Digital Breast Tomosynthesis Utilization in the United States: A Survey of Physician Members of the Society of Breast Imaging Lara A. Hardesty, MDa, Sarah M. Kreidler, DPT, MSb, Deborah H. Glueck, PhDb
Purpose: To assess utilization of digital breast tomosynthesis (DBT) and examine criteria for offering DBT to patients. Methods: We created an online survey for physician members of the Society of Breast Imaging to assess their use of DBT. The questions covered availability of DBT at the participant’s practice, whether DBT was used for clinical care or research, clinical decision rules guiding patient selection for DBT, costs associated with DBT, plans to obtain DBT, and breast imaging practice characteristics. Fisher’s exact tests and logistic regression were used to compare DBT users and nonusers. Results: In all, 670 members responded (response rate ¼ 37%). Of these, 200 (30.0%) respondents reported using DBT, with 89% of these using DBT clinically. Participants were more likely to report DBT use if they worked at an academic practice (odds ratio [OR], 2.07; 95% confidence interval [CI], 1.41 to 3.03; P < .001), a practice with more than 3 breast imagers (OR, 2.36; 95% CI, 1.62 to 3.43; P < .001), or a practice with 7 or more mammography units (OR, 3.05; 95% CI, 2.11 to 4.39; P < .001). Criteria used to select patients to undergo DBT varied, with 107 (68.2%) using exam type (screening versus diagnostic), 25 (15.9%) using mammographic density, and 25 (15.9%) using breast cancer risk. Fees for DBT ranged from $25 to $250. In addition, 62.3% of nonusers planned to obtain DBT. Conclusion: DBT is becoming more common but remains a limited resource. Clinical guidelines would assist practices in deciding whether to adopt DBT and in standardizing which patients should receive DBT. Key Words: Digital breast tomosynthesis, mammography, breast cancer J Am Coll Radiol 2014;-:---. Copyright © 2014 American College of Radiology
INTRODUCTION
Digital breast tomosynthesis (DBT), used in combination with digital mammography, has been shown to improve the diagnostic accuracy of breast cancer screening and diagnosis [1-21]. However, little is known about how and where DBT is being used in the United States. Without understanding which radiologists have or have not chosen to employ DBT, it is difficult to identify barriers to adoption of this new technology. In addition, knowledge of current use of DBT will inform the development of clinical guidelines for DBT. DBT obtains multiple low-dose images of the breast from multiple angles and reconstructs those images into a
Department of Radiology, School of Medicine, University of Colorado Denver, Aurora, Colorado. b Department of Biostatistics, Colorado School of Public Health, University of Colorado Denver, Aurora, Colorado. There are no conflicts of interest to disclose. Corresponding author and reprints: Lara A. Hardesty, MD, Department of Radiology, University of Colorado Denver, Leprino Building, 12401 East 17th Avenue Mailstop L954, Aurora, CO 80045; e-mail: lara.hardesty@ ucdenver.edu. ª 2014 American College of Radiology 1546-1440/14/$36.00 http://dx.doi.org/10.1016/j.jacr.2013.11.025
a 3-dimensional data set that is processed into thin slices for interpretation [22]. The thin-slice interpretation helps minimize the interference of overlapping normal breast tissue when the radiologist is searching for breast cancer on a DBT study [2,15]. Existing literature on DBT has focused on the description of the technology [2,15,22,23]. Additional authors have described the diagnostic accuracy of DBT for cancer detection in observer performance studies [1,3-11,13,14] and the diagnostic accuracy of DBT for breast cancer screening [12,16-21]. No studies to date have described where and how DBT is being utilized. In this study, our goal was to describe the current use of DBT for clinical practice and research. We sought to identify practice characteristics that encouraged or inhibited adoption of DBT. In addition, we examined fees for DBT and whether these costs were passed on to the patient. Lastly, we reviewed breast imagers’ criteria for offering DBT to their patients to determine if a consensus existed regarding clinical indications for DBT. In other aspects of breast imaging, clinical indications for use are 1
