Brest Cancer Screening—An Alternative View Point

The purpose of breast cancer screening is to save lives. To have confidence that the national programme can achieve this aim it must be shown firstly that mortality reductions have been conclusively demonstrated in a relevant pilot study and secondly that these benefits can be reproduced in everyday practice. The Forrest report [1] was released when the encouraging results of the Health Insurance Plan (HIP) [2] and Swedish Two Counties Trials [3] were known but since then three randomised trials have failed to show a benefit after screening [4, 5,6], The seminal HIP trial suggested mortality reductions but this study is not fully relevant to the UK in 1989. About two-thirds of cancers were detected by clinical examination but the UK programme will rely on mammography alone. The HIP trial employed clinical examination and mammogra-phy every year?we shall only be screening triennially. The risk rates of non-attenders in the different populations may also be a reason why the HIP results would overestimate benefits in the UK. The HIP authors concluded that "women with a higher risk of breast cancer tend to self-select themselves for screening" [2], whereas the UK trial experience was the opposite: "women at high risk of breast cancer tend to be non-participants" [4], It seems that the same screening activity in different countries may produce different results. Another factor to be borne in mind when considering results from the USA is the threshold for biopsy of screen detected lesions. Benign to malignant ratios of 10:1 have been considered acceptable [7], even optimal [8], in America, but the Forrest Report states that a ratio of up to 3:1 should be aimed for [1]. Several case control studies have been performed in Holland [9, 10] and Italy [11] but are unreliable due to the possibility of selection bias, with poor prognosis cases accumulating in the control group. This was illustrated in the UK Trial of Early Detection of Breast Cancer (UKTEDBC) in which 51% of breast cancer deaths in the study group occurred in the 28% of women who did not attend for screening, and the case fatality rate in this group was higher than for control women [4]. Therefore we cannot be sure what fraction, if any, of the lower death rates of attenders in these case control studies was due to the screening process?it may simply have been due to the inherently more favourable risk profiles of …