Progress in the development of integrated mental health care in Scotland.

Abstract The development of integrated care through the promotion of ‘partnership working’ is a key policy objective of the Scottish Executive, the administration responsible for health services in Scotland. This paper considers the extent to which this goal is being achieved in mental health services, particularly those for people with severe and enduring mental illness. Distinguishing between the horizontal and vertical integration of services, exploratory research was conducted to assess progress towards this objective by examining how far a range of functional activities in Primary Care Trusts (PCTs) and their constituent Local Health Care Co-operatives (LHCCs) were themselves becoming increasingly integrated. All PCTs in Scotland were surveyed by postal questionnaire, and followed up by detailed telephone interviews. Six LHCC areas were selected for detailed case study analysis. A Reference Group was used to discuss and review emerging themes from the fieldwork. The report suggests that faster progress is being made in the horizontal integration of services between health and social care organisations than is the case for vertical integration between primary health care and specialist mental health care services; and that there are significant gaps in the extent to which functional activities within Trusts are changing to support the development of integrated care. A number of models are briefly considered, including the idea of ‘intermediate care’ that might speed the process of integration.


Introduction
The nature of integrated care, partnership working and mental health policy Delivering integrated care is a key policy objective of Scotland's new devolved government and 'partnership working' is the current approach to its achievement w1x. Successive policies have set out to address the issues associated with this objective in relation to joint planning and strategy, finance and information systems, communication, and continuity of care. Recently, there has been renewed emphasis on the implementation of pooled budgets, devolved decision making and needs-led, person centred care w2x. There can be little doubt about the importance of achieving enhanced service co-ordination, especially for people with severe and enduring mental illness in the health care organisations responsible for them. In Scotland these organisations are Primary Care Trusts (PCTs), and their associated Local Health Care Co-operatives (LHCCs), details of which can be found elsewhere w1x.
These developments can be interpreted as an attempt to close the 'policy gap' in UK mental health care identified by Secker et al. w3x. This gap is described as a consequence of two opposing trends in policy. The first trend entails the promotion of a primary careled NHS, which has followed a course from GP fundholding to the establishment of new forms of primary care organisations established after the election of a new Labour government in 1997 w4x. One consequence of this has been that GPs have been using their increasing influence to meet the needs of patients with more moderate mental health problems w5x. The second trend set out in mental health policy (details available from www.show.scot.nhs.uk) aims to ensure that specialist services retain a focus on This article is published in a peer reviewed section of the International Journal of Integrated Care people with severe mental illness. Secker and colleagues maintain that the repeated concern articulated in policy to assure a co-ordinated service for those with severe mental illness has resulted in expectations that primary health care will work with specialist mental health services to share the care of this group w3x. It has been reported that as many as 40% of all patients with serious mental illness, including 25% of those with psychotic illness are cared for entirely in primary care, and, therefore, the scope for greater integration of generalist and specialist care is considerable w6x.
A number of models have been developed to elaborate the respective roles and responsibilities of primary care and secondary care w7x. These are frequently articulated with specific reference to the interface between Community Mental Health Teams and primary care and include the following: -Shifted out patient: mental health professionals work from primary care based clinics -Liaison attachmentdesignated mental health worker acts as liaison with primary care -Consultationadvice, support and training provided to primary health care team by mental health workers -Integrated workingjoint register of patients developed by primary and secondary care, along with agreements on good practice, e.g. assessment, management of specific conditions w8x.
A number of studies have sought to evaluate such models, and in a systematic review of the effect of on-site mental health professionals on GP behaviour Bower and Sibbald w9x concluded that referral to a mental health professional in a primary care setting resulted in fewer prescriptions for psychotropic drugs, and reduced numbers of referrals to specialist secondary care, although these effects were not consistent. A subsequent commentator w10x has questioned the validity of this headline conclusion because of shortcomings in the design of some of the reviewed trials. Nonetheless, there is some evidence that closer working of primary care and specialist mental health care professionals can be beneficial.
Against this background, the objective of the research reported in this paper was to explore the extent to which the conditions for more effective integration of care were being put in place in Scotland. As a study of process rather than outcome it is appropriate to observe that integrated care may not necessarily flow from integrated structures and processes; for instance, Community Mental Health Care Teams (CMHTs) may have integrated working practices but they may not provide services which are integrated from the user's point of view, or more integrated care overall. With this important qualification in mind the specific aims of the study were to: -Examine the structures in place within Primary Care Trusts, including Local Health Care Co-operatives for the delivery of integrated care for adults with severe mental illness -Consider the interface within these Trusts between specialist community mental health services and primary care services -Explore Trusts' relationship with social work and with the voluntary sector.
The study contained both descriptive and evaluative elements, the former to document structures and approaches to deliver integrated care, the latter to consider the relationship between structure and the delivery of integrated care and to identify success factors and key challenges. The study set out to cover a wide landscape, with a view to focusing down on selected areas in more depth as themes of interest emerged. The intention was to highlight a range of illustrative experiences and examples from across Scotland that would inform wider service and practice development. As this was a relatively short-term project started in November 2000 and reported seven months later w11x it did not consider the impact of integration on quality of care or client outcome. Rather, the focus is on the features of organisations and interprofessional relationships conducive to the delivery of integrated care. The study offers a baseline against which the effects of subsequent organisational development can be assessed.

