CAMHS presented a potential barrier to working as partners. Some CAMHS leaders appeared to justify their lack of communication because of what they perceived as a lack of need for early-intervention services, with one commenting that "There is something about the relative infrequency and the incidents of first-episode psychosis in children." Another CAMHS leader added, "I have had absolutely no involvement with early-intervention services, so I couldn't possibly comment or add anything of use at all."

A lack of understanding between the CAMHS and early-intervention services often appeared to impede working together. However, CAMHS and early-intervention service team members who had participated in joint training and educational initiatives described how an improved awareness of each other's priorities, philosophy of care, and ways of working had helped in breaking down some of these barriers. Two early-intervention service leaders commented on the value of joint training: "Through this [joint training] we have better knowledge and understanding of the needs of a young person being transferred from CAMHS, not just 'Here you are, your new team—goodbye!'" and "[Without joint training sessions] it would have been harder to convince people internally and at the practitioner and directorate level about the need to bring together adult and CAMHS teams as some sort of transition service based within early-intervention service."

One of the most important facilitators at the interface between early-intervention services and CAMHS was senior support either from an individual, such as a senior PCT or SHA executive, or through the involvement of an individual who had greater expertise in the area of mental health and early-intervention service development. Such persons appeared to facilitate the exchange of information and resources and to help those involved at the interface to negotiate past any difficulties that arose. One early-intervention service manager commented about an individual who had helped: "He's very useful because he's clicked in to all of the services right across the region. Because of his involvement he is aware of some of the barriers that other services have come across when trying to work with CAMHS." A senior SHA leader identified how useful a facilitator could be: "The chairman of one of the trusts regularly attends the early-intervention network, and that's good because chairmen meet chairmen and he's quite a champion for mental health and early intervention." A CAMHS leader agreed, reflecting on the value of facilitation and support: "Our manager—our direct manager above me—is very supportive and so are some of the PCT commissioners higher up within the trust, and I think we've even got a champion in the chief exec."

Several interviewees described how they had developed a particularly collaborative relationship with CAMHS. With support from their PCT or other trust, these early-intervention service leaders and teams had created innovative solutions to the problems at the interface between early-intervention services and CAMHS. Diverse methods of making the partnership work emerged, each requiring various levels of collaboration, innovative thinking, and commitment and each with different levels of success. These methods included basic protocol development, requiring the least commitment and innovative thought; hiring a generic children-adult service link worker; creating a position for a link worker who would focus on the interface between early-intervention services and CAMHS; and the most successful method, the development of a new service model altogether. This fulfilled all of the criteria set forth in the Mental Health Policy Implementation Guide (MHPIG) from the Department of Health but also went above and beyond that guidance, developing the service into what might be termed a "youth-focused model."

The youth-focused model was the development of a service that was neither CAMHS, nor AMHS, nor early-intervention services but included elements from all of these teams. This was the most complex method of bridging the early-intervention service-CAMHS divide. It required innovative thinking, senior support, and considerable commitment of resources, time, and energy. Diverse age-appropriate services were gathered in one place, which facilitated access and enhanced continuity of care for patients. In addition, housing multiple teams and organizations under one roof was beneficial for staff too, enabling access to each other for meetings, advice, and training. This type of model lessened the need for more formal service support, and although a number of youth services had considered the development of protocols, few had yet found the need to do so. In addition, some of the barriers to a workable partnership that early-intervention services had experienced with CAMHS, such as historical, geographical, and philosophical differences, were overcome because there was a greater sharing of knowledge and working practices with team members because they were at the same site. Two senior managers described the underlying processes that resulted in the development of the youth-focused model: "Social enterprise is very much the focus. You look at … who we want to set up a youth-focused service with to deal with these young people with psychosis or whatever. Who are the partners they want to be engaged with? You might want to consider something like pregnancy advice services, but there are loads of key agencies that need to be coordinated and pulled in. That's where these youth centers can fill the gap" and "I think the strength in developing this one-stop-shop service is this grouping—this peer support group—they swap and support each other."

This study adds to the relatively small body of literature exploring the interface between CAMHS and early-intervention services. We have described how poor communication can affect many areas of service development, including strategic planning and the ability of teams to work in partnership. In addition, this study has identified that the gap that exists between CAMHS and AMHS is also present at the interface between CAMHS and early-intervention services. This discovery is important because one of the underlying premises of early-intervention services is to strengthen collaboration between CAMHS and early-intervention services.

However, this study also offers some optimistic findings, with actions and behaviors from CAMHS and early-intervention services that encouraged the development of good working relationships. Innovative early-intervention and CAMHS services collaborated more effectively through joint training and educational initiatives, the development of a youth-focused model of working, and the involvement of key individuals who acted to facilitate this process.

