Interval Breast Cancers in the NHS Breast Screening Programme: Does the Current Definition Exclude Too Many?

Abstract

Objectives—: To examine the impact of the definition of interval breast cancers on interval cancer rates arising from the prevalent (first) screening round. Design—: Interval breast cancers arising from the prevalent (first) screening round at the Warwickshire, Solihull and Coventry Breast Screening Unit (17 April 1989 to 31 March 1992) were identified by comparison of data held at the unit with records at the West Midlands Cancer Intelligence Unit. Exclusion criteria used in National statistics were applied to this sample to quantify their impact on achieved interval cancer rates. The round lengths experienced by individual women at the unit were determined from the prevalent and incident invitation dates for 155 women with incident (re-screen) breast cancers detected in the second round. Setting—: Warwickshire, Solihull and Coventry Breast Screening Unit. Subects—59 017 women screened between 17 April 1989 and 31 March 1992 with a negative screening result and 155 women with incident screen detected cancers. Results—: A total of 278 interval cancers were identified, giving an overall rate from the prevalent screening round of 47.1/10 000 women screened. Of these, 213 met the criteria used in the definition of interval cancers for National statistics and were termed “core” interval cancers. The overall “core” interval rate was 36.1/10 000 women screened, similar to interval cancer rates found in the north west of United Kingdom. Thus applying commonly used exclusion criteria produced a 23.4% reduction in the apparent interval cancer rate, with the largest decrease resulting from the exclusion of cancers arising at 36 months or more from the last screen. Conclusions—: The exclusion criteria used in the definition of interval cancers have a significant impact on observed interval cancer rates. Of particular concern is the exclusion in the current National definitions of cancers arising at 36 months or more from the last screen, which may mask a problem with significant implications for the success of the NHSBSP.