Objective
To determine the effect of false negative screens (missed cancers) in the prevalent screening round on the relative size distribution of cancers detected at the first incident screen.
Setting
The Bolton, Bury, and Rochdale breast screening programme.
Methods
One hundred and three breast cancers detected in the first incident round of screening were analysed. The previous (prevalent round) screening films taken between two and four years earlier were subjected to blinded review and classified as either true negatives (no significant abnormality visible) or false negatives (a suspicious abnormality visible at the site of the subsequently detected cancer). The pathological size, type, and grade (where appropriate) of the cancers were recorded
Results
Fifty one of the 103 cancers (49%) detected in the first incident round screen measured <15 mm in diameter. A total of 32 cases were classified as false negatives. Of these, 12 (38%) measured <15 mm in diameter. If all the false negative cancers had been detected at the prevalent screen, 39 (55%) of the remaining 71 cancers detected in the first incident round screen would have measured <15 mm. A relative excess of lobular carcinomas was found among the false negatives.
Conclusions
The findings suggest that although false negative screens in the prevalent round increase the number of cancers available for detection at the first incident round screen, many of these cancers are still <15 mm in diameter at detection. Cancer detection performance in the prevalent screening round has only a minor influence on the relative proportion of small and large cancers detected at the first incident round screen.
