Greater than the sum of our parts? Emerging lessons for UK health and social care

Background and introduction Although most developed countries are currently pursuing greater integration of health and social care, the current evidence base is limited by a number of key weaknesses in the existing literature. Chief amongst these is the tendency to focus on issues of process (‘how well are we working together’?) not on outcomes (does it make any difference to people who use services?). As a result, there is a danger that integration can become an end in itself, rather than a means to an end (of better services and better outcomes). Understanding context, process and outcomes To guard against this danger, this policy paper sets out a number of theoretical and conceptual frameworks to help policy makers, managers and practitioners remain focused on the outcomes that their joint work is meant to achieve. This includes different approaches to being clear about what integration is trying to deliver (outcomes), understanding where we are now (context) and how we get from where we are now to where we want to be (process). Conclusion Rather than assume that integration is automatically a ‘good thing’, the frameworks presented in this paper suggest a more critical approach in which policy makers, managers and practitioners focus in more detail on what they are trying to achieve for the people they serve, viewing integration as a means to an end and never an end in itself.


Background and introduction
In the UK, integration and inter-agency collaboration are key themes in current government policy, both for adults and for children. Despite this policy enthusiasm, our knowledge of what works when it comes to collaboration and integration remains limited by a number of key weaknesses in the existing literature, in current approaches to research/evaluation and in current UK policy and practice [1][2]: • Most studies focus on issues of process (how well are we working together?) rather than on outcomes (does it make any difference for people who use services?).

•
Most research fails to fully involve people who use services and (sometimes) front-line staff.
• Many UK policy makers assume that integration and inter-agency collaboration are inevitably a 'good thing' (that lead to better services and hence to better outcomes).

•
Most research focuses on the potential positives of integration, without necessarily considering some of the negatives. • Most studies only consider a narrow range of quantitative indicators (rather than adopting a broader approach that combines qualitative and quantitative research with consideration of the expertise of front-line practitioners and the lived experience of people who use services). • Most research finds it hard to identify the specific impact of the integration or collaboration (as opposed to the many other services and changes underway at the same time).
• Above all, most integration/collaboration becomes an end in itself rather than a means to an end.
That this is the case is deeply to be regretted, as integration can be extremely disruptive and can have significant negative effects for both staff and service users. Indeed, research suggests that [3][4][5][6]: • Structural change alone rarely achieves its stated objectives. • In addition to stated drivers for integration, there are usually unstated drivers (such as addressing managerial or financial deficits and responding to local or national politics). • The economic benefits of integration are modest at best, and may be out-weighed by unanticipated direct costs and unintended negative consequences (such as a decline in productivity and morale). • Senior management time is often focused on the process of integration, and this can stall positive service development for at least 18 months (if not longer). • The after-effects of integration can continue for many years after the change has taken place.
As a result, it is crucial that policy makers, managers and practitioners are clear about what outcomes they are trying to achieve, are ready for the upheaval integration can cause and are sure that the outcomes at stake are worth it! Against this background, this paper reviews emerging lessons with regards to UK health and social care partnerships. While the UK system is very different from the situation in other countries, some of the underlying frameworks and approaches suggested here may be relevant in other contexts. For those interested in exploring these issues in more detail, there is helpful additional material available in wider literature on organisational development, culture, leadership, networks and systems theory [see, for example, 7-10].

