Breast cancer screening

In 2001 a Cochrane meta-analysis of randomised trials reported that mammography screening reduces breast cancer related mortality by 15% and that one in three cancers are overdiagnosed.1 No randomised trials have been performed since then, except an update of a Canadian trial that found no effect of mammography screening on breast cancer mortality at a cost of 22% overdiagnosis of cancer.2 Results from observational studies can be dichotomised in to those concluding that mammography screening is associated with little or no benefit but substantial overdiagnosis and those reporting larger benefits and only limited overdiagnosis.

In this issue, a new study by Autier and colleagues (doi:10.1136/bmj.j5224) analysing data from the Dutch screening programme is in the first category. It finds at most a 5% reduction in breast cancer related mortality and overdiagnosis in one out of three cancers detected.3 Since there was no reduction in late stage …