Digital mammography screening: Weighing reduced mortality against increased overdiagnosis Rianne de Gelder a, ⁎, Jacques Fracheboud a , Eveline A.M. Heijnsdijk a , Gerard den Heeten b , André L.M. Verbeek c , Mireille J.M. Broeders b, c , Gerrit Draisma a , Harry J. de Koning a a Erasmus MC, Department of Public Health, P.O. Box 2040, 3000 CA, Rotterdam, The Netherlands b National Expert and Training Centre for Breast Cancer Screening, Radboud University Nijmegen Medical Centre, PO Box 6873, 6503 GJ Nijmegen, The Netherlands c Radboud University Nijmegen Medical Centre, Department of Epidemiology, Biostatistics and HTA, PO Box 9101, 6500 HB Nijmegen, The Netherlands abstract article info Available online 21 June 2011 Keywords: Breast neoplasms Mass screening Early diagnosis Mammography Carcinoma Intraductal Noninfiltrating Computer Simulation Risk assessment Mortality Objective: Digital mammography has been shown to increase the detection of ductal carcinoma in situ (DCIS) compared to screen-film mammography. The benefits and risks of such an increase were assessed. Methods: Breast cancer detection rates were compared between 502,574 screen-film and 83,976 digital mammograms performed between 2004 and 2006 among Dutch screening participants. The detection rates were then modeled using a baseline model and two extreme models that respectively assumed a high rate of progression and no progression of preclinical DCIS to invasive cancer. With these models, breast cancer mortality and overdiagnosis were predicted. Results: The DCIS detection rate was significantly higher at digital mammography (1.2 per 1000 mammograms (95% C.I. 1.0–1.5)) than at screen-film mammography (0.7 per 1000 mammograms (95% C.I. 0.6–0.7)). Consequently, 287 (range progressive- non progressive model: 1–598) extra breast cancer deaths per 1,000,000 women (a 4.4% increase) were predicted to be prevented. An extra 401 (range: 165–2271) cancers would be overdiagnosed (a 21% increase). Conclusion: Modeling predicted that digital mammography screening would further reduce breast cancer mortality by 4.4%, at a 21% increased overdiagnosis rate. The consequences of digital screening, however, are sensitive to underlying assumptions on the natural history of DCIS. © 2011 Elsevier Inc. All rights reserved. Introduction Because of its improved contrast resolution, digital mammography has the potential to improve test accuracy compared to screen-film mammography (Lewin et al., 2002; Lewin et al., 2001; Pisano et al., 2005; Skaane et al., 2003). Digital mammography in women aged 45–69 resulted in significantly higher referral and cancer detection rates than screen-film mammography (Skaane et al., 2007). Other trials, however, showed that an improved accuracy was limited to women under the age of 50, women with dense breasts and pre- or peri-menopausal women (Lewin et al., 2002; Pisano et al., 2005). The long term benefits and risks of digital mammography in population-based screening have not yet been assessed. In 2004, a feasibility study of screening women with digital mammography was started in the Netherlands. Here, the results were modeled and used to predict the benefits and risks of implementing digital mammography as compared to screen-film mammography. Because various digital mammography studies observed an increased detection of ductal carcinoma in situ 1 (DCIS) and micro-calcifications frequently related to DCIS (Del Turco et al., 2007; Vigeland et al., 2008), our study focused on the benefits and risks of an increased detection of DCIS. However, the extent to which such lesions have the potential to become invasive cancers remains uncertain (National Institutes of Health State-of-the-Science Conference Statement). Detecting DCIS may prevent progression to invasive cancer, but may also imply that a lesion is diagnosed that would not have progressed to invasive cancer during the woman's lifetime (i.e. ‘overdiagnosis’). The main purpose of this study is to assess the consequences of an increased detection of DCIS by digital mammography relative to screen-film mammography screening. Preventive Medicine 53 (2011) 134–140 ⁎ Corresponding author at: Erasmus MC, Department of Public Health, Room AE - 137, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands. Fax: +31 10 7038474. E-mail addresses: r.degelder@erasmusmc.nl (R. de Gelder), j.fracheboud@erasmusmc.nl (J. Fracheboud), e.heijnsdijk@erasmusmc.nl (E.A.M. Heijnsdijk), g.denheeten@lrcb.nl (G. den Heeten), A.Verbeek@ebh.umcn.nl (A.L.M. Verbeek), M.Broeders@ebh.umcn.nl (M.J.M. Broeders), g.draisma@erasmusmc.nl (G. Draisma), h.dekoning@erasmusmc.nl (H.J. de Koning). 1 List of abbreviations: Ductal carcinoma in situ (DCIS); Medio-lateral oblique (MLO); Cranio-caudal (CC); Micro-simulation screening analysis (MISCAN); Invasive tumor diameter ≤5 mm (T1a); Invasive tumor diameter 6–10 mm (T1b); Invasive tumor diameter 11–20 mm (T1c); Invasive tumor diameter N 20 mm (T2+); Screen- film mammography (SFM); Digital mammography (DM). 0091-7435/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ypmed.2011.06.009 Contents lists available at ScienceDirect Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed