The ‘collaborative’ method, developed by the US Institute of Health Improvement, has succeeded in motivating staff and responding to patients' concerns about acute in-patient psychiatric care across 37 trusts in the Northern, Yorkshire and Trent regions. The method puts into practice the new values underpinning government policies on ‘modernising’ and ‘shifting the balance of power’ in the NHS, i.e. seeing things through the eyes of patients, empowerment of frontline staff, learning networks and focus on measured outcomes. The evaluation suggests that organisational and management culture crucially determined the level of achievement of the trusts taking part in the collaborative. Evaluations of other collaboratives have raised doubts about the sustainability of the improvements achieved. It is argued that refining the collaborative method is less important than incorporating its principles into the existing management and organisational cultures of NHS trusts, and the leadership styles of chief executives and clinical directors.

The United States Institute of Health Improvement has developed a method for continuous incremental improvement of clinical services, which is now sweeping the NHS. According to the Department of Health (2002), during the past 2 years, 16 million patients have been affected in hundreds of trusts by the work of thousands of improvement teams. The service areas involved include cancer, orthopaedics, primary care, coronary heart disease, older people and mental health. ‘Collaboratives’ are seen by the Department of Health to be a vital expression of NHS ‘modernisation’ policy (Department of Health, 2000) and the more recent policy on ‘shifting the balance of power’ (Department of Health, 2001). The method designates responsibility to clinical teams to redesign local services around the needs and convenience of their patients, and is a vehicle for spreading good practice from centre to centre. However, experience in the US and UK suggests that the service improvements achieved are often modest, and only a small proportion are sustained a year or more after completion of the collaborative programme (Reference KiloKilo, 1998; Reference Robert, Hardcare and LocockRobert et al, 2002).

Is this yet another fashion that will sweep through the NHS only to disappear without a trace? This paper argues for optimism if the nature of the challenge is shifted from improving the method and implementation of collaboratives to changing the managerial and organisational cultures in which they are applied.

The mental health collaborative

National and regional surveys carried out by the Sainsbury Centre for Mental Health (1998) and the Northern Centre for Mental Health (Reference KennedyKennedy, 2001) found that patients, carers, professionals and managers alike were very concerned about poor standards of care in acute psychiatric in-patient wards. The Northern Centre for Mental Health in partnership with the Trent and Northern & Yorkshire regional offices, and with advice on the collaborative method from the Leicester Centre for Best Practice, organised a ‘reference group’ day meeting of about 50 people. Those attending were influential representatives of all the professions involved with acute in-patient care and patients with recent experience of these services. After listening to patients describe their journeys through the process of admission, stay in the ward and discharge, there was unanimity of view among all those taking part that the patients' experiences could be greatly improved by redesigning the service through incremental changes, to match a couple of dozen clear and achievable standards.

There were two remarkable observations about this initiating event. First, inter-professional differences in perceptions that are notorious for inhibiting progress in the mental health field began to evaporate when exposed to patients' real and recent experiences. Second, representatives of a mental health workforce currently suffering from widespread problems of recruitment, burnout and premature retirements became enthusiastic about the potential of the collaborative approach and optimistic that things could be significantly improved. Thirty-seven multi-disciplinary teams with local users were recruited from all trusts with mental health services in the two regions. Local project managers were appointed. All were invited to the first collaborative workshop to learn about the collaborative method, which focuses on ideas for change being rapidly put to the test on a small scale with the results informing further work. The method assumes that such small changes aggregate to larger improvements through successive ‘Plan—Do—Study—Act’ (PDSA) cycles. Local teams could choose the priority attached to particular standards, and set their own targets and the times it would take to achieve them. The commitment was made to share information on progress and attend four workshops over the next 12 months to learn from each other.

Commitment rose from tentative interest to some reporting the events as being the most energised conferences on mental health they had ever attended. Standardised measurements of progress were reported by all trusts towards achieving their targeted improvements in care. Differences in achievement between centres subsequently became a source of pride or a stimulus for striving harder. However, there was concern that these data might be released to those involved in performance management within trusts and health authorities. It became apparent that whereas the performance management system inevitably encourages ‘massaging’ the data to match requirements of higher levels of the NHS, reporting within the collaborative was more truthful and therefore, quite often discrepant.

