Introduction/Objective: The aim of this project is to implement and evaluate existing national guidelines for referral and management of breast disorders. Breast pain and lumpiness have a high prevalence in the community, and referral rates appear to be increasing leading to psychological morbidity and unwarranted surgical intervention. Correct management of young, relatively low risk women is reassurance and symptomatic treatment in primary care. The aim of the guidelines is to promote appropriate referral.

Methods: This is a practice-based pragmatic randomised controlled trial. Practices are randomly assigned to receive one of two algorithms relating to breast lump and breast pain. Each practice acts as its own control; data is collected on both presentations in all practices, both before and after the introduction of the guidelines (Latin square design). Outcome measures relate to referral rates, physical and psychological morbidity and cost-effectiveness. Additional studies include a natural history study of presentation and referral, and a case-control study of risk perception.

Results: Participation rates are excellent, with 40 urban and rural group practices involved. The natural history study elicited that 13 women present per GP per year. The case control study of risk perception achieved an 83% response rate. Only 17% of 'cases' could correctly estimate the population risk from breast cancer. Further results will be available for presentation at the meeting.

Discussion: This study uses innovative methods for evaluating guidelines, which minimize the problems of the Hawthorne effect. The structure of the trial allows further observational and qualitative research to add meaning to the final results. The implications of the methods used for community based trials will be discussed in detail.