2 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
specified clearly and are widely agreed upon [24]. Such guidelines do not exist for the clinical use of DBT. METHODS
An online survey was created to assess breast imagers’ use of DBT. The survey included a total of 23 questions. The questions covered several aspects of DBT use, including the availability of DBT at the participant’s clinic; whether DBT was used for clinical care, research purposes, or both; the clinical decision rules guiding patient selection for DBT; the patient costs associated with DBT; the practice’s future plans to obtain DBT; and general characteristics of the participants’ breast imaging practices. To limit the time burden for participants, branching logic was used. Thus, a survey respondent would only be presented with questions that were relevant based on previous responses. For example, a participant working at a clinic without DBT would not be asked to respond to questions about whether DBT was used for research or patient care. The survey was built using Survey Monkey’s “Select Plan” [25]. Written permission was obtained from the Society of Breast Imaging (SBI) to utilize its membership list for the study. An invitation to participate in the survey was sent by e-mail in November 2012 to all physician members of the SBI. These radiologists are experts in the field of breast imaging and are certified by either the ABR or the American Osteopathic Board of Radiology [26]. Radiologist members of the SBI were selected for 2 reasons: they are often the primary physicians interpreting breast imaging studies, and they commonly influence decisions regarding the adoption of new breast imaging technology into their practices. Only practicing members were invited to participate. Trainees and retired physicians were excluded. Descriptive statistics were calculated for practice demographics, both overall and stratified by DBT use. The size of a respondent’s practice was measured using 2 metrics: the number of radiologists at the practice reading breast imaging and the total number of mammography units at the practice. A map was produced showing the percentage of respondents using DBT in each geographic region of the United States. Frequencies were calculated for each of the survey responses. Separate Fisher’s exact tests [27] were used to examine the association between the odds of DBT use and practice type, and the association between DBT use and practice size. Logistic regression [28] was used to test the association between the odds of DBT use and the setting of the respondents’ breast imaging practice (rural, small town, small city, suburban, or major metropolitan area). For all hypothesis tests, odds ratios, 95% confidence intervals, and P values were produced. All hypothesis tests were performed with an alpha level of .05. Analyses were performed using R version 2.15.3 [29] and SAS version 9.3 [30].
RESULTS
We identified 1,930 eligible physicians from the SBI membership list. For 127 members, delivery of the e-mail invitation failed. Survey invitations were e-mailed successfully to 1,803 SBI members. A total of 670 physicians responded to the survey (response rate ¼ 37%). Number of Digital Breast Tomosynthesis Units
Of the 670 respondents, 200 (29.9%) reported using DBT, 102 (51%) of DBT users had only a single DBT unit at their practice, and 12 (6%) worked in practices with 7 or more DBT units. Only 11 (5.5%) DBT users worked in practices where all mammography units were DBT units. Demographics of Practices Offering DBT
Practice demographics are summarized in Table 1. Figure 1 shows the percentage of respondents using DBT by region of the United States. Academics versus Private Practice. Of respondents working in an academic institution, 38.8% (73 of 188) had DBT available in their practices, whereas 23.5% (104 of 443) of those working in private practice had DBT. Clinicians working in an academic environment were 2.07 times more likely to report using DBT (95% confidence interval [CI], 1.41 to 3.03; P < .001) than those working in private practice. Table 1. Practice demographics Overall DBT Users Characteristic (N [ 670) (n [ 200) Type of Practice Academic Private No response Practice setting Major metropolitan area Small city Suburban Small town Rural No response US region Northeast North central South West Other No response Number of breast imaging radiologists 3 or fewer More than 3 No response Number of mammography units 6 or fewer 7 or more No response