Constructs and definitions in exploring integrated care
In order to be able to gauge progress towards the delivery of integrated care for people with severe mental illness it is necessary to clarify some of the terms and constructs that are in common usage in this field. For instance, 'integration' and 'partnership' are frequently used to describe a widely varied set of actions concerned with, respectively, the objectives of policy and the means of achieving its implementation.
This research project proceeded to explore integration using a widely accepted distinction between horizontal and vertical integration. Horizontal integration refers to the bringing together of professions, services and organisations that operate at similar levels in the care hierarchy. Collaborative working arrangements between health and social work in multidisciplinary community teams are an example of horizontal integration in direct care delivery. Vertical integration refers to the bringing together of different levels in the hierarchy of care (Figure 1). This article is published in a peer reviewed section of the International Journal of Integrated Care Community mental health and well-being Non-medical approaches to positive health, to create the conditions that promote and sustain the well being of individuals and communities Self care and self help Enabling and facilitating people to look after themselves, using purposively designed information and educational materials. Can include electronic resources as well as interpersonal resources (group and one to one) Extended primary care Enhanced capacity within primary health care to respond to needs, by including for example counselling, psychology. Can also encompass more effective signposting to and use of wider community resources, such as educational facilities, money advice and recreational activities Intermediate care Drawing on a range of professional skills, ensuring smooth and timely throughput at points of transition by the provision of respite, rehabilitation, step down and step up services Secondary care Multidisciplinary and multi-agency care generally focusing on the needs of those with severe and enduring mental health problems, in both community and hospital settings Tertiary care Highly specialist advice, treatment and care that combine direct provision with advice and consultation to other services and other levels of care On the vertical dimension, our research looked in particular at the relationship between primary and secondary care in providing services for people with a severe mental illness.
The extent of integration was explored by reference to a number of functional activities undertaken in health and social care organisations: The direct delivery of care

Strategic planning and development
The study also considered the links and relationship between these functional activities in relation to mental health care and the extent to which these activities were themselves horizontally and vertically integrated.
Partnership is an attractive concept since it conveys the collective efforts of diverse talents being employed to overcome difficult problems. It is also a concept that is easy to grasp since partnerships are such a common feature of human relations that there are numerous points of reference that help us gain an insight to the idea of partnership working. But, what makes a successful partnership and how can that success be measured? A distinction can be made between co-operative partnerships and co-ordinating partnerships w12x.
-Co-operative partnerships are characterised by enlightened self interest; where partners pursue own goals most effectively by co-operating with others; low investment of time and effort to maintain; low trust, high control, tendency to focus on more superficial issues and produce short term solutions.
-Co-ordinating partnerships are characterised by: mutual trust, deeper level of understanding of purpose, added value that will accrue and improved outcomes anticipated.
As current Scottish Executive policy appears to seek a shift from the former model to the latter, it is helpful to be able to place partnerships on a spectrum of relationships ( Figure 2).
While policy might be said to give primacy to organisational restructuring as a lever for reform, there is evidence that change in organisational and professional cultures does not necessarily occur spontaneously when organisational architecture is redesigned w13x.
A substantial literature w14x exists on the factors that need to be addressed to achieve the shifts in profes-This article is published in a peer reviewed section of the International Journal of Integrated Care  The development of new ways of working giving primacy to reflection and learning that is also inclusive and involving.

Methods and approach
The research was exploratory and gathered a broad range of information both on structures and on current practice in developing and delivering integrated care for people with a serious mental illness. Methods were largely though not exclusively qualitative. Information collection involved: a survey using a questionnaire with pre-coded items, open ended questions and self ratings on a scale provided; semi structured interviews (phone and face-to-face (which followed a semistructured format, with direct questions and follow on probes to elicit more detailed information and to explore experiences and perceptions in greater depth.
The study was carried out in three phases: First: A postal questionnaire was distributed to the 14 Chief Executives of all Primary Care Trusts and Scotland's three Island Health Boards. These latter serve small communities and the creation of separate trusts was considered inappropriate. All 14 PCTs responded. The Island Boards were not able to complete the questionnaire within the time scales available but a representative from each agreed to be interviewed by phone. The survey gathered general information on structures and key objectives, interfaces and links between primary and secondary care, between hospital and community mental health services and between the Trust, their local authority social work colleagues and the voluntary sector. Respondents were asked to identify factors that facilitated and inhibited each of these partnerships and to give examples of effective partnership. The questionnaire asked for named contacts in community based mental health services, in primary careyLHCCs and in the local authority social work department, to take part in follow up stages. Interviews amplified on the survey responses and provided access to a range of different perspectives.
Third: A series of six case studies in selected LHCCs. These were purposively chosen to include areas from different parts of Scotland. Selection was informed by the first two stages of the study that indicated varying degrees of maturity in partnerships as well as differences in approach to integration and partnership working. Case studies were, therefore, chosen to allow the research to explore aspects of integration and partnership working in some detail. In three areas, representatives from the non-statutory services were also interviewed. In one LHCC a Com-munity Psychiatric Nurse involved with a primary care based mental health service, separate from the CMHT was also interviewed. To preserve the confidentiality of respondents, case studies are not named in this paper.
A summary of the case studies is provided elsewhere w11x.
Each of the PCTs was invited to nominate a representative to attend a Reference Group, which met on two occasions: at an early point in the data collection process and as this was nearing conclusion. The Reference Group acted as a sounding board to allow the research team to test out emerging themes and to provide an additional insight into experiences in a range of Trusts. Representatives from seven PCTs attended the first meeting and all but one of these attended the second. Participants included people with responsibility for clinical andyor operational management of services. The meetings were facilitated by the principal researchers and a written summary of the main points of discussion was shared with participants.
Survey responses were summarised on a database and coded. Notes were made of the phone interviews and a written account completed for each one. Data from the phone follow-up and face-to-face case study interviews were analysed using qualitative techniques. Case study interviews were recorded and transcribed. The data were subjected to staged content analysis, building out from the initial themes investigated, to develop and elaborate on themes in progressively more detail.
Initial analysis of the survey and phone interviews was undertaken prior to refining the areas for exploration in the case studies, to use these as an opportunity to consider key areas in depth and to contrast differing perspectives.

