Karla Kerlikowske, MD; Rebecca A. Hubbard, PhD; Diana L. Miglioretti, PhD; Berta M. Geller, EdD; Bonnie C. Yankaskas, PhD; Constance D. Lehman, MD, PhD; Stephen H. Taplin, MD, MPH; Edward A. Sickles, MD; for the Breast Cancer Surveillance Consortium
Acknowledgment: The authors thank the participating women, mammography facilities, and radiologists for the data they have provided for this study. A list of BCSC investigators and procedures for requesting BCSC data for research purposes can be found at http://breastscreening.cancer.gov/.
Grant Support: By the National Cancer Institute–funded BCSC cooperative agreement (grants U01CA63740, U01CA86076, U01CA86082, U01CA63736, U01CA70013, U01CA69976, U01CA63731, and U01CA70040) and National Cancer Institute grants R03CA150007 and RC2CA148577. The collection of cancer data used in this study was supported in part by several state public health departments and cancer registries throughout the United States; for a full description of these sources, see http://breastscreening.cancer.gov/work/acknowledgement.html.
Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M10-2769.
Reproducible Research Statement:Study protocol: Available from Dr. Kerlikowske (e-mail, Karla.Kerlikowske@ucsf.edu). Statistical code: Available from Dr. Hubbard (e-mail, firstname.lastname@example.org). Data set: Available after approval by the BCSC Steering Committee at http://breastscreening.cancer.gov/.
Requests for Single Reprints: Karla Kerlikowske, MD, San Francisco Veterans Affairs Medical Center, General Internal Medicine Section, 111A1, 4150 Clement Street, San Francisco, CA 94121; e-mail, Karla.Kerlikowske@ucsf.edu.
Current Author Addresses: Dr. Kerlikowske: San Francisco Veterans Affairs Medical Center, General Internal Medicine Section, 111A1, 4150 Clement Street, San Francisco, CA 94121.
Drs. Hubbard and Miglioretti: Group Health Research Institute, 1730 Minor Avenue, Suite 1600, Seattle, WA 98101-1448.
Dr. Geller: Health Promotion Research, University of Vermont, 429AR4, One South Prospect Street, Elevator C-4426, Burlington, VT 05401-3444.
Dr. Yankaskas: Department of Radiology, CB #7515, MRI, University of North Carolina, 106 Mason Farm Road, Chapel Hill, NC 27599-7515.
Dr. Lehman: Seattle Cancer Care Alliance, 825 Eastlake Avenue East, G2-600, Seattle, WA 98109.
Dr. Taplin: National Cancer Institute, 6130 Executive Boulevard, Rockville, MD 20852-7344.
Dr. Sickles: Department of Radiology, Box 1667, UCSF Medical Center, San Francisco, CA 94143-1667.
Author Contributions: Conception and design: K. Kerlikowske, R.A. Hubbard, D.L. Miglioretti, B.M. Geller, B.C. Yankaskas, S. Taplin.
Analysis and interpretation of the data: K. Kerlikowske, R.A. Hubbard, D.L. Miglioretti, B.M. Geller, B.C. Yankaskas, E.A. Sickles.
Drafting of the article: K. Kerlikowske, R.A. Hubbard, B.M. Geller, B.C. Yankaskas, C.D. Lehman, S. Taplin.
Critical revision of the article for important intellectual content: K. Kerlikowske, R.A. Hubbard, B.M. Geller, B.C. Yankaskas, C.D. Lehman, S. Taplin, E.A. Sickles.
Final approval of the article: K. Kerlikowske, R.A. Hubbard, D.L. Miglioretti, B.M. Geller, B.C. Yankaskas, C.D. Lehman, S. Taplin, E.A. Sickles.
Statistical expertise: R.A. Hubbard, D.L. Miglioretti.
Obtaining of funding: K. Kerlikowske, D.L. Miglioretti, B.C. Yankaskas, S. Taplin.
Collection and assembly of data: K. Kerlikowske, R.A. Hubbard, D.L. Miglioretti, B.M. Geller, B.C. Yankaskas.