Nonusers (n [ 470)
188 (28.1%) 443 (66.1%) 39 (5.8%)
73 (36.5%) 104 (52.0%) 23 (11.5%)
115 (24.5%) 339 (72.1%) 16 (3.4%)
304 (45.4%)
99 (49.5%)
205 (43.6%)
152 146 21 8 39
(22.7%) (21.8%) (3.1%) (1.2%) (5.8%)
34 39 3 2 23
(17.0%) (19.5%) (1.5%) (1.0%) (11.5%)
118 107 18 6 16
(25.1%) (22.8%) (3.8%) (1.3%) (3.4%)
157 140 196 122 16 39
(23.4%) (20.9%) (29.3%) (18.2%) (2.4%) (5.8%)
58 35 36 40 8 23
(29.0%) (17.5%) (18.0%) (20.0%) (4.0%) (11.5%)
99 105 160 82 8 16
(21.1%) (22.3%) (34.0%) (17.4%) (1.7%) (3.4%)
409 (61.0%) 222 (33.1%) 39 (5.8%)
89 (44.5%) 88 (44.0%) 23 (11.5%)
320 (68.1%) 134 (28.5%) 16 (3.4%)
434 (64.8%) 215 (32.1%) 21 (3.1%)
95 (47.5%) 99 (49.5%) 6 (3.0%)
339 (72.1%) 116 (24.7%) 15 (3.2%)
Data presented are n (%). DBT ¼ digital breast tomosynthesis.
Hardesty et al/DBT Utilization in the US 3
Fig 1. Percentage of respondents using DBT by region of the United States.
Practice Size. Respondents from practices with more than 3 breast imagers were 2.36 times more likely to report DBT use (95% CI,1.62 to 3.43; P < .001) when compared with respondents from practices with 3 or fewer breast imagers. Similarly, respondents from practices with 7 or more mammography units were 3.05 times more likely to report using DBT (95% CI, 2.11 to 4.39; P < .001) when compared with radiologists in practices with 6 or fewer mammography units. Practice Setting. Practice setting was significantly associated with the odds of DBT use (P ¼ .02). Radiologists working in small cities were 0.52 times less likely to report DBT use (95% CI, 0.34 to .81; P ¼ .004) than breast imagers working in major metropolitan areas. The odds of reporting DBT use were similar between clinicians working in rural (odds ratio [OR], 0.6; 95% CI; 0.12 to 3.04; P ¼ .54), suburban (OR, 0.66, 95% CI, 0.43 to 1.01; P ¼ .06), and small town (OR, 0.3; 95% CI, 0.09 to 1.05; P ¼ .06) settings when compared with those working in a major metropolitan area. DBT Usage
Clinical Versus Research Use. Of the responders, 141 (70.5%) breast imagers using DBT reported using it for clinical patient care, with 36 (18%) using DBT for both research and patient care, and 17 (8.5%) using DBT for research only. Among respondents using DBT for both research and clinical care, only 3 (8.1%) reported having specific DBT units in their practices dedicated to research
(2 had a single dedicated unit, and 1 had 2 dedicated units). Among respondents using DBT for research only, 13 (76.5%) had a single dedicated research unit and 4 (23.5%) had 2 dedicated research units). Criteria for Offering DBT to Patients. Of 177 physicians using DBT for clinical care, only 20 (11.3%) reported that all clinical mammograms were performed using DBT. Among the 157 physicians who did not use DBT for all clinical mammograms, 107 (68.2%) offered DBT to patients based on the type of exam required (screening or diagnostic), 25 (15.9%) offered DBT to patients based on the patients’ breast tissue density, and 25 (15.9%) offered DBT to patients based on other risk factors, such as family history of breast cancer. By Screening Versus Diagnostic Mammogram. Of 103 physicians offering DBT to patients based on exam type, 26 (25.2%) offered DBT to all patients undergoing screening mammography, 15 (14.6%) offered DBT to all patients undergoing screening mammography at the main office only, 41 (39.8%) offered DBT to some, but not all, patients undergoing screening mammography, and 19 (18.4%) did not offer DBT to any screening mammography patients. Two survey participants did not respond to the question about screening mammography. Twenty (19.4%) offered DBT to all patients undergoing diagnostic mammography, 7 (6.8%) offered DBT to all patients undergoing diagnostic mammography at