Understanding context, process and outcomes
Against this background, a useful guiding framework is set out in Figure 1, asking front-line services and practitioners to consider: • What they are trying to achieve for local people (outcomes). • How well (or otherwise) current services do this already (context).
• The structures they need to develop as a result (process).
While this is inevitably an over-simplification, thinking of integration and collaboration in these terms can be a helpful way of staying focused on what really matters. Depending on the outcomes concerned, local partners may well decide that they do need some form of integration/collaboration, but this is not automatically assumed, and the form of collaboration required depends very much on where local services are starting and what they want to achieve.
Simple though this may appear, remaining focused on outcomes is extremely difficult in current UK public services, as many policy makers, managers and practitioners tend to find it easier to think in terms of structures and processes rather than outcomes (see Box 1). This seems to be the result of a series of inter-related issues-the historical legacy which current services have inherited, the tendency for current performance management systems to focus more on process than on outcomes, and the fact that many front-line practitioners are often never asked to think about what they are there to achieve on behalf of the people they serve. In addition, the process of determining outcomes may not be simple and can reveal that different stakeholders have quite different perceptions about what the collaboration/integration is there to do. If such an exercise does indeed reveal significant differences, it may well help to explain why integration and collaboration may have been problematic-because different people had very different expectations of the relationship. However, with all these caveats, the fact remains that such an approach can help potential partners to retain a focus on collaboration/integration as a means to an end, rather than as an end in itself.
Against this background, the remainder of this paper builds on the framework in Figure 1, drawing on research, policy and practice in order to explore how to improve the positives associated with inter-agency collaboration/integrated care and how to implement and use the results of research.

Being clear what we are trying to achieve (outcomes)
When seeking to make a clear statement about the outcomes that any given collaboration or integration is designed to achieve, recent policy and practice reveals a number of helpful hints. In particular, anecdotal evidence from recent UK reforms in children's services 1 Personalcommunications,DirectorsofChildren'sServices,West MidlandsHealthandSocialCareLeadersnetwork,UK.
suggests that this can be most powerful when outcomes are focused on potential benefits for people who use services, when they are expressed positively (that is, 'keeping children safe and well' rather than 'preventing abuse'), and when they are simple and easy to communicate to staff, service users and the public alike.
In addition, it is our belief that it is at the stage of articulating desired outcomes that user and carer involvement is most important. All too often, in our experience, services consulting about future provision tend to ask service users for their views on issues of structure or process. While users may have a view on service structures, we believe that their input is crucial from day one, and that they should be centrally involved in discussions about desired outcomes-what do they want their lives to be like and how can services support this? After this, involvement in critiquing the current context is also important (in order to establish the extent to which current services deliver user aspirations)-but involvement in what sort of structure is needed as a result is much more technical and much less relevant to many service users.
As an example of such an approach, Box 2 summarises HSMC's Partnerships Outcomes Evaluation Tool (POET) as a possible way of approaching this issue.
In devising a list of desired outcomes, it is also important to consider the type of outcomes that the collaboration is trying to achieve. Typically, many collaborations have (both stated and unstated) outcomes that focus on benefits to service users (better services), benefits for staff (a richer, more satisfying environment) and benefits for the organisation (better use of scarce resources). Perhaps it is this promise of a potential 'win-win' situation that makes integration and collaboration such tempting concepts. However, in considering which type of outcome is being sought, it is important to be honest and open. The danger with some collaborations is that they are seeking economic benefits for partners, but dress this up in the language of service user benefits, and this quickly leads to cynicism and disengagement (particularly for service users who typically see straight through such claims).

Understanding where we are now (context)
When deciding what is possible in the current context, the literature suggests that it is important to be honest and open about the extent to which different agencies are genuinely committed to the outcomes in question (as well as about any pressures that may prevent them from prioritising these issues as much as they would like). Without this, there is a danger that organisations sign up enthusiastically to joint desired outcomes that they will never be able to prioritise in practice, and that the outcomes at stake are not really important enough to all parties to ensure that the necessary action actually materialises. To guard against these dangers, potential partners can seek to focus collaborative activity on areas that are of importance to both organisations, with successful outcomes much less likely if the resulting collaboration/integration is focusing on something that is important to only one agency and not the other. Where an issue is not strategically important to both parties, then a contractual relationship may be a better way forward. As a simple illustration, Figure 2 provides a brief summary of this approach, and this diagram may be helpful for potential collaborators to use when analysing and developing both current and future relationships.

How do we get from where we are now to where we want to be (process)?
Once potential collaborators understand and are clear with each other about what they want to achieve together and about where they are now, the bit in between-what do we do?-ought to be a lot less controversial than is often the case. However, even where partners have used the approach in Figure 1 above to explore the context and desired outcomes together, it is nearly always issues of structure and process (the 'what do we do?') that generates the most potential tension. When this happens, some of the frameworks and concepts below may help to explore the type of relationship that is needed.