Results
The next part of the paper considers the main themes and issues to emerge. We begin by exploring horizontal integration by reference to Trusts' relationship with social work and with the voluntary sector; and then consider the vertical integration of primary and secondary care and the integration of functional activities.

Horizontal integration: Primary Care Trusts and local authority social work
This sections considers progress towards and factors associated with horizontal integration between two sets of partners: firstly Trusts and local authorities, (principally social work) and secondly between Trusts and voluntary sector (not-for-profit) organisations. Partnership working and integration are considered in relation to the range of functional activities outlined earlier, spanning from the direct delivery of care, through management processes to strategic planning and development.

Building partnerships
Fractures in the relationship between health and social care responses for people with severe mental illness have been a recurring concern of much of recent policy and practice guidance in the UK. It was striking in this study that PCTs and Island Health Boards regarded partnership with social work as key in order to achieve their organisational objectives. Indeed, local authority social work was one of the main, if not the main partner for Trusts in the development and delivery of mental health services. Trusts had put considerable energy into developing their relationship(s) with social work colleagues and several respondents observed that this might have had the effect of deflecting attention away from the primaryysecondary care relationship. One Trust respondent went so far as to posit whether primary care might in fact be jealous of the relationship that secondary care services had developed with social work. A local authority case study respondent observed that: 'Primary care does not have mental health high on their agenda. We need to get sorted in secondary care first and then make it more attractive to people.' Some Trusts (e.g. Greater Glasgow) were explicit about the investment they had made in promoting integrated health and social care as the top priority to date for this client group, an investment that had until recently overshadowed ambitions to attain closer integration between primary care and community mental health teams (CMHTs).
The interface between PCTs and local authorities was reported to have developed through a combination of formal partnership agreements, regular opportunities for meeting and discussion and informal contacts and networks. These elements were considered to complement one another and to contribute to varying degrees to the different aspects of partnership working. Several Trusts and local authorities related local progress on partnership working to the recent recommendations from the Scottish Executive w16x and to experiences with other client groups, particularly older people.

Benefits of partnerships between health and social work
Respondents from Trusts and local authorities gave various accounts of the added value that flowed from working in partnership. Experience of close working between health and social work was reported to have produced tangible benefits, by enabling partners to: -Extend the range of services and supports that can be provided, e.g. incorporate into CMHTs support staff whose roles include help with budgeting and home making -Develop effective links with a wider array of services including housing and drug and alcohol services, money advice services etc -Tackle effectively complex problems and needs without having to pass the individual on to another service -Pool expertise and streamline care to avoid duplication of effort -Work together on local issues of pressing concern -Access additional resources available for partnerships In Highland, a Trust respondent drew attention to demonstrable benefits that closer working had achieved in terms of altered patterns of service usage. This was reported to include reduced admission rates to acute psychiatric inpatient beds as a result of improvements to community services and local capacity to support people in their own homes in times of crisis.
A number of respondents suggested that the establishment of decentralised organisational structures facilitated more effective collaboration and stronger partnerships between social work and health, in addressing agreed local needs and local priorities for development. Both local authority and Trust respondents perceived that social work had accrued a significant amount of influence in working in partnerships that operated at local level. Social work departments were able to field people with authority to take part in planning and development. Local implementation groups were influential in creating closer integration between health and social work, when empowered to make decisions about spending priorities. Several local authorities were able to point to increases in spending on social care services for people with mental health problems and in the number of staff working in these services.

Integrating functional activities
The research suggested that health and social work were looking systematically at partnership and its implications, to consider collaborative working at different levels from individual direct care and service delivery through management to strategy and service development. These are considered in turn.
This article is published in a peer reviewed section of the International Journal of Integrated Care Single key worker More effective use of time Easier for user to know how to make things happen or draw attention to issues of concern Devolved budget Provides easier and more direct access to services and resources Allows monies to be used flexibly and creatively Can facilitate service change and development Can make it clearer where responsibility lies Integrated care management Any trained member of staff can assess and put forward recommendations for services Provides access to a structured assessment Care co-ordinated by one care manager Ensures people are tied into a system to review their needs on a regular basis Provides a means of involving all the key people, including the user and carers

Direct care
The case studies were used to consider elements of care delivery that policy and practice guidance indicate are critical gauges of effective joint working in providing integrated direct care. Case study respondents were asked about their understanding of the purpose and benefits of these elements (see Table 1 below).
Overall, it emerged that respondents were clear that there were good grounds for striving to develop these joint processes, to streamline service provision and to ensure that service responses were matched to individual needs. Progress was greatest in the development of joint systems to assure a single point of access, joint assessment and joint care records and single key worker arrangements. The development of joint information systems and the integration of the care management function were the two elements of care delivery least in evidence. It may be that the depth and complexity of the systems changes required to bring these developments about lie outside the jurisdiction of mental health services and would have implications for wider organisational systems.

