A new approach to research and development (R&D) for mental health services is developing in the North-East, Yorkshire and Humber. It derives from experience with service redesign using the ‘collaborative’ approach developed by the US Institute for Health Improvement. Kennedy & Griffiths (Reference Kennedy and Griffiths2003) described such an approach involving 37 mental health trusts, each with a multidisciplinary team, with the aim of improving acute in-patient wards. After studying the patient's journey through care they agreed a set of improvement targets that all teams would work to achieve. Progress towards targets was measured and reported by all teams, who met periodically to compare performance and learn from each other. Remarkable energy to achieve objectives was released among front-line staff involved. Focus was sustained on what most concerned and benefited patients. Good ideas and results quickly spread to all these services affecting thousands of patients.

Research with service development

This acute mental health collaborative was a powerful and practical approach to achieving some of the changes called for in the NHS Plan and the National Service Framework for Mental Health (Department of Health, 1999, 2000). The same emphasis on collaboration, with a more rigorous approach to collection and standardisation of data, is now being applied to promote wide engagement in service-based research. A network of 29 assertive outreach teams in 12 trusts is taking part in multi-site evaluation.

The first step in developing this collaboration was the convening of a large workshop involving over 80 assertive outreach staff and managers at which participants identified and debated a range of questions which might be answered by empirical evidence. The underpinning philosophy was the utilisation-focused approach, described by Patton (Reference Patton1997). Here, stakeholders, working in large and small groups, are encouraged not only to formulate questions but to identify who specifically would be interested in the findings and, most importantly, how the findings could be used to improve services. Only when audiences and uses were established were the research questions agreed.

The task of the researchers was to propose methodologies by which the research questions might be answered. At a follow-up meeting, the appropriateness and feasibility of the methods were debated by the clinical teams, modified and accepted. The methods included structured interviews about team organisation, qualitative enquiry into principles and practice, and a population survey of assertive outreach service users. A key concern was to benchmark services across the region in relation to what previous research had shown to work and to national policy guidance (Department of Health, 2001). Variations according to the urban/rural mix found in this region of England were also of importance to this study.

This programme of research is being managed and coordinated by a partnership of clinical and academic staff experienced in health and social services research. The leading academics are from the Universities of Durham and Hull. Meetings to report back on the findings have had full attendance of front-line staff and managers, and user and carer involvement is growing. Teams are using the results. A Delphi study (Reference Linstone and MunyLinstone & Muny, 1975) has been completed in order to agree what additional clinical measures will be used to obtain more detailed information on outcomes.

We think this approach illustrates how health services research can become dynamic and influence practice across large populations, and help deliver the intended objective of the NHS Priorities and Needs R&D Funding Stream (PNF) - research should make a difference to clinical practice. This is in line with a number of recent papers which have advocated and described practice-based research (Reference Audin, Mellor-Clark and BarkhamAudin et al, 2001; Reference Gilbody, House and SheldonGilbody et al, 2002; Reference SladeSlade, 2002; Reference Proctor, Mitford and PaxtonProctor et al, 2004). This is research that uses real-world patient populations in ordinary service settings, rather than highly selected patient groups in university clinics. Practice-based research can deliver randomised controlled trials (RCTs) and other research methods that are needed at different stages. This approach is most likely to yield results that are immediately and directly incorporated into practice. To tackle some questions about complex interventions there is a need for rigorous qualitative as well as quantitative research (Reference Campbell, Fitzpatrick and HainesCampbell et al, 2000; Medical Research Council, 2000), and in mental health research a need to span both the natural and social sciences (Reference Slade and PriebeSlade & Priebe, 2001).

Organisational challenges

The above are fine ideas perhaps, but making them work in practice is another matter. Practice-based research requires the kinds of service and academic partnerships specified in the guidance of the PNF funding stream, but making such partnerships work is challenging. The challenges include workload, financial pressures, differing priorities between the NHS and universities, and the belief of some NHS staff and organisations that R&D is a frill rather than a worthwhile NHS activity.

Considerable attention needs to be given to supporting and developing partnership working if large and diverse groupings are to collaborate successfully on a continuing basis in order to deliver the research the NHS most needs. Our assertive outreach example has successfully engaged 12 trusts with other service organisations, and 4 universities (Newcastle, Durham, Hull, Bradford) over 4 years to date. It is part of a wider collaboration working on service development and research and evaluation of the new mental health service components: assertive outreach, crisis resolution/home treatment and early intervention. Some features of the programme are that it:

1. • aims to carry out research on priorities and problems identified by practitioners, their patients and their provider organisations;

2. • promotes wide involvement in designing and carrying out the research, and ensures that service personnel understand and are committed to answering the research questions;

3. • is flexible in helping to develop high-quality services and building research capacity;

4. • values and uses diverse research methodologies;

5. • begins by describing services as they actually are, studies and learns from existing service differences, and moves on to evaluating changes generated within services, including comparative evaluation across the partnership.

Progress has been substantial. Reports analysing the characteristics and circumstances of over 800 users of the assertive outreach teams have been produced, with comparisons to the users of community mental health teams. This has enabled clinicians and managers to benchmark their work (Reference Schneider, Brandon and WooffSchneider et al, 2006; Reference Slade, Schneider and BrandonSlade et al, 2006). Detailed descriptions of team functions and structures have been developed, compared, and fed back. Several service development networks have merged into the partnership, adding strength and avoiding duplication. Tools to measure the fidelity of the new services to the policy implementation guidance models (Department of Health, 2001) have been produced and used. The mental health minimum data-set is being used to investigate wider service impacts. At a regional conference in December 2003, it was clear that engagement in the partnership, and receipt of research information produced by the partnership, were highly valued by both clinical and management staff. Following a delay in funding for a researcher in 2004, the project has been re-energised and a follow-up survey of teams and their case-loads is underway.

Helpful tactics

Despite continuing financial and workload pressures, and the substantial task of coordinating and sustaining such a large and diverse group, service and academic partners remain engaged and enthusiastic about the further potential of the partnership. Factors that seem to have helped are:

1. • the pursuit of mutual benefits from projects: people - whether academic, service provider, patient or manager - will not give their time, energy and ideas unless there is a pay-off that each values;

2. • close engagement with and support from the National Institute for Mental Health regional development centre;

3. • establishing programme coordination as a function that all participants recognise explicitly as a main ingredient for success that has to be adequately funded;

4. • establishing the programme coordinator (both personally and on an organisational basis) as a neutral honest broker, whose requirements from each participant organisation must be met to deliver the benefits to all members of the partnership;

5. • all partnership organisations ‘buying in’ to multicentre projects with a (usually small) resource contribution in the form of a local project manager, some of whose time is spent on liaison with the overall programme coordinator and data handling.

6. • the constant endeavour to make it easy for organisations to participate, but having explicit entry criteria that participating organisations must meet;

7. • building trust through openness, even-handedness, and fair acknowledgement of personal contributions.

Where next?

The National Institute for Mental Health regional development centres have a natural interest in helping to connect research and practice. We need to strengthen and extend this partnership with the help of our regional development centre. In particular, we wish to look to the centre for the administrative leadership and central costs of coordinating the partnership. We also hope that through the establishment of a north-east hub of the Mental Health Research Network we will be able to contribute more to the further development of research which is grounded in the real world.

Declaration of interest

None.