Age-specific sensitivities of mammographic screening for breast cancer.

The sensitivity of the mammographic screening test in the biennial screening program of Nijmegen is assessed by analyzing the occurrence of interval cancers, i.e. cancers surfacing clinically in the interval between a negative screening examination and the subsequent scheduled examination. The difference between the observed number of interval cancers and the expected number of clinically manifest cancers in the absence of screening for the interval period reflects the number of cancers detected by screening. The expected number should be limited by the number of those cancers that were not detectable at the time of the screening examination because their size was under the threshold of mammographic detectability (5 mm). In contrast to other sensitivity studies we took these 'fast growing' cancers into consideration, the numbers of which are estimated in each of the six-month periods of the two-year interval using age-specific tumor volume growth rates for three age groups: < 50, 50-69, and > or = 70 years. In patients under age 50, the sensitivity was 64% for cancers which would become clinically manifest within one year after screening. This sensitivity was lower than those obtained from the 50-69 and > or = 70 age groups, being 85% and 80%, respectively. For cancers that would become clinically manifest 12-18 months after screening, sensitivity decreases to 22% in the under age 50 group, and to 56% and 65% in the two above age 50 groups, respectively. We conclude that even when adjusted for growth rate, the mammographic screening test has a poor performance in the under age 50 group.