Predicted long-term mortality reduction associated with the second round of breast screening in East Anglia

Randomized trials have demonstrated that mammographic screening can reduce breast cancer mortality. Our aim was to estimate the reduction in mortality expected from the East Anglian Breast Screening Programme. Breast screening achieves benefit by improving cancer prognosis (reducing tumour size, nodal involvement and possibly grade) through earlier diagnosis. We compared cancer prognosis between women invited for screening and those not yet invited in East Anglia, UK, in order to predict the mortality reduction achievable by screening, independently of any reduction due to changes in treatment and underlying disease. Participants (both invited and not-yet invited) were women eligible for invitation to first and second screens and diagnosed with invasive breast cancer in 1989-96. Death rates were predicted based on the observed distribution of tumour grade, size and node status amongst 950 cancers diagnosed following first invitation, up to and including at second screen (excluding those detected at first screening), and 451 cancers presenting symptomatically in women awaiting first invitation during the staggered introduction of screening, after adjustment for lead time amongst screen detected cases. For all ages, the ratio of predicted breast cancer mortality in the invited compared with the uninvited group was 0.85 (95% CI 0.78, 0.93). It was 0.93 (0.80, 1.08) for women aged 50-54 at diagnosis and 0.81 (0.72, 0.91) for those aged 55-64. We conclude that, by 2004, the second round of screening in East Anglia should reduce mortality by around 7% in women below age 55 at diagnosis, and by around 19% in those aged 55-64.