Answers are grounded in this paper and cite their sources.
Breast cancer screening: false positive rate is lower in older womenEditor-Werneke and McPherson 1 question the detection rates of cancer observed in the breast screening programme in East Sussex, Brighton, and Hove that we reported in our recent paper. 2 They suggest that these are the result of a high proportion of false positive screens.It is unfortunate that they did not consult us beforehand.The paper gives results of an ongoing study, but the cancers reported are those confirmed by histological examination after surgery, not preliminary suspected cancers at "positive" screens.The statistical programme was not run until some months after the end of the screening period, when all the screening episodes had been completed and the women concerned had been assessed, diagnosed, and operated on as appropriate.The East Sussex area has always tended to have a high detection rate for cancer, possibly because of the local sociodemographic composition of the population, and the rate over this period was higher than in the following year, without being exceptional.The 1995-6 result for the whole South West Thames region was 7.6 women with cancer for every 1000 women screened overall. 3he small study of older women in north London found 14.2 women with cancer per 1000 women screened aged 65-9. 4uring the period reported in our study the recall for assessment rate in women aged 65-7 was 4.6% (95% confidence interval 3.9% to 5.2%) and 6.2% (5.2% to 7.2%) in those aged 68-9, compared with 6.1% (5.8% to 6.4%) in those aged 50-64.Given the higher number of cancers detected this indicates a lower false positive rate in older women.This is to be expected because of the relative ease of interpreting mammograms and decreasing incidence of benign breast lesions in older women. 5 We now have the first year's results from the other two pilot sites, which will be published in due course.They confirm higher detection rates of cancer in older women but similar recall rates.Attendance is also encouragingly high.
No takes yet. Share an insight, caveat, or question.
Tozzi et al. (1998) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: