Implementing evidence-based changes in healthcare Edited by David Evans and Andrew Haines Radcliffe Medical Press 320 pages £27. 50 paperback

In the aftermath of the NHS plan, the priority is securing effective service redesign and changes in clinical practice. As the plan acknowledges, model services exist in some parts of the NHS and the challenge is to 'universalise the best'. This book is therefore timely and essential reading. For students of decision-making and implementation its key messages are not surprising. However, although clinicians have come rather late to appreciate that a rational, linear model of change does not explain how change actually occurs, it is better than that they should not have done so at all.

The book's focus is on turning research evidence into practice. It documents 11 research and development implementation projects in North Thames region between 1996 and 1998. The descriptions of the projects are preceded by an opening chapter by the external evaluators which offers their overall impressions of the projects. The evaluators provide a short commentary at the end of each project.

The authors were instructed to be honest and record failures as well as successes, and for the most part they have succeeded. Each chapter ends with two sections headed 'lessons learned' and 'things we would do differently if we were starting tomorrow'.

The case histories described amply reaffirm the barriers to change familiar in the literature on implementation and the need for multifaceted approaches to overcome them.

Some projects give rise to contradictory lessons. For example, several stress the need for ownership of the process, including protocols. But one suggests it is not ownership that is critical but professionals being satisfied that the protocol is well-designed, easy to use and helps their practice.

Such differences demonstrate the complexity of securing change and the naivety of those who believe that if something works well in one place it is a fairly straightforward matter to transfer it elsewhere using the same change technologies. The accounts of these projects challenge that view.

Achieving change is messy, incremental, time-consuming, expensive, frustrating, and invariably incomplete. It is also context-specific, and while there may be a repertoire of techniques and tricks to draw upon, their precise configuration and weighting is likely to vary according to the context.

There were lessons common to the projects, and they include finding organisations and individuals that are ready, picking a good topic, picking the right project worker and 'ringfencing' time, and locating the project within the sphere you want to influence. Six of the eight projects in primary care found achieving change much harder due, in part, to the heterogeneity evident in primary care organisations.

The book should be required reading for those engaged in modernising the NHS, including those at the centre intent on micro-managing the change agenda.

They will discover that the linear model of change did not work for any of the projects; that, in the words of the external evaluators, without a deeper examination of processes 'it is difficult to determine exactly why certain project teams are better able to reach every project's goal of becoming a part of routine practice', and that securing commitment to change is best done on an informal, individual basis that is inevitably time-consuming.

There are no quick fixes and no easy solutions to be achieved by sending in hit squads or inspectors.

David Hunter Professor of health policy and management, Durham University.