Effect of participation on the cumulative risk of false-positive recall in a breast cancer screening programme

Abstract

One of the major concerns in breast cancer screening programmes is the number of women with mammographic abnormalities requiring further investigation that finally turn out to be negative. It has been reported that, apart from giving rise to anxiety1 and higher costs,2 such false-positivemammogram results might also affect subsequent screening attendance. Several studies have estimated the cumulative risk of falsepositive mammograms during a woman’s lifespan, and all have found a high cumulative false-positive recall rate, ranging from 20% to 50% after 10 rounds of screening. Elmore et al., analysed a cohort of women with irregular programme attendance, while other authors analysed cohorts of women participating in all screening rounds. Comparison of their results is difficult not only because of differences between health systems and protocols, but also due to differences in the selection criteria of the target population to estimate the cumulative risk of false-positive recall. It was hypothesized that estimations which only included women who participated in all screening rounds might be biased since, if false-positive results affect attendance at subsequent screening rounds, this might lead to cumulative false-positive rate underestimation in this cohort. To investigate this issue, the cumulative risk of false-positive recall was estimated for two groups of women with different profiles of adhesion to the programme using a cohort of women participating in a population-based breast cancer screening programme. Furthermore, the association between women’s characteristics and the cumulative risk of false-positive