A number of limitations were identified in this study, one being that perhaps too many interviews were carried out during the data collection period. However, midway through the study, SHAs and PCTs underwent major reorganization, so we felt it important to continue to interview throughout this time of significant change. During the data collection, relatively few individuals from social care were interviewed and no service users were interviewed, which might also influence the usefulness of the findings.

The findings of this study share similarities with the findings of other studies of partnership at the interface between AMHS and CAMHS in that communication tends to be generally poor, there is difficulty in engaging the services to develop a collaborative partnership, there is a lack of strategic planning, and the liaison between agencies is poor (23,24,25,26). In addition our findings reflect the findings of recent studies that have explored potential service models for early-intervention services and identified how a greater focus should be placed on neglected areas such as relapse prevention, functional recovery, vocational recovery, youth-focused services, or a specialist youth mental health model (27,28,29,30).

This study has, however, also highlighted the importance of the role of the "interface facilitator" in improving the provision of care for young people with first-episode psychosis. Indeed, the role of an interface facilitator was described by many early-intervention services and CAMHS leaders as key to service development. This individual could be employed at any level of the health service but often had a mental health background or an interest in early-intervention services for first-episode psychosis and mental health. One function appeared to be to steer and coordinate the different individuals involved at the interface of early-intervention services and CAMHS.

One of the most successful models of service development in terms of reducing the gap between CAMHS and early-intervention services was the development of an all-encompassing youth service. This service generally did not require the role of a link worker as a contact between CAMHS and early-intervention services, because communication in this instance was enhanced because both services were housed in one building and were part of an overarching organization, known as the youth service. This also promoted increased frequency of contact, meetings, joint training, and educational initiatives between the two services.

Many studies exploring the challenges in creating effective partnerships in mental health have suggested that hierarchical policy-driven solutions may be required to drive through change (31). This study differs in that cooperation between the different parties involved, which were supported and facilitated by a "champion" who steered the different groups toward a common goal or purpose, resulted in more successful partnership practices at the interface between early-intervention services and CAMHS. An important factor in that success was that the balance of power between the different groups appeared relatively equal, with each group sharing and participating in decision making.

This study suggests that the more successful models of partnership at the interface between early-intervention services and CAMHS had a number of shared characteristics, which included an element of leadership and a named individual who steered and facilitated the relatively autonomous individual stakeholders toward joint goals by enhancing communication and coordination among them. In general this individual tended to focus on the facilitative aspects, rather than on the leadership aspects of their roles, thus encouraging the groups involved to become more interdependent. Administrative direction and control were dispersed among the participants who had diverse roles and who acted as organization representatives. Decisions were made through negotiation and were coordinated by the facilitator. These different strategies enhanced a number of processes, including greater equity and sharing of funds, resources, and knowledge; increased communication and consultation; and more inclusive decision making. All the relevant parties involved participated in decision-making processes in a nonhierarchical fashion, which increased the concept of joint ownership and encouraged a local focus to service development.

At the managerial or administrative PCT level, however, there were many perceived challenges to providing a youth-focused intervention embedded within an adult mental health service. It has been argued that early-intervention services themselves are an expensive treatment option, and the additional demand to develop transitional youth-focused models of care further adds to the costs of developing and implementing these services (32,33). A number of health care executives continued to feel that the cost-benefit rationalization of providing early-intervention services was an ongoing challenge and needed more consideration before further expenditure. It is possible that the development of service models reflecting elements of the youth-focused service may overcome some of these challenges through encouraging negotiation and communication between the different members of the group and the development of potentially innovative solutions.

Early-intervention services are given clear guidance in the MHPIG and in other key documents that support early intervention's developing partnerships with CAMHS. However, there is little guidance on the model or the quality of relationship expected and little policy support to underpin this relationship. The complexities of the development of an interface between early-intervention services and CAMHS have implications for policy development and training. Future policy and guidance will need to clarify these issues, and training will need to emphasize the need for collaboration and communication between the two services.

The aims of this study were to explore, in detail, the different types of relationship between CAMHS and early-intervention services. It appears that despite policies encouraging the development of, and supporting the need for, effective partnership at the interface between early-intervention services and CAMHS, cooperation was rare. This study found that rather than hierarchical policy-driven solutions, cooperation between the different parties involved, supported and facilitated by a "champion" who steered the different groups toward a common goal or purpose, resulted in more successful working practices between the two services. Elements that led to more successful partnership development included joint learning and training, designation of senior-level champions, joint operational policy or protocol development, and use of specific link workers to facilitate the interface. The most successful approach was to develop a separate youth-focused service that placed multiple teams and organizations responsive to younger people's needs (such as education, employment services, social activities, pregnancy services, and peer support) under one roof. The findings from this study may help determine the best model of partnership development for early-intervention services and CAMHS in England.

This work was funded by a trainee award from the National Institute for Health Research (Department of Health) to Dr. England. Ethics Approval was given by South West Multi-Research Ethical Committee Form MREC/03/6/54 and MREC/03/6/59.

The authors report no competing interests.