Management processes
The study highlighted a number of areas of innovative working where Trusts and their social work partners were reshaping their management functions to support integrated care delivery. However, these tended to be the exception rather than the norm. The general pattern appeared to be that of two separate agencies struggling to find solutions that could accommodate both sets of organisational requirements and that were also able to underpin and maintain services, which assure integrated care.
There was an emerging recognition that effective joint working at one level of the organisation needed to be matched by appropriate partnership arrangements at other levels. The majority of Trusts acknowledged in the survey that the development of joint management and accountability arrangements with social work were important, to enhance capacity for integrated care.
However, relatively few were at the stage of having joint processes and systems in place to achieve this, with negative consequence for the effective delivery of integrated care. This was illustrated in a case study area where well-developed multi-agency teams were inhibited by the immaturity of joint management arrangements between the parent organisations and a lack of clarity about the extent of the teams' delegated authority.

Strategic planning and development
The survey indicated that in all areas, agreements and processes to ensure joint planning and commissioning of services were being pursued. However, only five of the 14 PCTs reported that a joint financial framework This article is published in a peer reviewed section of the International Journal of Integrated Care -Working within agreed financial framework with clear budgets, especially where the structures are complex and agencies cover different 'patches' -Securityycertainty about financial position. Imbalances in financial position of partners can be destabilising. Trust is likely to be relatively buoyant, local authority to be cash starved and less able to commit to future expenditure -Coterminosity of PCT and local authority -Co-location of key personnel -Sufficient integration of management and planning to support joint service delivery -Addressing structural complexity, without duplication and excessive bureaucracy -Acknowledging the diversity of organisational and professional cultures was in place in their area between health and social work.
In Dumfries and Galloway, a joint Mental Health Board had been recently established with responsibility for strategic planning and commissioning. Mental health services were jointly managed through a joint director of mental health services post.
Several Trusts described strategic partnerships linked to projects, generally associated with hospital closure and service re-provision. The Edinburgh Mental Health Partnership (EHMP) was established to plan the closure of long stay hostel beds at the Royal Edinburgh hospital and develop alternative resources and was reported to have achieved effective cross agency and cross-sectoral involvement. A notable feature of this partnership project was that it had a budget and was empowered to make joint decisions. Its work included overseeing a joint needs assessment process and the development of joint commissioning protocols. Other examples of project based activities included service renewal in West Lothian, to steer the development of an infrastructure of community based services to replace the local psychiatric institution.

Factors that facilitate partnerships between health and social work
Standing back from these details it is apparent that a number of recurring themes affect the development of partnerships and services integration (Figure 3). Within a single Trust area the number of social work partners could be large. Partnerships with different local authorities within this area could vary in both style and substance for a number of reasons. LHCCs were at differing stages of readiness to engage with local authorities. Trustylocal authority partnerships were also reported to be influenced by the level of development of community mental health services. For example the establishment of multidisciplinary, multiagency teams required a level of explicit agreement and understanding about the basis for partnership working. In Renfrew and Inverclyde PCT, it was reported that team development had moved at a different pace in different parts of the area and the longer established and more integrated team in Inverclyde acted as an impetus for the parent agencies to consolidate their relationship.
Trusts and local authorities made frequent reference to the cultural differences between their agencies. These differences were expressed in professional values and attitudes that came to the fore both in service delivery but also in different approaches to management. An interesting perception from one Trust was that, while the health services were now redefining 'management' as a more facilitative, enabling function operating within flatter structures, local authorities remained wedded to more hierarchical structures and problem solving models of management.
Without underplaying these cultural differences, most respondents tended to be optimistic that increased exposure to one another's values and ways of working, the growing interdependence of health and social care services in supporting people in the community and the recognition of the benefits to both users and professionals of collaborative working were powerful countervailing forces: 'People get thrown together and acquire more insight into how other professionals work'

Horizontal integration: Primary Care Trusts and the voluntary sector
Within the Primary Care Trusts, there was a common recognition of the valuable role that voluntary sector organisations played in providing more informal services and in maintaining a capacity to be innovative. It was striking that in the study the impetus and energy to develop stronger links with the voluntary sector was discernible within LHCCs. One LHCC respondent observed that not-for-profit organisations were able to make and sustain long-term relationships with clients that could be problematic for specialists in health and social services. He also considered that voluntary sector services were more conducive to supporting the individual in creating 'an ordinary life' and an identity that was not centred on the experience of mental illness.
This article is published in a peer reviewed section of the International Journal of Integrated Care A second LHCC respondent observed that the not-forprofit sector was highly skilled at networking and far more sophisticated at partnership working than were the statutory services. A third person stated that GPs were becoming more aware of the voluntary sector and what they could offer. In this area, the development of a specific primary care based initiative, run by a voluntary organisation and aimed at people with moderate mental health problems had helped to nurture links and relationships across sectors. Another LHCC respondent considered that government policy expected collaboration with the voluntary sector. This was part of the process of recognising that the promotion of health involved more than health care services.
In some Trusts it was acknowledged that health services had not developed as mature relationships with the voluntary sector as social work had. The latter had extensive experience of funding and monitoring nonstatutory services. For example in one case study area, the CMHT Senior Social Worker acted as supervising officer for voluntary sector providers, with responsibility for overseeing their evaluation and development.
A number of health respondents felt that it was important to ensure that voluntary organisations were able to be appropriately represented at strategic level and get involved in planning. The perspective from voluntary organisations indicated that they welcomed the increasing acknowledgement of their role in supporting people with mental health problems. However, it appeared that readiness to work with the voluntary sector was not universal: it emerged in one case study that some health professionals were perceived to be unwilling to operate in partnership with the sector and would not enter into dialogue.
Voluntary organisations were also concerned that they remained at one remove from decision making, with only limited ability to effect influence. This sidelining was reinforced where important decisions were determined outside of the forums to which voluntary organisations had access.

