Letters/Book reviews be a local collaborative activity. Academic departments of general practice can design or adapt guidelines in consultation with other academic or hospital departments and FHSAs and medical audit advisory groups. FHSAs need to be responsible for supporting practices to help practitioners cope with the increased time required for structured care. Purchasers and regional health authorities need to ensure funding for local initiatives, and departments of public health should

more »

... priorities for guidelines according to local epidemiological factors and health needs assessment. To evaluate the effect of guidelines suitably large studies are required to detect changes in process and outcome measures. These studies should be restricted to answering questions about effectiveness; local projects on guidelines will need to be audited to ensure that their content and implementation fit local conditions. In addition, clinical care guidelines as part of service development require understanding of how change happens as well as being grounded in evidence that procedures are effective. This relies on action research and audit. A simple notion of guidelines as intrinsically virtuous and necessarily improving the quality of care is irresponsible. The current enthusiasm for general practice guidelines and other clinical guidelines is not matched by a commitment to evaluating their effect on practice. Evaluation should have a high profile within the NHS research and development strategy, and it needs national and regional coordination.