Return to the windmill - behavioural modelling and the future

The lack of a 'big picture' of where reforms will take us means investment and strategic planning are severely hampered. Alasdair Liddell and Laurie McMahon describe a behavioural modelling approach that can help

The NHS reform agenda is huge - payment by results, patient choice, practice-based commissioning, provider plurality - and brimming with acronyms - ISTCs, DTCs, APMS. The impact and intent of individual policies and initiatives are often well understood, but their combined effect? Nobody knows.

Even if policy makers thought they knew, had mapped it all out on a giant whiteboard in Richmond House, they would be wrong. It simply isn't possible to compute the sum of all the individual behaviours of all the different interests as they respond to the changed incentives and opportunities created by the reform programme. And that's before we think of overlaying all of this onto the existing health system, with its pre-existing powerful economic, technological and demographic forces for change.

But does this really matter? The health secretary made a point of claiming &Quot;no end game&Quot; at a recent King's Fund conference. The reforms were intended to stimulate new ways of working, new efficiencies, and to unlock the creative energies of all the players in the new NHS market to achieve NHS objectives.

The problem is that without a big picture of where the reforms might take us, the future remains so uncertain that everyone's ability to plan investments and set business strategy for the medium to longer term is compromised.

There is a real danger that the opportunities presented by the reforms will not be fully exploited. And there is an equal danger of unintended adverse consequences - bear traps that will provoke a damping down of all that creative energy.

Conventional methods of predicting the future will not help. Waiting to see what happens then muddling through is not an acceptable strategy either. What is needed is a way of modelling how the system will actually respond. This requires a 'soft' qualitative approach that uses a behavioural simulation to draw on the knowledge and instincts of real players in the system to understand the future.

That's why the King's Fund, in conjunction with Loop2 and with support from Monitor and Nuffield Hospitals, is planning just such a simulation. But it's not the first time this approach has been used, just as it's not the first time an NHS market had been created.

1990 and all that

The first time behavioural simulation methodology was used to understand the future of the health system was in East Anglia in 1990, when the conditions were somewhat similar to those existing now.

The Government of the day was introducing a major reform package of new incentives and new organisational structures to support the new internal market arrangements. Then, as now, it was impossible to map the outcome - positive or negative - of the new arrangements using conventional predictive methods.

East Anglian Regional Health Authority commissioned the Office for Public Management to design and run a simulation event over three days in a Norfolk hotel. The results have passed into the folklore of the NHS, as has the name the initiative was given: the Rubber Windmill.

The objective was to explore the dynamics of the market in three separate years, and in particular how the behaviour and interactions of actual participants would have an impact on the ultimate aims in securing better value for money in providing real benefits for patients and populations.

There were about 40 participants in all - about half from East Anglia, the others from the Department of Health, and other regions and districts. These included Peter Griffiths, then deputy chief executive of the NHS, Ron Kerr, Chris Spry, Graham Winyard, Alan Burns, Stephen Thornton, Pat Troop and Niall Dickson, then social affairs correspondent of the BBC.

The participants were divided into groups representing their real-life positions in the system. There was no role play - the GPs, managers, hospital clinicians and authority members were playing 'for real' in the simulated setting.

While the overall structure of the simulation had been designed with care, there were no rules for the play, and no rules to 'win by'. Participants were invited to identify their objectives for the year and set out to achieve them, recording their interactions and agreements with other players for later analysis and discussion.

Participants acted and behaved in ways which reflected their own professional and institutional interests. Partway through the third year of play the system 'collapsed' under the weight of conflicting financial and competitive pressures.

This was not - as the dramatic press reports of the time suggested - a surefire indication that the market would fail. It had simply been 'tested to destruction' for the purposes of the simulation. But the messages that emerged when the participants reflected on what had happened were compelling. They said that if the reforms were going to work there needed to be:

• the development of the 'purchasing' role to ensure that it delivered health and quality outcomes as well as financial performance
• a clear corporate sense of direction for the service to avoid fragmentation
• financial stability, to cope with the complexity of the contracting process, and the greater vulnerability of participants to sudden changes in cost and revenue
• better information and new systems of accountability to reflect the centrality of GPs in the new system.

Recognise any of these?

Windmill 2007

Work on the 2007 simulation begins with a 'system dynamics' event on 4 December bringing together around 35 key thinkers and commentators to explore how the roles and relationships between the major components of reform might evolve - and also to identify the drivers and tensions that might emerge. The intention is that this should provide a framework to ensure that during the simulation event (to be run across two days in February) we focus on the key issues and concerns.

A few of the questions to emerge from the preparation for this event are:

• Will there be a single regulator for quality and financial performance?
• Will trusts be allowed to go bankrupt as a result of shifts in patient flows?
• Will PCTs divest themselves of provider services?
• Will the Office of Fair Trading and the EU have an increasing influence over the health care system?
• Will foundation trusts become providers of primary and community care?
• Will ISTCs continue in place after five years?
• How will the system deal with the conflict of interest between practices as commissioners and providers?
• Will tariffs be replaced by price competition?

Of course, we're not claiming Windmill 2007 will give us detailed answers to all of these questions - plus the many more not listed here. But we are convinced it will provide real insight to the key consequences and, most importantly, what we need to do now to give the reforms the best chance of success.

Alasdair Liddell was chief executive of the East Anglian RHA and commissioner of the 1990 Rubber Windmill. He is now leading the development of Windmill 2007 for the King's Fund, together with Laurie McMahon, now director of Loop2 and designer of the 1990 Rubber Windmill. The report of Windmill 2007 will be published by the King's Fund in spring 2007.