One-view breast tomosynthesis vs two-view mammography: a methodological issue

We read with interest the Article by Sophia Zackrisson and colleagues, 1 Zackrisson S Lang K Rosso A et al. One-view breast tomosynthesis versus two-view mammography in the Malmö Breast Tomosynthesis Screening Trial (MBTST): a prospective, population-based, diagnostic accuracy study. Lancet Oncol. 2018; 19: 1493-1503 Summary Full Text Full Text PDF PubMed Scopus (117) Google Scholar which compared the diagnostic accuracy of one-view digital breast tomosynthesis—a potential alternative to breast cancer screening—with standard two-view digital mammography in population-based breast cancer screening. The authors prospectively enrolled 14 848 women from Jan 27, 2010, to Feb 13, 2015. For digital breast tomosynthesis, the sensitivity was 81·1% (95% CI 74·2–86·9) and the specificity was 97·2% (97·0–97·5) for screening malignancy, with a negative predictive value of 99·8% (99·7–99·9) and a positive predictive value of 24·1% (20·5–28·0). By contrast, for digital mammography, the sensitivity was 60·4% (52·3–68·0) and the specificity was 98·1% (97·9–98·3), with a negative predictive value of 99·6% (99·4–99·7) and a positive predictive value of 25·9% (21·6–30·7). The authors concluded that, if cost-effective, one-view digital breast tomosynthesis warrants consideration as the preferred breast cancer screening method. One-view breast tomosynthesis versus two-view mammography in the Malmö Breast Tomosynthesis Screening Trial (MBTST): a prospective, population-based, diagnostic accuracy studyBreast cancer screening by use of one-view digital breast tomosynthesis with a reduced compression force has higher sensitivity at a slightly lower specificity for breast cancer detection compared with two-view digital mammography and has the potential to reduce the radiation dose and screen-reading burden required by two-view digital breast tomosynthesis with two-view digital mammography. Full-Text PDF One-view breast tomosynthesis vs two-view mammography: a methodological issue – Authors' replyWe thank Xiang Fang and colleagues for their interest in our Article.1 However, as already stated in the Article and by Fang and colleagues, a receiver operating characteristic (ROC) curve based on the data available in this study would be of questionable value and was thus not a predefined outcome measure included in the trial protocol. The reading data can be considered as binary: a recommendation to recall or not to recall the participant for further investigation. Because of this design, we would have been forced to base a ROC analysis on a single datapoint, which is essentially equivalent to fitting a curve to a single datapoint. Full-Text PDF