Breast cancer screening: should it be introduced and how?

This brief discussion of the papers by Gästrin and Paccaud is looking at differences in the results of different study types. Introducing screening means changing unsystematic to systematic screening: only the latter has been shown to be effective. The comparison between randomized trials, and follow-up of acceptors only, suggests that the non-acceptors have much higher risks. Programmes should therefore concentrate on trying to reach non-acceptors. The combination of mammographic screening and self-examination in a randomized trial could add to the existing evidence.