4 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
the main office, 64 (62.1%) offered DBT to some, but not all, patients undergoing diagnostic mammography, and 10 (9.7%) did not offer DBT to any diagnostic mammography patients. Two survey participants did not respond to the question about diagnostic mammography. By Breast Density. There were 25 breast imagers offering DBT to patients based on breast tissue density. Radiologists were allowed to select one rationale or multiple rationales for their decision to offer DBT to patients based on breast tissue density. Twenty-four of these 25 (96%) offered DBT to patients with extremely dense breasts, 24 (96%) offered DBT to patients with heterogeneously dense breasts, 11 (44%) offered DBT to patients with scattered fibroglandular tissue, and 6 (24%) offered DBT to patients with fatty breasts. By Risk Factors Other Than Mammographic Density. Among the 25 physicians offering DBT based on the patient’s risk factors, 12 (48.0%) offered DBT to patients with a personal history of breast cancer and 12 (48.0%) offered DBT to patients with a high risk of developing breast cancer in the future based upon family history or upon a personal history of atypical ductal hyperplasia or lobular carcinoma in situ. Patient Costs
Of the 177 respondents using DBT clinically, 47 (26.6%) charged patients an upfront fee for the service, 122 (68.9%) did not charge a fee, and 8 (4.5%) declined to answer. Fees for DBT ranged from $25.00 to $250.00. The average fee for DBT services was $69.07 (standard deviation ¼ $40.92). Of the respondents using DBT clinically, 43 (24.3%) reported that patients were required to sign a waiver accepting costs for DBT that were not covered by insurance and 36 (83.7 %) of these reported that non-covered costs were passed to the patient. Plans to Obtain DBT
Of the 470 respondents who did not use DBT in their practice, 293 (62.3%) indicated that they had considered obtaining DBT. Among these considering purchase of DBT, 37 (12.6%) planned to do so within 6 months, 57 (19.5%) planned within 6 to 12 months, 85 (29.0%) planned within 1 to 2 years, and 73 (24.9%) did not know when they planned to obtain DBT. Finally, 41 (14.0%) reported that they would obtain DBT once a valid Current Procedural Terminology (CPT) code was available. DISCUSSION
There are several limitations to the survey. First, the study targeted only radiologist members of the SBI. Nonmembers may also play an important role in the adoption of new breast imaging technologies. Therefore, the sample may not reflect DBT use in the population of breast imagers as a whole. In addition, some respondents chose not to answer all of the questions that they were asked. It is unknown if DBT users were more
inclined to respond to the survey than non-DBT users, which may have created a nonresponse bias. Finally, e-mail invitations to the survey were sent to radiologists, rather than to breast imaging practices. Therefore, practices with greater numbers of physician members of the SBI are more likely to be represented in the survey results than practices with fewer SBI members. Our results indicate that DBT has crossed from a research technique to a clinically used technology. The data show that 70.5% of breast imagers who reported use of DBT used it for clinical patient care, and 18% used DBT for both research and patient care. A majority of practices (62.3%) without DBT plan to obtain it soon. However, DBT is far from being the accepted standard of care. Even among those using DBT for patient care, only 11.3% reported performing all clinical mammograms using DBT. Several barriers block adoption. The cost of converting all of a practice’s existing mammography units to DBT units at one time is likely prohibitive and has probably slowed adoption. Lack of a CPT code for DBT may be a barrier to implementation, because 14.0% of radiologists report that they will obtain DBT once a valid CPT code is available. Without a CPT code, radiologists may be reluctant to adopt the new technology and pass the costs on to patients. With no published guidelines regarding DBT use, criteria for offering DBT to patients