Kerlikowske K., Hubbard R., Miglioretti D., Geller B., Yankaskas B., Lehman C., Taplin S., Sickles E., ; Comparative Effectiveness of Digital Versus Film-Screen Mammography in Community Practice in the United States: A Cohort Study. Ann Intern Med. 2011;155:493-502. doi: 10.7326/0003-4819-155-8-201110180-00005
Download citation file:
Published: Ann Intern Med. 2011;155(8):493-502.
Few studies have examined the comparative effectiveness of digital versus film-screen mammography in U.S. community practice.
To determine whether the interpretive performance of digital and film-screen mammography differs.
Prospective cohort study.
Mammography facilities in the Breast Cancer Surveillance Consortium.
329 261 women aged 40 to 79 years underwent 869 286 mammograms (231 034 digital; 638 252 film-screen).
Invasive cancer or ductal carcinoma in situ diagnosed within 12 months of a digital or film-screen examination and calculation of mammography sensitivity, specificity, cancer detection rates, and tumor outcomes.
Overall, cancer detection rates and tumor characteristics were similar for digital and film-screen mammography, but the sensitivity and specificity of each modality varied by age, tumor characteristics, breast density, and menopausal status. Compared with film-screen mammography, the sensitivity of digital mammography was significantly higher for women aged 60 to 69 years (89.9% vs. 83.0%; P = 0.014) and those with estrogen receptor–negative cancer (78.5% vs. 65.8%; P = 0.016); borderline significantly higher for women aged 40 to 49 years (82.4% vs. 75.6%; P = 0.071), those with extremely dense breasts (83.6% vs. 68.1%; P = 0.051), and pre- or perimenopausal women (87.1% vs. 81.7%; P = 0.057); and borderline significantly lower for women aged 50 to 59 years (80.5% vs. 85.1%; P = 0.097). The specificity of digital and film-screen mammography was similar by decade of age, except for women aged 40 to 49 years (88.0% vs. 89.7%; P < 0.001).
Statistical power for subgroup analyses was limited.
Overall, cancer detection with digital or film-screen mammography is similar in U.S. women aged 50 to 79 years undergoing screening mammography. Women aged 40 to 49 years are more likely to have extremely dense breasts and estrogen receptor–negative tumors; if they are offered mammography screening, they may choose to undergo digital mammography to optimize cancer detection.
National Cancer Institute.
John D.Keen, Radiologist
Cook County John H. Stroger Hospital, Chicago IL
October 17, 2011
What about computer-aided deception?
Dr. Kerlikoske et al have studied the utility of expensive digital imaging technology "after the fact" (1). Previously Dr. Kerlikowske correctly called for evidence that benefits outweigh the harms before the medical community adopts new technologies, specifically regarding computer -aided detection (CAD). (2) The current paper does not address this digital tag-along technology (1), but CAD uptake has important implications before anyone recommends digital screening, or concludes that digital technology is superior to film in community practice and hence the transition was a good idea.
My research has shown that almost all United States digital mammography facilities use CAD (RSNA abstract, 2011). Computer-aided "deception" in community practice makes overall screening accuracy worse by increasing recalls and decreasing specificity.(3) CAD also likely increases the harm of overdiagnosis by preferentially detecting DCIS.(4) Consequently, digital technology with CAD, the reality in community practice, may well make women worse off than film screening alone. This debate assumes that screening mammography by any means is a good idea for women, however, new evidence confirms that the harms are substantial compared to the limited mortality benefit.(4)
1. Kerlikowske, et al.
2. Kerlikowske K. A call for evidence of benefits outweighing harms before implementing new technologies: comment on "Diffusion of computer- aided mammography after mandated Medicare coverage". Arch Intern Med;170(11):990-1.
3. Fenton JJ, Abraham L, Taplin SH, et al. Effectiveness of Computer-Aided Detection in Community Mammography Practice. J Natl Cancer Inst;103(15):1152-1161.
4. Jorgensen KJ, Keen JD, Gotzsche PC. Is mammographic screening justifiable considering its substantial overdiagnosis rate and minor effect on mortality? Radiology;260(3):621-7.
StephanImfeld, MD PhD
University Hospital, Basel
October 18, 2011
Comparative Effectiveness of Digital Versus Film-Screen Mammography in Community Practice in the United States
TO THE EDITOR:
I have read with interest the recent study by Kerlikowske and colleagues (1), who compared the accuracy of digital mammography with that of film-screen mammography in a large population-based cohort of community -based imaging facilities.
The overall statistics of the main results show a very small but statistically significant difference in specificity (0.6 percent, p<0.001) and no statistical difference in sensitivity (p=0.21). Nevertheless the main focus as described in the abstract is set to a subgroup analysis of the latter (Table 4) comparing sensitivity for different estrogen receptor status and age strata without any statistical adjustment for multiple testing.
Subgroup analyzes of variables with no overall significance must be performed very cautiously due to multiple testing. In such cases, the significance levels require adjustments as for example a Bonferroni correction (2). Adjusting for the 10 comparisons in Table 4, the resulting p level of 0.005 would mean that none of the subgroup comparisons reaches significance, and therefore would invalidate the author's conclusions.
Bonferroni corrections themselves have been criticized, but even the most critical epidemiologists (3) consider them as indicated in such difficult cases.
Even though information on digital and analog mammography performance is of outmost interest, I have serious concerns that the results of the study, which could well influence future recommendations by the NIH on screening procedures, are heavily affected by the above mentioned statistical weakness.
(1) Kerlikowske K, Hubbard RA, Miglioretti DL, Geller BM, Yankaskas BC, Lehman CD, Taplin SH, Sickles EA, for the Breast Cancer Surveillance Consortium: Comparative Effectiveness of Digital Versus Film-Screen Mammography in Community Practice in the United States. A Cohort Study. Ann Intern Med. 155:493-502, 2011
(2) Bland JM, Altman DG, Multiple significance tests: the Bonferroni method. BMJ 310: 170, 1995
(3) Perneger TV: What's wrong with Bonferroni adjustments. BMJ 316:1236- 1238, 1998
Paul F.Pinsky, MD
Division of Cancer Prevention, National Cancer Institute
October 24, 2011
Why Not Make Recommendation Based on Breast Density, Not Age?
I found the article, "Comparative Effectiveness of Digital Versus Film-Screen Mammography in Community Practice in the United States" (1) to be very informative concerning the relative performance of these two modalities overall and in various subsets of women. However, I found the authors' conclusion that digital mammography should specifically be recommended for women 40-49 to be somewhat puzzling.
With digital compared to screen-film, the sensitivity is 6.8% higher in the 40-49 age group, 4.6% lower in women 50-59 and then again 6.9% higher in those 60-69 (and 1.4% higher in those 70-79). Although true quadratic-type relationships do occur in biology, in the absence of a mechanistic explanation, chance or some subtle form of bias are much more likely explanations. Therefore, in terms of age, it is not clear whether there is indeed any actual relationship between age and the difference in sensitivity between digital and screen-film. Further, as noted above, the difference in sensitivity is the same for 60-69 as for 40-49 year old women; however, in the former specificity is the same for both modalities (slightly, but not statistically significantly, higher with digital) while in the latter specificity is significantly lower with digital. Therefore, from the data on these age groups, there seems to be no reason to select out women 40-49 for digital over those 60-69.
The rationale for the selecting out of those 40-49 for digital, according to the authors, is that these women are more like to have ER negative tumors, for which sensitivity is significantly better with digital, and that these women are more likely to have extremely dense breasts, for which, again, sensitivity is better (borderline significant) with digital. With respect to the former, women can't know beforehand what type of ER status their tumor will be, so the only thing that is relevant is the overall sensitivity for the age group. Additionally, by the data in this study, the percent of cancers that are ER negative is 19.2% for women aged 40-49 versus 16.5% for those 50-79, hardly a difference upon which to make a clinical recommendation. With respect to breast density, this is a factor that women can be informed about; therefore, why not make the recommendation in terms of breast density and not age, especially since only a small minority of women 40-49, roughly 13% (2), fall into this (extremely dense breasts) category.