Vertical integration: primary and secondary care
Recent years have seen the development of a range of opportunities for General Practitioners (GPs) to gain increased influence over secondary care, through initiatives such as the (now abolished) GP fundholding scheme and related derivatives. The initial findings of an evaluation of the impact on mental health services w5x indicate that a consistent stated objective was to improve communication between primary and secondary care. A popular means of achieving this was to increase the number of specialist staff based with or attached to primary care. However, it was noted that these developments were viewed with some caution by Trusts who were concerned about the possible diversion of resources and staff away from those with severe and enduring mental illness. In a review of the developing relationship between Primary Care Groups and local authorities in England w17x, it is noted that primary care has tended to use collaborative opportunities with social services (and arguably with secondary caresee above) to pursue operational developments and improve the co-ordination of services. There is evidence that GP fundholders tended to use their financial flexibility to purchase practice based services to alleviate pressures on their own services rather than to engage in a strategic review of local needs with partner agencies and services w18x.
In Scotland, LHCCs are an attempt to carry forward some of the benefits of improved liaison between care levels that fundholding brought whilst avoiding the inequity of access and bureaucracy its critics alleged. In considering the locus and influence of LHCCs within PCTs in relation to services for people with a severe mental illness, it is, therefore, important to be aware of the relative youth of LHCCs. Despite the relative inexperience in primary care of joint planning and commissioning this study was able to identify encouraging indications of the emergence of new forms of collaboration.
These developing relationships were explored by reference to the different functional activities set out earlier.

Direct care delivery
Previous research has illustrated that GPs have a significant role in the sole management of severe mental illness w6x but have limited involvement in the care and treatment of people with a severe mental illness who are known to secondary care services w19x. In our study half of the Trusts reported that the integration of primary care within Trusts had led to enhanced mutual understanding between GPs and secondary care at local level rather than to a substantive shift in power or influence. Progress tended to have come about in areas where steps were being taken to devolve responsibility for secondary care services to local level, with the explicit intention of achieving closer links between primary and secondary care. One respondent observed that LHCCs in that PCT were acquiring more understanding of mental health issues and a sense that relationships were changing (Figure 4).  In this area, the LHCC had been working on an exercise to map the patient's journey, involving primary care and mental health services, including health and social work and the voluntary sector. Taking a whole systems approach to assess the operation of current processes and services had uncovered a lack of integration that the LHCC was seeking to remedy. 'Nobody round the table thought it was joined up, but it was more fragmented than any of us had really understood. It was a shock'. This LHCC manager described how the LHCC decided to use some of its budget to pay for clerical support to enable the CMHT to develop shared case notes. This decision was driven by the desire for better communication from the CMHT to primary care: 'We said ''If we can sort out your case notes, you can sort out communication with us further down the road''.' Within this overall picture there were differences in the types of relationship between primary and secondary care: The research identified a number of ways in which PCTs were developing processes that lead to 'seamless' care across the primaryysecondary care interface for people with mental health problems, e.g. care pathways, referral and discharge protocols. In one case study area, the LHCC had funded a primary care register of people with enduring mental illness, maintained at GP level, not centrally, which was reported useful in generating information on needs and on users views on and experiences of services.
However, it was evident that some of the relationships between primary care and secondary mental health care remained fragile. This was attributed to poor communication and a lack of clarity about the role primary care might play in the continuing care of people with an enduring mental illness. There was also a common concern among secondary care respondents that the involvement of GPs and the primary health care team in mental health would pose challenges to the resources and skill base of secondary care services and lead to services such as Community Psychiatric Nursing being 'controlled' or 'taken over' by primary care. This was a strongly held recurrent view that that would seem to illustrate the perceived differences in the mental health priorities of primary and secondary care. It also points to a lack of mutual confidence, that sufficient common ground could be established without either service having to cede control or relinquish resources.

Partners in management
There were relatively few examples of primary and secondary care partnerships in this area of functional activity. Those identified included: Operational management In Mid-and East Lothian, the LHCC managed all community based mental health services on an integrated basis.
Training In Ayr the LHCCs were regarded as an important means of identifying and addressing mental health training and support needs of primary care.

Liaison and communication
-Lothian PCT reported that the delivery of seamless services and a consistent approach was furthered by the cultivation of close working links between the consultant psychiatrist and primary care colleagues, involving regular meetings and the development of protocols. This was also supported by joint agreements on specialist services.
-Some areas referred to the practice of identifying a named member of the CMHT to liaise with the primary health care team.