varied among radiologists in our survey. Exam type (65.2%) was the most common criteria for offering patients DBT. Among those who selected patients for DBT by type of mammogram, 25.2% offered DBT to all screening mammography patients and 19.4% offered DBT to all diagnostic mammography patients. There seems to be an emerging consensus among radiologists who use tissue density as their rationale for offering DBT to their patients. Physicians overwhelmingly (96%) reported offering DBT to patients with extremely dense or heterogeneously dense breasts. The rationale is likely that DBT use may reduce superpositioning errors in some dense or heterogeneously dense breasts. Physicians probably choose DBT for these women in hopes of improving detection of cancers that are occult on 2D mammography, while simultaneously reducing false-positive diagnoses. The literature does, in fact, suggest that DBT may perform better clinically than 2D mammography. Reader performance studies on case sets containing a higher number of cancers than would be found in a clinical setting showed that DBT increased the rate of cancer detection [12,17]. One such study found a 22% increase in sensitivity for invasive cancers and a 3% increase for in situ cancers [17]. Another showed an 8% increase in cancer detection with DBT [12]. Reader performance studies also showed that DBT increased sensitivity by decreasing recall rates, with one study
Hardesty et al/DBT Utilization in the US 5
demonstrating a 40% decrease in the recall rate [1,4]. DBT was also shown to be useful in a diagnostic mammographic setting. One reader performance study compared the evaluation of masses using DBT versus using mammographic spot compression views and found similar visibility ratings, reader performance, and BI-RADS assessments, suggesting that spot compression views may not be necessary for mass characterization if a woman has undergone DBT [9]. Another showed that cancer size assessed with DBT was significantly more similar to tumor size at pathology than was size assessed by digital mammography alone [6]. These and other reader performance studies [7] led the FDA to approve DBT for clinical use in the United States on February 11, 2011 [31]. This FDA approval was for DBT to be used in combination with conventional mammography. In current clinical usage in the United States, “DBT” refers to the combined use of DBT plus conventional 2D digital mammography. Not surprisingly, use of DBT in combination with conventional 2D digital mammography increases the radiation dose that a patient receives relative to digital mammography alone [31,32]. However, the mean glandular dose that a patient receives from undergoing DBT and digital mammography in combination is less than the Mammography Quality Standards Act limit for a 2-view mammogram [33]. A more recent reader performance study found that compared with additional mammographic views, DBT significantly improves diagnostic accuracy for noncalcified lesions, including masses, distortions, and asymmetries [14]. Another showed that DBT can replace additional mammographic views obtained to evaluate noncalcified findings recalled from screening mammography, while maintaining similar sensitivity and specificity [19], and a third showed that the diagnostic accuracy of DBT is at least equal to digital mammographic spot compression views [10]. Four recent publications have demonstrated that clinical use of DBT in women undergoing screening mammography led to increased cancer detection rate (increased sensitivity) while simultaneously decreasing recall rates (improved specificity) [16,18,20,21]. One of these showed a 27% increase in cancer detection rate and a 15% decrease in recall rate with the use DBT relative to conventional digital mammography [16]. Another found that cancer detection rate increased from 4.0 per 1,000 without DBT to 5.4 per 1,000 with DBT while recall rates decreased from 8.7% to 5.5% [21]. A third study found that the cancer detection rate increased, but not significantly, from 5.2 per 1,000 without DBT to 5.7 per 1,000 with DBT, while recall rates decreased significantly from 12.0% to 8.4% [20]. The fourth study found that DBT significantly improved cancer detection rates from 5.3 cancers per 1,000 screens without DBT to 8.1 cancers per 1,000 screens when DBT plus digital mammography were used [18]. It is unusual for a new technology to improve
