Objectives -To determine general practitioners' responses to and explanations for variation in rates of referral to hospital and how feedback of data on rates of referral could be used to facilitate practices in auditing their own referral behaviour.
Design -Visits by audit facilitators to general practices after feedback of details of rates of referral to hospital derived from annual reports in general practice.
Setting -92 general practices in East Anglia.
Results -General practitioners judged that access to specialist care, the individual skill of general practitioners, patient demand, and fear of litigation were major determinants of referral behaviour. Because there was widespread scepticism about the accuracy of the data on which the feedback was based and because there is no clear relation between rates of referral and quality of care, it was extremely difficult to encourage doctors to use the feedback as a basis for auditing their own hospital referrals.
Conclusion -If general practitioners are to contribute meaningfully to monitoring future changes in referral patterns it will be essential to develop reliable information systems in which doctors have confidence. Furthermore, audits need to be based on analysis of clinical cases rather than on rates of referral.