A Fuhs1, S Bartholomäus, O Heidinger, H-W Hense. 1. Institut für Epidemiologie und Sozialmedizin, Universität Münster, Domagkstr. 3, 48129, Münster, Deutschland, fuhs@uni-muenster.de.
Abstract
BACKGROUND: In the German Mammography Screening Program (MSP), women aged 50-69 years are offered X-ray-based mammography for the early detection of breast cancer (BC) every 2nd year. To maintain the licensing of the MSP, evidence of a positive benefit-risk assessment of the radiation-related health risk has to be provided. Therefore, long-term effects of the program have to be shown by evaluating the BC mortality reduction based on the MSP. A current feasibility study attempts to develop different data flow models to evaluate whether, based on current legislations on data confidentiality, secondary data routinely available in the German health system can be linked in such a way that they can be used for epidemiological evaluation studies. We present the"NRW model," which builds on procedures developed and evaluated in the Epidemiological Cancer Registry of North Rhine-Westphalia (EKR-NRW). METHODS: Data of the Association of Statutory Health Physicians in Westfalen-Lippe (KVWL) are used to enumerate the cohort of women in WL who are entitled to MSP participation and their use of curative mammography outside of the MSP. The EKR-NRW provides epidemiological and medical data on all BC cases in WL, on cohort mortality, and on causes of death. The central MSP database MaSc offers the screening history of all MSP participants. The established uniform encryption methods employed in the EKR-NRW are used for linking records from the three data sources in one data-merging center (DZS). To this end, data are first captured in standardized formats, variably aggregated and transferred in an encrypted format, checked for anonymity and diversity level in an encrypted form, and eventually stored in a factually anonymized manner in the evaluation center (ES). Researchers can obtain data sets with plain text epidemiological-medical data from the ES for analyses.
BACKGROUND: In the German Mammography Screening Program (MSP), women aged 50-69 years are offered X-ray-based mammography for the early detection of breast cancer (BC) every 2nd year. To maintain the licensing of the MSP, evidence of a positive benefit-risk assessment of the radiation-related health risk has to be provided. Therefore, long-term effects of the program have to be shown by evaluating the BC mortality reduction based on the MSP. A current feasibility study attempts to develop different data flow models to evaluate whether, based on current legislations on data confidentiality, secondary data routinely available in the German health system can be linked in such a way that they can be used for epidemiological evaluation studies. We present the"NRW model," which builds on procedures developed and evaluated in the Epidemiological Cancer Registry of North Rhine-Westphalia (EKR-NRW). METHODS: Data of the Association of Statutory Health Physicians in Westfalen-Lippe (KVWL) are used to enumerate the cohort of women in WL who are entitled to MSP participation and their use of curative mammography outside of the MSP. The EKR-NRW provides epidemiological and medical data on all BC cases in WL, on cohort mortality, and on causes of death. The central MSP database MaSc offers the screening history of all MSP participants. The established uniform encryption methods employed in the EKR-NRW are used for linking records from the three data sources in one data-merging center (DZS). To this end, data are first captured in standardized formats, variably aggregated and transferred in an encrypted format, checked for anonymity and diversity level in an encrypted form, and eventually stored in a factually anonymized manner in the evaluation center (ES). Researchers can obtain data sets with plain text epidemiological-medical data from the ES for analyses.
Authors: Bettina Braun; Laura Khil; Joke Tio; Barbara Krause-Bergmann; Andrea Fuhs; Oliver Heidinger; Hans-Werner Hense Journal: Dtsch Arztebl Int Date: 2018-08-06 Impact factor: 5.594
Authors: Ingo Langner; Christoph Ohlmeier; Ulrike Haug; Hans Werner Hense; Jonas Czwikla; Hajo Zeeb Journal: BMJ Open Date: 2019-07-26 Impact factor: 2.692