Nationwide mammography screening registers: the best way to go for mortality evaluation – but there is no free lunch

Abstract

R Ljung et al. focus on an important issue in a letter to the Editor in this issue of Acta Radiologica (1). Access to screening data and data quality represent a great challenge in many western European countries having a nationwide mammography screening program. Sweden has been a pioneer country in mammography screening. The Swedish randomized controlled trials started in the late 1970s, but screening was not nationwide until about 20 years later. Although Sweden has one of the oldest Cancer Registries (established in 1958), a national health data register and some regional databases of high quality as well as a personal identity number which makes linking from several databases possible: the lack of a centralized screening register has made a national mortality evaluation of screening mammography extremely difficult (1). The Norwegian Breast Cancer Screening Program (NBCSP) started as a pilot project in 1995–1996 and became nationwide in 2005. The program is managed by the Cancer Registry of Norway which was established in 1951. Since 1953 all cancers diagnosed in Norway must, by law, be reported to the Registry. Due to an 11-digit personal identification number given to all inhabitants, screening information for all women invited to the program can be linked to the national cancer database and the cause of death registry. Publications have shown that a nationwide evaluation is possible (2). The latest development in Norway has shown that a centralized well-working nationwide screening register is no guarantee for ongoing evaluation. After running the NBCSP for more than 10 years, the mere participation in the program was considered inadequate by the lawyers of the Norwegian Data Inspectorate and the implementation of informed consent was mandatory. The Inspectorate decided in March 2008 that all data on normal findings in the NBCSP without informed consent had to be deleted by the end of 2011. An appeal against this decision was rejected, and the consequence would be that all future scientific evaluations of theNBCSP from its early beginningwould be impossible.At the time ofwriting, theNorwegianMinistry of Health and Care Services is working on an amendment of the Cancer Registry regulation, and hopefully this will save the screening data. The Advisory Board of the NBCSP has for some years strongly recommended an update of the screening database for including the BI-RADS lexicon and the rating scales recommended by the European Guidelines. The Nordic countries have been reluctant to incorporate the BI-RADS lexicon into their practice, but standardized international descriptors and indicators are of utmost importance for comparison of screening programs. During the first years of the NBCSP the Cancer Registry of Norway was responsible for the quality assurance as well as financial aspects. Due to reorganization of the healthcare system in Norway some few years ago, the Regional Health Enterprise has the financial control of the NBCSP, and an update of the screening database is obviously more difficult than when one institution has the responsibility for all the screening activity. The Editor agrees with the authors (1) that the recent publication in the New England Journal of Medicine by M Kalager et al. (3) suffers from serious limitations and does not provide a thorough and complete evaluation of the screening program. The conclusion of this study is in contrast with results recently published on case-control evaluations of European mammography service screening programs (4). The publicity the article in N Engl J Med has been given internationally shows how important an ‘official’ mortality evaluation of the NBCSP would be. Such evaluations were scheduled for 2009 but were stopped due to the decision by the Norwegian Data Inspectorate. The problem in Sweden addressed by Ljung et al. (1) and the problems arising in Norway demonstrates a common challenge to the Nordic countries: implementation of screening and the establishment of screening databases does not represent a guarantee for proper evaluation. Continuous collaboration between healthcare providers (authorities) and screening institutions is mandatory in order to solve the challenges regarding quality assurance, updating databases, access to data for scientific purposes, and consequently the ability to present valid mortality evaluations of organized mammography screening.