Is breast compression associated with breast cancer detection and other early performance measures in a population-based breast cancer screening program?
We aimed to investigate early performance measures in a population-based breast cancer screening program stratified by compression force and pressure at the time of mammographic screening examination. Early performance measures included recall rate, rates of screen-detected and interval breast cancers, positive predictive value of recall (PPV), sensitivity, specificity, and histopathologic characteristics of screen-detected and interval breast cancers.
Information on 261,641 mammographic examinations from 93,444 subsequently screened women was used for analyses. The study period was 2007–2015. Compression force and pressure were categorized using tertiles as low, medium, or high. χ 2 test, t tests, and test for trend were used to examine differences between early performance measures across categories of compression force and pressure. We applied generalized estimating equations to identify the odds ratios (OR) of screen-detected or interval breast cancer associated with compression force and pressure, adjusting for fibroglandular and/or breast volume and age.
The recall rate decreased, while PPV and specificity increased with increasing compression force (p for trend <0.05 for all). The recall rate increased, while rate of screen-detected cancer, PPV, sensitivity, and specificity decreased with increasing compression pressure (p for trend <0.05 for all). High compression pressure was associated with higher odds of interval breast cancer compared with low compression pressure (1.89; 95% CI 1.43–2.48).
High compression force and low compression pressure were associated with more favorable early performance measures in the screening program.
KeywordsMammography Breast cancer screening Early performance measures Breast compression
We would like to thank Hilde Trå Hervig and Grethe Johansen, radiographers at the breast diagnostic center of Stavanger University Hospital, and Berit Hanestad, radiographer at the breast diagnostic center of Haukeland University Hospital, for help and support in collecting and processing the density data used in this study. We would also like to thank Kaitlyn Tsuruda, consultant at the Cancer Registry of Norway, for assistance with statistical analysis and epidemiologic insight in the study.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- 1.Perry N, Broeders M, de Wolf C, Tornberg S, Holland R, von Karsa L (2006) European guidelines for quality assurance in breast cancer screening and diagnosis, 4th edn. European Communities, LuxemburgGoogle Scholar
- 4.Kopans DB (2006) Mammography and the normal breast imaging. Lippincott Williams & Wilkins, London, pp 357–363Google Scholar
- 14.Holland K, Sechopoulos I, den Heeten G, Mann RM, Karssemeijer N (2016) Performance of Breast Cancer Screening Depends on Mammographic Compression. In: Tingberg A, Lång K, Timberg P (eds) Breast imaging: 13th international workshop, IWDM 2016, Malmö, Sweden, June 19–22, 2016, proceedings. Springer, Cham, pp 183–189Google Scholar
- 17.Matakina Technology Limited (2015) Volpara solutions—Volpara density. http://volparasolutions.com/solutions/volparadensity/. Accessed 1 Oct 2015