Interval breast cancers in New South Wales, Australia, and comparisons with trials and other mammographic screening programmes.

Objective: To present information on 1st year interval breast cancer from the New South Wales mammographic screening programme and to compare with published results from trials and services.

Setting: New South Wales data were derived from a population based biennial mammographic screening programme, which achieved statewide coverage in 1995. Women aged 50-69 years screened during 1995-7 were included.

Methods: Bilateral two view mammography with reading by two radiologists is used for biennial screening examinations. Interval cancers were detected by the screening programme and by linkage with the statewide cancer registry. In situ carcinoma was excluded. Incidence of interval cancer was estimated as a proportion of the expected underlying incidence of breast cancer. Comparative data were derived from the published literature and meta-analyses were performed.

Results: Although randomised trials of screening have a proportional incidence by meta-analysis of 19% (95% confidence interval (95% CI) 12% to 25%), service studies yield a proportional incidence by meta-analysis of 27% (95% CI 25% to 30%), and more than half report proportional incidences greater than 25%. In the New South Wales mammographic screening programme the proportional incidence of interval breast cancer was 33% (95% CI 29% to 38%) for ages 50-59 years, 28% (95% CI 24% to 32%) for ages 60-69 years, and 31% (95% CI 28% to 34%) for ages 50-69 years combined. Proportional incidence in the New South Wales programme for ages 50-69 years was not significantly different from the rate for service studies by meta-analysis.

Conclusions: Effectiveness of mammographic screening for reducing mortality from breast cancer needs to be examined relative to rates of interval cancer from actual service situations as trials may concentrate resources and expertise in ways which may be less replicable in routine delivery of the service.