Recently government ministers have set out their vision of the future of purchasing. Ineffective treatments will be discarded and purchasing will be based on guidelines or protocols rather than activity. But have the advocates of this approach considered all the issues? This paper examines the challenges of balancing the desire for protocol based uniformity with the needs of individual patients, explores the extent to which existing purchasing structures can support this process, and questions whether such moves will actually lead to reduced costs. In each case it is concluded that oversimplistic analyses are likely to be misleading and that much of the current debate fails to recognise the complexity of health care.

A health service in which decisions are rational and based on evidence has obvious benefits. Some current treatments are ineffective and others that are effective are underused. Much effort is being devoted to developing clinical guidelines and protocols. Amid continuing disillusion with audit some politicians, managers, and health serviceresearchers see the purchasing process as the way forward. There have been several initiatives to increase the effectiveness of health care through the purchasing process, described in recent executive letters,1 2 3 including the “Effective Healthcare” bulletins, the “Outcomes Clearing House,” and the changing focus of the research and development programme. These seek to gather together the best available evidence. But it is less clear how this information should be translated into practice.4

This vision sees all interventions progressively being evaluated by randomised controlled trials and purchasers undertaking comprehensive assessments of need that will indicate which services should be purchased: locally adapted guidelines will be implemented by all providers; purchasers will buy not activity but guidelines or protocols5; change will come about through the “incredibly powerful lever of contracting”6; and Dr Mawhinney's seven steps …