Paul F. Pinsky Division of Cancer Prevention National Cancer Institute
1. Kerlikowske K, Hubbard RA, Miglioretti DL, et al (2011). Comparative Effectiveness of Digital Versus Film-Screen Mammography in Community Practice in the United States. Annals Int Med 155: 493-502.
2. Schousboe JT, Kerlikowske K, Low A, Cummings SR (2011). Personalizing mammography by breast density and other risk factors for breast cancer: analysis of health benefits and cost effectiveness. Ann Int Med 155: 10-20.
KarlaKerlikowske, MD, Rebecca Hubbard, PhD, Biostatistics Unit, Group Health Research Institute, Seattle, WA, Department of Biostatistics, University of Washington School of Public Health, Seattle, WA
Department of Epidemiology and Biostatistics, University of California, San Francisco, CA, General I
December 5, 2011
We compared the interpretative performance of digital versus film- screen mammography in community practice and were reassured to find no subgroups of women experience substantially worse performance with either modality. In fact, digital mammography tended to perform as well or slightly better in most subgroups (1). Keen suggests had we taken into account use of computer-aided detection (CAD), we would have found more unfavorable performance results for digital compared to film-screen mammography since CAD is widely used to interpret digital mammograms. CAD was more commonly used for digital than film-screen mammography (56% vs. 23%) in our study. However, this differential use is incorporated in our performance estimates and despite the more frequent use of CAD with digital mammography, we found similar performance results for both modalities.
We did not adjust for multiple comparisons. While adjustments such as the Bonferroni correction control the type I error rate, they inflate the type II error rate, making it more likely we would miss actual differences (2). It is possible some of the significant differences we report are due to chance alone. However, by using an alpha-level of 0.05 we were able to identify differences in subgroups of women undergoing digital versus film- screen mammography for verification in future studies. Pinsky suggests our results support using breast density as a criterion to recommend digital mammography in women aged 40-49 years. Currently, breast cancer guidelines are based on age with consensus that women aged 50-74 years undergo mammography and debate about whether women aged 40-49 years should be screened. Although our results show there is slight variation in cancer detection by decade of age that could be due to chance, both modalities are acceptable for women aged 50-74 years since sensitivity is relatively high and similar for digital (range 80.8-90.3%) and film-screen (range 83.3-85.6%) mammography. Among women aged 40-49 years, sensitivity is lower for film-screen (76.1%) compared to digital (82.7%) mammography consistent with prior reports (3). Thus, women aged 40-49 years electing to be screened may choose digital mammography to optimize cancer detection.
We agree breast density is an important risk factor that influences breast cancer risk and detection (4) and can be used along with other risk factors to recommend age to start screening and screening frequency (5). In the Breast Cancer Surveillance Consortium, 78% of radiologists report BI-RADS breast density. For breast density to be used to apply risk-based screening in clinical practice, breast density would need to be routinely included in mammography reports.
1. Kerlikowske K, Hubbard RA, Miglioretti DL, et al. Comparative- effectiveness of digital vs. film-screen mammography in community practice in the U.S. Ann Intern Med 2011;155:493-502. [PMID: 22007043]
2. Rothman K. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1:43-46.
3. Pisano ED, Gatsonis C, Hendrick E, Yaffe M, Baum JK, Acharyya S, et al; Digital Mammographic Imaging Screening Trial (DMIST) Investigators Group. Diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med. 2005;353:1773-83. [PMID: 16169887]
4. Houssami N, Kerlikowske K. The impact of breast density on breast cancer risk and on breast screening. Current Breast Cancer Reports 2012;4:xx, In press
5. Schousboe JT, Kerlikowske K, Loh A, Cummings SR. Personalizing mammography by breast density and other risk factors for breast cancer: analysis of health benefits and cost-effectiveness. Ann Intern Med 2011;155:10-20. [PMID: 21727289]
to gain full access to the content and tools.
Learn more about subscription options.
Register Now for a free account.
Hematology/Oncology, Breast Cancer, Prevention/Screening.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only