EFFECT ON BREAST-CANCER MORTALITY OF BIENNIAL MAMMOGRAPHIC SCREENING OF WOMEN UNDER AGE 50

The effects on breast cancer mortality seen after 16 years of biennial screening of younger women ave assessed in this prospective cohort st udy. Since 1975 some 13,500 women, aged 35-49 in 1975, were invited to participate in the Nijmegen screening programme comprising a mammogra phic examination every 2 years. By the end of 1990, 75 women had died of breast cancer out of the 332 cases diagnosed after the start of the screening project. Women from the same birth cohort, living in Arnhem , a neighbouring city with a comparable population and without a scree ning project, were used as controls. In this city, 74 breast cancer de aths out of 284 cases occurred during the same period. In Nijmegen, af ter 16 years of follow-up, breast cancer mortality showed a non-signif icant reduction of 6% (95% confidence interval: 32% reduction, 29% exc ess). In the relevant period, after a time lag of 10 years from the st art of the programme, this reduction rose to 20% (95% confidence inter val: 48% reduction, 23% excess). No reduction in breast cancer mortali ty was observed in the first decade of screening. For a later period, a shift towards a reduction emerges, but the data are as yet inconclus ive. (C) 1995 Wiley-Liss, Inc.