Strategic planning of mental health services
The research pointed to widening opportunities for primary care to be involved in mental health planning, through LHCCs and through representation on strategic bodies. In nearly all instances, LHCCs were represented at strategic level within the Trust, with opportunity to participate in and influence service planning as one of a number of key participants. Examples of involvement of primary care and secondary care in planning and development included: -In Highland, the LHCC manager and the local CMHT manager take a joint lead in local service development and liaise with the senior management group that takes the strategic lead Highlandwide. In Forth Valley, the secondary care services are structured within a specialist co-operative to create opportunities for parity in dialogue with primary care through the LHCCs, with both being brought together in Local Implementation Groups.
-In Ayrshire and Arran, primary and secondary care clinicians have been involved in local service planning and needs assessment.
One in three respondents to the survey noted that although LHCC influence was increasing, it was not yet possible to point to specific impacts on policy or strategy. A number of factors may be at work here. LHCCs have been developing gradually and may not yet be in a position to grasp opportunities for influence. This themethat time was essential to allow LHCCs to mature and develop capacity and confidence in their relations with other parts of the Trust and beyondrecurred throughout the research.
There was also evidence that LHCCs had variable levels of active involvement in mental health developments. A minority of LHCCs was reported to have a mental health subgroup, for example. This appeared to vary within Trusts as well as across Trusts. In areas where there were firm organisational structures in place on which to build relationships between primary and secondary care services Trust wide and locally, there appeared to be considerable potential for twoway influence. For example, in some areas of Greater Glasgow, the CMHT lead sat on the LHCC executive. In Tayside all LHCCs Boards include a secondary care representative.
There is a distinction between engagement with and involvement of primary care at local level in shaping local mental health priorities and the influence of primary care at Trust level over the wider mental health agenda. Highland and Glasgow exemplified two different approaches to this. In Highland, the PCT was aiming for a tiered approach, in which the majority of care services would be delivered locally with access to specialist support as required. In Greater Glasgow, there had been a deliberate policy decision to separate the primary and secondary mental health agendas. Other PCTS, such as Lomond and Argyll, were taking steps to establish a joint commissioning board for mental health services, in which LHCCs would be one of several partners.
The view was frequently expressed by respondents in different positions in PCTs (including people with strategic responsibilities and others with operational positions, those working in secondary and those in primary care) that the mental health agendas in primary care and secondary differed substantively. As noted, Greater Glasgow had followed this course to the stage of producing a separate primary care mental health strategy and was in the process of defining and developing models of primary mental health care services to complement the network of community mental health teams providing secondary care. These primary care models were to address the needs of those with mild to moderate mental illness and to include health and social care, counsellors and clinical psychology inputs. The Trust recognised the challenge involved in ensuring effective links and co-ordination with secondary care mental health services. The reasons cited for this course of development were pragmatic: to ensure that the relatively well developed secondary care services were not hampered by the variable pace of LHCC development and to allow LHCCs to focus initially on primary care development without being swamped by an enormous mental health agenda.

Factors that facilitate partnerships between primary and secondary care
A series of factors emerged from the research that appear to be significant in influencing the pace of integrating primary care and secondary care mental health services. They can be summarised in the following terms: Developing trust and mutual understanding.

-
The maturity of community based secondary mental health services and the confidence that developments in primary care mental health can be taken forward without prejudice to existing secondary services for people with severe mental illness.
-Fears persist in many areas among secondary care about being 'taken over' by primary care and about the risk that the focus on the needs of those with severe and enduring mental illness will be lost.

Recognising a common purpose locally.
-Finding common ground, to go beyond perceptions that primary and secondary care focus on different populations and have different priorities.
-Jointly agreeing policy objectives and establishing good inter-professional relationships: in our study these were considered to be more powerful influences than national policy on the interface between primary and secondary care.
Incentives and engagement.

-
The difficulties of engaging GPs as independent (self-employed) contractors in joint working initiatives when they have to bear the costs of involvement.

-
The perceived lack of centrally driven, national initiatives to stimulate GP involvement in mental health.
This article is published in a peer reviewed section of the International Journal of Integrated Care -The extent to which LHCCs are used by PCTs as a means of engaging primary care and secondary mental health services in dialogue. In one Trust, it was reported that GPs preferred to link with CMHTs on a practice-by-practice basis only and argued that the LHCC structures did not add value to this relationship.
-An over emphasis on joint planning and strategy and a lack of attention to service change and development means that partnership working can fail to deliver any immediate benefits for those 'on the ground' and thus undermine the will to integrate services.

Integration of functional activities
Whilst previous sections have explored horizontal and vertical integration by looking at functional activities, the study examined the extent to which the functional activities of Trusts (direct care, management processes and strategic planning) were themselves vertically and horizontally integrated. This provided a means to begin to explore the inter-relationship between different organisational functions, the direction of influence and the notion of traction: whether progress in one functional area might produce forward movement in other functional activities.
The connection and interaction between the different functions was described in a number of ways. Several case study respondents noted that strategy statements were useful in providing the justification for action for those who were already eager to move in that direction and in creating expectations that 'dragged' the more reluctant or agnostic along. An LHCC respondent stated that policy from the centre helped to 'jump start' local initiatives that could then make use of incentives such as development monies. Another described how strategies and policies gave permission to people who had been struggling on the ground to legitimate partnership working. Elsewhere a secondary care manager noted that when plans and strategies were not congruent, this was experienced as enormously disruptive.
The argument for integrating functional activities was cogently put by several case study respondents. A secondary services manager in a case study area, who had responsibility for an integrated health and social work service observed that: 'To offer a truly integrated service it has got to be joined at all levels and that includes the top of service fusing budgets together, fusing the strategic and philosophical.' This respondent went on to note that the appointment of two members of staff senior to her, who had a commitment to integrated working had made it possible to move forward in a number of areas where progress had previously been blocked.
The evidence gathered in the study described how PCTs, along with health boards and their local authority social work colleagues, were collaborating in planning services. Equally, there were repeated examples of local partnership working in the delivery of services to people with severe mental illness. There was less direct evidence, however, of PCTs ensuring that their internal organisational processes and functions were geared up to support partnership working in the delivery of integrated care as illustrated by the following comments from respondents in two different case studies: 'The way it works is strategy comes out and says all the right things. Everyone involved signs up to it, then someone somewhere within the organisation is not acting consistently with what is trying to be achieved.' 'Partners are all signed up to integrated working. The problem is that we do not have the structures and processes to allow us to deliver it, except at operational level' Integrated structures offered one means of achieving closer integration of functions according to one case study LHCC respondent: 'If you want to truly integrate primary care and secondary care in the community then the barriers and false boundaries that appear in management structures hinder that. If people were managed by the same management structure they would see themselves much more as a team... Language and structures can unfortunately sometimes influence behaviour and cultures and they don't let go.' However, structural change to align partners and joint processes to encourage integrated working were necessary but not sufficient, without concomitant attention to the development processes needed to achieve changes in attitudes: 'We need structural change and changes of attitudes' Managers had to be enabled to take on developmental roles to introduce new strategic ideas and to enable stakeholders to assimilate and apply these. A case study manager of an integrated secondary care service described his role as facilitator, to enable the management group of senior operational staff to reach consensus, sometime referee and a conduit for angst. The role also required that this person acted as the interface with both parent agenciesthe PCT and the This article is published in a peer reviewed section of the International Journal of Integrated Care Trusts are working to integrated working evident on redefine the centrespecific projects or issues periphery relationship with the move to more devolved working local authority social work department. However, the ragged nature of the joins at this level were illustrated by his observation that both agencies regarded him as 'their own' employee rather than as a shared resource.

