Von Willebrand's disease.

matically mean that efficiency (as judged by some chosen yardstick) will rise. Again, the evidence is largely lacking2 and such experience as we have suggests that changes in utilisation and resource allocation are the product of a sudden and profound switch in attitude determined by conviction rather than appeal to facts, by new ideas suddenly born rather than the deliberate and intelligent review of figures and graphs. Such comment may sound cynical and destructive, but it is not: merely to point out that audit is still in its infancy. What Irving and Temple, among others, have done is to focus attention on the possibility, provided enthusiasm and conviction are there, of "closing the feedback loop"-bringing information derived from clinical experience to bear on future behaviour. What we now need is a prospective study to show whether useful as distinct from acceptable results can be obtained. Here the techniques of evaluation of the utility of audit-and, moreover, the additional skills in persuasion which are required to make clinicians more self-critical and to persuade them to act on available data-still have to be developed. Irving and Temple rightly remark that we do not want to expend our strictly limited money and resources on a procedure that is not cost-effective. Clinical genius needs to be focused on finding ways and means of digging out of the mass of information made available from medical audit those critical items which are not only relevant but also influential as levers for change. We look forward to hearing more of the Salford experience on these lines.