[Determination of interval cancer rates in the German mammography screening program using population-based cancer...

The evaluation of interval cancers (ICA) and false-negative diagnoses (FND) is important for the quality assurance of the mammography screening program (MSP) for women aged 50-69 years. We aimed to investigate ICA in the states of Lower Saxony (NDS) and North Rhine-Westphalia (NRW). ICA could be identified by record linkage of data from screening attendees (TN) with data from population-based cancer registries. The frequencies of ICA and FND are described, and problems in the categorization of ICA according to EU guidelines are discussed.

Record linkage in the cancer registries was performed with the encrypted data from 83,724 TN in NDS and 885,940 TN in NRW. ICA were identified; in NDS, ICA were categorized by the regional MSP reference center.

In all, 208 ICA in NDS and 2.018 ICA in NRW were ascertained. In relation to the regional background incidence rate (IR), the relative ICA rates were 22.5%/26.7% in the first year after screening (recommendation of EU guidelines <30% of IR) and 55.6%/54.2% in the second year after screening (recommendation of EU guidelines <50% of IR). It was found that 17.3% of all categorized ICA in NDS were FND (recommendation of EU guidelines <20%).

The relative ICA rates for the 2-year interval are consistent with the recommendations of the EU guidelines. Unlike in NDS, the categorization of ICA was not possible in NRW. The MSP needs the diagnostic mammograms of all ICA in order to identify FND. However, data protection regulations are hindering this process and nationwide comparability of FND is therefore not guaranteed. In NDS, the state government has enacted laws to facilitate the transfer of diagnostic mammograms to the MSP. There is need for this sort of action in other federal states too.