Discussion and conclusions
The study was able to paint a broad-brush picture of the structures and activities Trusts have developed to deliver integrated care. This concluding section aims to provide an overview, summarise the principal themes that emerge and suggest areas that might repay more in-depth development subsequently.
A theme that runs consistently through much of this study is that structures are only one of a number of factors that influence the development and maintenance of partnerships and promote capacity for integrated care. One of the key messages that respondents repeated was that structures can help by creating opportunities and a context that is conducive to partnership working, but are not in themselves sufficient. Where partnerships are problematic, structural change was not regarded as a solution. On the other hand, where the basic necessities of trust and common purpose were in place organisational structures could enable dialogue and engagement.
Primary Care Trusts placed great weight on partnerships as a means to integrate mental health care. However, they faced challenges both in balancing a range of organisational objectives and balancing the imperatives of different partnerships: internally between primary and secondary care and externally with social work and increasingly with the voluntary sector. The typology of relationships set out above ( Figure 1) provides a framework, on to which the key partnerships described in the preceding section can be mapped, and the results are summarised in Table 2 (horizontal integration) and Table 3. (vertical integration). A number of themes become apparent and are discussed below.

Emergent themes Priming partnerships
Our study has identified a number of features that appear to be conducive to the development and maintenance of partnerships in delivering integrated care for this client group.

Preconditions for partnership
-Trust and commitment are essential ingredients in all aspects of partnership working.
This article is published in a peer reviewed section of the International Journal of Integrated Care there is a foundation of pre-existing local relationships to build on. Often where significant progress in partnership working has occurred, there has tended to be a confluence between external influences (e.g. policy imperatives and incentives) and a local pre-existing capacity and commitment. Relationships between secondary care and social work have taken years to reach their current stage of maturity. LHCCs are still in their relative infancy and their place in the system of care and support for people with mental health will take time to emerge.
-Continuity of personnel and of relationships, recognising the importance of interpersonal dimensions of partnership. The tendency in the health service for personnel to move on rapidly from one post to the next poses particular challenges here.
-Attitudes and values are the life-blood of partnership. Attention to structures should not overshadow the need to create a culture that values and rewards partnership.

Organisational development
The study indicated that the achievement of effective partnership working, in terms of horizontal and vertical integration, is hampered by the absence of a whole systems approach, which would require that PCTs consider how all aspects of the organisation's functions (human resources, finance, estates management) could better support the objective of delivering integrated care.
The shift from management to enabling leadership remains a key challenge for the future delivery of integrated care. There is a need for further work to explore how management roles can and should be transformed to provide the support and facilitation that partnership working requires. There is an urgent need in this respect to direct more attention to the professional development agenda for the middle to senior tiers of managers who increasingly have responsibility for multi-agency and multi-disciplinary services.

Organisational processes that underpin partnership
There was evidence that PCTs and local authorities were actively seeking to move partnership working from projects and initiatives, to ensure it is embedded in mainstream working. The obstacles that lie in the path of this journey are familiar and are currently being considered at national as well as local level w16x.
It was reported in the study that aspirations to develop pooled budgets are often impeded by difficulties with information and financial systems and the disaggregation of discrete budgets. However, a counter opinion was expressed by several respondents who maintained that the care outcomes desired did not require jointypooled budgets, on the grounds that integrated care does not necessarily require service integration.
Joint information systems continue to pose considerable difficulties. Trusts and their social work partners were struggling to address this and were clear that it was an important element in supporting integrated working.

Incentives
Promoting partnership working involves creating appropriate incentives and removing disincentives that influence both horizontal and vertical integration and that have a bearing on different aspects of functional activity. Incentives operate at several levels, to shape the behaviour of individuals, professions, services and organisations. A key challenge for Trusts is to nurture the motivation and commitment that brought people into health care in the first place and ensure these qualities are not eroded.
Incentives also have to be linked to what motivates people to work in partnership. The motivators identified in this study include: -Sharing a common purpose to make a difference to people's lives -The added value that can follow from addressing complex issues by working in partnership -The learning and enrichment that comes from exposure to other perspectives -Economic incentives, to reward those who take on additional roles and responsibilities -Compliance with requirements, such as performance management or clinical standards

Differentiation and diversity
The study brought to light a high degree of diversity in the structures and processes Trusts had developed to deliver integrated care. Diversity was in part a consequence of geography and historical patterns of development. It was also noticeable that there were considerable differences within Trusts, and divergence was likely to be intensified by moves to devolve to LHCC level. These trends create challenges in ensuring that responsiveness to local needs can be held in balance with equity. Diversity can lead to enrichment or fragmentation and this begs questions about the role of the PCT in facilitating learning and the exchange of experiences within a common framework of shared aims and values.

