Published: 21/11/2002, Volume II2, No. 5832 Page 30 31

Measuring Up Improving health system performance in OECD countries Publisher: OECD.

ISBN: 9264196765 380 pages.£46.

This is a useful - if dense - guide to the international terrain as the golden age of data tourism draws to a close and the performance managers establish their logging camps.We used to study the mysterious peaks and troughs of performance; now we are supposed to do something about altering them.

This volume gives the first accessible opportunity to compare how the US, UK, Canada, Sweden and the Netherlands are doing.The other Organisation for Economic Cooperation and Development members (especially Australia) are curiously absent, but there are some rather clear results.

Figures that create local incentive through a disease focus are likely to have much more impact than broadbrush national indicators. In Ohio, the Cleveland health quality chain established an agreed set of measures with funders and providers that led to a 13 per cent reduction in mortality for the eight pathologies covered.

A good review of the US by French authors finds that 'the establishment of comparative databases on hospital performance, if they are constructed with rigorous methodological principles and based on co-operation with participating hospitals, can provide highly useful knowledge for improving performance.'

And Coulter and Cleary show how it is possible to develop measures to improve the care process from the patient's point of view.

The UK experience is well charted by Clive Smee, albeit with a focus on government rather than bottom-up professional efforts such as the Confidential Enquiry into Perioperative Deaths by the Royal College of Surgeons.

Smee makes it all too clear that the UK has begun to move right away from all the lessons of international best practice.Our system is now based on broad-brush indicators, collected and disseminated in a way that maximises the threat effect and reduces local cooperation.

The star-rating system has probably not led to visible improvement for a single patient in the whole of the UK. It may have had negative effects in diverting management and professional time from actual improvements in service.

The Commission for Health Improvement is taking over the system and it should make an urgent study of the international evidence.This suggests the outcome should be a move towards basing indicators on disease management and getting much more local and professional co-operation.

CHI can look at Canada as well as Ohio and Sweden.Canada has shown how it is possible to use province-wide data to improve surgical performance and to answer questions about whether specialisation gets better results for patients.

If we want to move on to influence performance, we have to produce data that really measures it.Outcome and patient-related indicators need much more accurate definition.

This OECD volume will not only be a godsend for professional exams - it is essential background for CHI as it sets out how to develop credible indicators and get away from the dismal heritage of the star-ratings system.