Box 1. The difficulty of focusing on outcomes
When seeking to apply Figure 1 to health and social care practice, one local area struggled to remain focused on outcomes when seeking to reform local older people's services. After detailed reflection, they were certain that the 'outcome' they were trying to seek for older people was an 'integrated management team'! Clearly, an integrated management team is merely one way of trying to achieve broader outcomes for older people-while it could be a helpful way forward, it is hard to know without greater clarity about desired outcomes. However, having an integrated management team should never be an end in itself, and the authors are yet to meet an older person that needs an integrated management team per se!  In a classic article on 'the five laws of integration', Walter Leutz [12] argues that different services may need to work together in different ways depending on what they are trying to achieve. This includes three different levels of integration: • Linkage: appropriate for people with mild, moderate or new needs, linkage involves everyone being clear what services exist and how to access them, so that support is provided by autonomous organisations, but systematically linked. • Co-ordination: with more explicit structures in place, co-ordination involves being aware of points of tension, confusion and discontinuity in the system and devising policies and procedures for addressing these. • Full integration: for people with complex or unpredictable needs, full integration involves the creation of new services and approaches with a single approach and pooled funding.
In a similar approach, Edward Peck [13] and Jon Glasby [14] challenge partners to consider the balance they need to strike between depth and breadth of relationship in order to achieve desired outcomes (see Figure 3 and Box 3). In any given local community, there will be a range of inter-agency relationships in different places

Box 2. The POET approach
The Partnerships Outcomes Evaluation Toolkit (POET) is a resource developed by the Health Services Management Centre in order to assist health and social care partnerships to evaluate their work. This web-based resource recognises the importance of both process (i.e. how well do partners work together?) and outcome (i.e. does the partnership make any difference to those who use services?) As a result, POET takes a two-pronged approach: • Inviting all staff members to complete an online survey which analyses how the partnership 'feels' to them and also surfacing all the underpinning assumptions about what the partnership is aiming to achieve in terms of outcomes for service users • Using the information from the staff survey, a research schedule is designed which checks out with service users and carers whether these are the 'right' outcomes to be aiming for and the degree to which the partnership has been successful in changing these outcomes In this way, POET is both: • Formative-it seeks to evaluate how well partners are working together, helps people to understand and make sense of their current context, and highlights both areas for celebration within the partnership as well as areas where development work is needed • Summative-POET is evaluative in that it requires partnerships to be explicit about desired outcomes and then analyses the degree to which the partnership is successful in achieving these aims (for further information, visit: http://hsmcfs3.bham.ac.uk/questionnaire/) As an example of the POET process, one health care community used this approach to begin evaluating its services for children and young people. However, having started this process they realised that there was insufficient agreement locally about what the partnership had been set up to achieve, and so focused on doing more detailed work to explore desired outcomes. In another area, a disabled children's service used POET to help staff share what they thought the team was there to achieve for children and families using services, before doing further qualitative work to understand how these views differed from those of children and families themselves in this matrix, and potential partners may well find it useful to consider what type of relationship they think they need in order to deliver their joint aspirations. Put another way, this framework encourages local organisations to ask themselves a key question about any given piece of inter-agency working: 'collaboration with whom and for what'?

Conclusion
Although integration and collaboration are currently being pursued in a number of developed countries, it is important to be cautious about some of the claims made for this way of working. While they might be the right way forward for some service user groups, this is not always the case. Rather than assume that integration and collaboration are automatically a 'good thing', UK policy makers, managers and practitioners need to focus in more detail on what they are trying to achieve for the people they serve and on how well (or otherwise) current services do this. Then, and only then, is it time to think about how we change services to get from where we are now to where we want to be. While this might involve some form of integration or collaboration, it would be integration and collaboration as a means to an end-not for their own sake and not as an end in themselves.