Relationship between different dimensions of integration
In this brief exploratory study it was not possible to determine whether there was a direct correlation between the maturation of partnership working at different functional levels, within PCTs and between Trusts and social work. However, there were indications that robust partnership working at one level does not appear automatically to 'cascade down' or 'trickle up'. This reinforces the case for organisational development to facilitate and support change in different aspects of the organisation's activities. There were also indications that strong partnerships in care delivery may at times create sufficient traction to generate movement in other parts of the partner organisations, leading to closer alignment. The study has identified that closer integration of the management functions of partner organisations could make a significant impact on capacity to deliver integrated care.

Models of care that promote integration
Attention has already been drawn to the suggestion that a 'policy gap' that has opened out as a result of the impact of UK policy relating to the care of people with mental health problems w3x and a range of service models to fill this void. The PCT study reported here  One case study area was developing a primary care based mental health services comprising a multi-disciplinary team of social work, CPN, clinical psychology, counsellors and a support worker. This team will offer short-term interventions for up to 12 weeks for people with acute rather than enduring mental health needs. The links with the local CMHT were still being discussed to clarify boundaries, cross referral and transfer arrangements and to ensure effective communication between the two services.
In another case study, the social work department had contributed to the funding of a pilot primary care mental health project. This will involve a CPN post attached to three general practices, whose role will be to support practices, provide education and training and to work with the primary care teams to enhance their procedures and referral processes to secondary care. The post is envisaged as a point of access with the CMHTs and a conduit between primary and secondary care. The CPN will have a caseload but will also train and support members of the primary care to work with patients with mental health.
In a third case study, this process of development was inverted in that a primary care based CPN service predated the formation of a multi-agency CMHT. There were tensions and diverging views among practitioners about proposals to align more closely the CPN service that focused on people with short term more acute needs with the CMHT that worked with people with more serious and long-term problems. The LHCC was keen that the two services should form an integrated team.
In this LHCC, there was reference also to a new type of service that was perceived as having potential relevance for other aspects of mental health work. The appointment of a systemic family therapist was heralded by the LHCC respondent as an opportunity to get away from 'the tyranny of the referral system where you end up playing pass the parcel'. This post was viewed as a means of changing the nature of the relationship between professionals from one predicated on referral and transfer of responsibility to one founded on consultation and shared responsibility, in which the patient was included. What appears to emerge from the study is a lack of clarity about how services designed for people with severe mental illness articulate with those designed for people with mild or moderate mental health needs, making it difficult to promote a whole systems approach. As the implementation of mental health policy gives renewed emphasis to the full spectrum of mental health needs, including the needs of people with severe mental illness, but not overlooking other groups, it seems apposite to re-consider this issue. There may be important differences in perspective within Trusts about the types of services that might facilitate the delivery of integrated care across primary and secondary care for the full range of mental health needs.
It was striking that there was little mention of specific services that had been developed jointly between primary and secondary care for people with severe mental illness, with the exception of out of hours arrangements. Generally developments involving primary and secondary care entailed clarifying boundaries to demarcate roles and responsibilities. There appears to be a lack of clarity about the role of primary care in developing and delivering care for people with serious mental illness. In some areas, it was assumed that responsibility for this group should lie exclusively with secondary services. In other (generally rural) areas, primary care was seen as one of the care partners that shared this responsibility, although the implications of this had often still to be fully developed.
The relationship between primary and secondary care has tended to be uneasy, in part because of issues relating to workload and priorities. To date, LHCCs vary in the extent to which they have identified mental health as a priority interest and in the extent to which they have as yet developed the organisational and management experience required to take an active part in service development. Juxtaposing the examples identified in the PCT study with those described in the literature suggests that models designed to operate at the primaryysecondary care interface may have a number of distinctive purposes: -Facilitativeto enable primary care patients with mental health needs to gain access to a wider range of services and resources -Supportiveto enable primary care to make best use of skills and experience in providing mental health care and treatment -Supplementaryto extend and enhance the capacity of primary care to address mental health needs -Systemicto support and enable different parts of the care system to work effectively individually and collectively, with particular attention to communication, clarifying roles and facilitating partnerships where appropriate Data from the study corroborate other evidence that a bifurcation is developing in mental health services, with primary and secondary care, respectively, leading two separate services, for those with acute or mild to moderate problems and those with severe and enduring problems. This may suggest the need to explore in this context the relevance of intermediate services that could ensure effective linkage, promote continuity of care and appropriate throughput, following, but not necessarily replicating, the models developed for other care groups, particularly older people. Current work to promote the development of psychological interventions may also be of value in supporting vertical integration and may provide models that help to accelerate progress in this area (see A Framework for Mental Health Services in Scotland Services Offering Psychological Interventions; Psychological Interventions Pilot Implementation Project: www.show. scot.nhs.uk).
The research has indicated that focusing on the primary careysecondary care interface should not hinder the potential for partnerships between primary care and other sectorssocial work and the non-statutory sectorwho have much to contribute in promoting 'an ordinary life' for people with mental health needs.

The forward agenda
The study poses challenges for policy makers and national bodies whose role is to support the imple-mentation of policy, in creating the conditions and incentives required to achieve the level of partnership working necessary for integrated care. It raises important questions for further research and development in relation to the primaryysecondary care interface to extend the range of models of care available to meet mental health needs.
The study's conclusions have implications for the way in which PCTs support integration within their organisations, across different functional activities and within their areas, in partnership with local authorities and the voluntary sector. The study has drawn attention to the challenges of organisational development and capacity building that would repay further more detailed research and development, to support new ways of planning, managing and delivering integrated mental health care.