Abstract
Several parameters for quality assurance of mammography screening require background incidence rates, i.e. the breast cancer incidence in the absence of screening. For new programmes, incidence rates for the years prior to the programme are usually taken as a substitute with the appropriate source being cancer registries. In Germany, the feasibility of quality-assured screening was examined in three pilot projects over the years 2001-2005, so that background incidence was needed for quality assurance. Since countrywide cancer registration was not yet well established and especially in the pilot area of Wiesbaden an active cancer registry did not exist, other ways to obtain accurate rates were needed. Thus, in this pilot a manual data collection of breast cancer cases from diagnosing pathologic institutes was conducted to quantify local breast cancer incidence and compare it with available regional registry data from a distant part of the country. Results showed large similarities of rates for the mid of the 1990s, but a stronger increase of incidence for the subsequent years specifically in the pilot area. This increase was likely related to particular efforts in favour of opportunistic screening in that area. They led to increased detection of small cancers, but were too unsystematic to initiate decline of advanced breast cancers on the population level. In view of getting appropriate figures for background incidence these results indicate that emphasis should be put on using rates which are less affected by opportunistic screening rather than obtaining geographically proximate data. We propose to use the 50-69 years incidence rate of breast cancer for the years 1996-2000 in the Saarland cancer registry as background incidence rate which is almost identical to the Wiesbaden data for 1996 and 1997 but increased less strongly during 1998-2000. This incidence rate is 2.5 per 1,000, while that found in the Wiesbaden area was 2.7 per 1,000.