Factors helping or hindering achievement

Measurement to improve was so central to the change process that it did not need a randomised controlled trial to confirm that widespread and worthwhile changes had occurred over time. The particular improvements patients had asked for were validated by them afterwards. All teams showed progressive improvement towards targets over a broad range of standards, demonstrating significant success for the project as a whole (Northern Centre for Mental Health, 2002). There were, however, large variations in levels of achievement across trusts.

Robert et al (Reference Robert, Hardcare and Locock2002), of the Health Services Management Centre, University of Birmingham, carried out a qualitative study to identify the characteristics and differences in organisational contexts and processes that might explain differences in achievements between successful and less successful project teams. From the 37 trusts participating in the collaborative, six sites were selected for detailed study using semi-structured interviews, direct observation of team meetings and a formal ‘readiness for change’ questionnaire (adapted from ‘The Organisational Change Manager’ developed by Professor David Gustafson, University of Toronto and Madison Healthcare Improvement Ltd., Wisconsin). The interviews and observational data were analysed and triangulated with the responses to the questionnaire, and the self-reported quantitative data on achievements made on each of the six sites. The study suggested five ‘contextual’ factors that affected the rate and degree of progress by the local project teams. These were: the extent of senior management support, leadership style, organisational adaptability, level of empowerment and attitudes to risk and innovation.

Senior management support seemed to make a large difference to team enthusiasm and confidence. Teams that felt well-supported were more resilient in overcoming practical obstacles and staff resistances. Senior leadership style that emphasised people, process, support and participation facilitated progress more effectively than rigid, controlling and directive approaches. In addition, those organisations that were ready to accept and adapt to different ways of doing things allowed more progress than those that required project teams to just fit in with existing conditions.

The level of empowerment of teams emerged as a key issue. Teams that felt both empowered by the collaborative approach and free to get on with making their own decisions made more progress. This was also the case for organisational attitudes to innovation and risk. Those trusts in which it was explicitly allowed to take calculated risks, and which encouraged innovation, saw a greater level of service improvement.

The evaluation emphasised just how important organisational culture or ‘context’ is in creating the right circumstances for learning, improvement and development. The more of the five key factors that are present in an organisation, the more likely it is to benefit from a collaborative approach.

Implications for chief executives, clinical directors, and organisations

Lessons learned from this mental health collaborative are too important to confine to the planning of future collaboratives. It is suggested that there are profound implications for how chief executives, medical directors, clinical directors and all service managers operate, and the kind of culture that patients and staff need them to promote.

The way the collaborative involved everyone in looking at the service through the eyes of patients, and focused the collective enthusiasm of professionals on ‘can do’ incremental cycles of improvement, could be the modus operandi of every clinical directorate and clinical team. Every service manager and clinical director should be armed with the skills and experience of a successful project manager in a collaborative. Chief executives need to understand what Grint (Reference Grint2000) found: the solutions to most organisational problems are already known to the workers, but their formal leaders often prevent them from implementing these solutions.

Much of the malaise within the NHS workforce can be attributed to ‘top down’ control that inhibits innovation and risk-taking, and seeks standardisation and equity that often means, at best, mediocrity. Commissioning of care, and promotion of beacon services, tends to impose other people's solutions on local conditions that are only fully understood by those working in them. Performance management arrangements sometimes tell higher levels of the service what they want to hear, not what they ought to hear. When performance indicators truly reflect the experiences and priorities of patients and front-line clinical staff, they are used to improve performance.

The energy generated in clinical teams taking part in the collaborative to overcome obstacles and get things done has startled chief executives. The evaluation showed that chief executives who listened and responded positively discovered levels of motivation from staff quite unprecedented in their experience. Disinterested chief executives and clinical directors rapidly sapped the energy of their staff and severely limited their achievements.

Our conclusion is that we should be less worried about modest achievements or lack of sustainability of achievements made by collaboratives and much more concerned to help managers and clinical directors to recognise the potential of nurturing a leadership style and culture that harnesses the principles of the collaborative.

Declaration of interest

The mental health collaborative and its evaluation was funded by the NHS Regional Offices of Trent and Northern and Yorkshire with contributions from participating trusts.