Surgical audit: variations in wound infection rates according to definition.

A prospective study of 702 postoperative surgical patients was undertaken to determine whether clinicians showed consistency in their interpretation of signs of infection in wounds. In the 62 cases where symptoms suggestive of such infection were noted, clinical signs were recorded, as were the clinicians' subjective impressions. The application of different criteria would have resulted in a substantial variation in the apparent rates of infection, which raises questions about the need for an agreed definition of the term 'wound infection'. The lack of a specific definition has implications for surgical audit.