both sensitivity and specificity. These studies suggest that DBT will be beneficial in clinical use. CONCLUSIONS
There is increasing evidence of the clinical efficacy of DBT for breast cancer screening and diagnosis. DBT is a promising new technology used by 30% of respondents, but it remains a limited resource with only 11% of users performing all mammograms with DBT, and methods to select which patients receive DBT vary as do costs to patients. Clinical guidelines for DBT usage would help practices determine whether to adopt DBT and what patients would benefit the most from receiving DBT. Standardization of reimbursement for DBT via the adoption of a CPT code would help practitioners make adoption decisions based on science, rather than cost. TAKE-HOME POINTS
DBT use is growing rapidly, with 29.9% of survey respondents using DBT, 88.5% of whom use it clinically. DBT usage is more likely in academic practices and in larger practices. The majority (62.3%) of respondents who do not have DBT plan to obtain it. Access to DBT units is limited, with only 11.3% of DBT users performing all mammograms using DBT. To determine who should receive DBT, respondents use varying decision points: screening versus diagnostic mammogram, breast density, and risk factors for breast cancer. REFERENCES 1. Poplack SP, Tosteson TD, Kogel CA, Nagy HM. Digital breast tomosynthesis: initial experience in 98 women with abnormal digital screening mammography. Am J Roentgenol 2007;189:616-23. 2. Park JM, Franken EA, Garg M, Fajardo LL, Niklason LT. Breast tomosynthesis: present considerations and future applications. Radiographics 2007;27(suppl 1):S231-40. 3. Good WF, Abrams GS, Catullo VJ, et al. Digital breast tomosynthesis: a pilot observer study. AJR Am J Roentgenol 2008;190:865-9. 4. Gur D, Abrams GS, Chough DM, et al. Digital breast tomosynthesis: observer performance study. Am J Roentgenol 2009;193:586-91. 5. Gennaro G, Toledano A, di Maggio C, et al. Digital breast tomosynthesis versus digital mammography: a clinical performance study. Eur Radiol 2010;20:1545-53. 6. Förnvik D, Zackrisson S, Ljungberg O, et al. Breast tomosynthesis: Accuracy of tumor measurement compared with digital mammography and ultrasonography. Acta Radiol Stockh Swed 1987 2010;51:240-7. 7. Gur D, Bandos AI, Rockette HE, et al. Localized detection and classification of abnormalities on FFDM and tomosynthesis examinations rated under an FROC paradigm. Am J Roentgenol 2011;196:737-41. 8. Spangler ML, Zuley ML, Sumkin JH, et al. Detection and classification of calcifications on digital breast tomosynthesis and 2D digital mammography: a comparison. AJR Am J Roentgenol 2011;196:320-4. 9. Noroozian M, Hadjiiski L, Rahnama-Moghadam S, et al. Digital breast tomosynthesis is comparable to mammographic spot views for mass characterization. Radiology 2012;1(262):61-8.
6 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
10. Tagliafico A, Astengo D, Cavagnetto F, et al. One-to-one comparison between digital spot compression view and digital breast tomosynthesis. Eur Radiol 2012;1(22):539-44. 11. Wallis MG, Moa E, Zanca F, Leifland K, Danielsson M. Two-view and single-view tomosynthesis versus full-field digital mammography: high-resolution X-ray imaging observer study. Radiology 2012;262: 788-96. 12. Skaane P, Gullien R, Bjørndal H, et al. Digital breast tomosynthesis (DBT): initial experience in a clinical setting. Acta Radiol 2012;53:524-9. 13. Thibault F, Dromain C, Breucq C, et al. Digital breast tomosynthesis versus mammography and breast ultrasound: a multireader performance study. Eur Radiol 2013;23:2441-9. 14. Zuley ML, Bandos AI, Ganott MA, et al. Digital breast tomosynthesis versus supplemental diagnostic mammographic views for evaluation of noncalcified breast lesions. Radiology 2013;266:89-95. 15. Alakhras M, Bourne R, Rickard M, Ng KH, Pietrzyk M, Brennan PC. Digital tomosynthesis: A new future for breast imaging? Clin Radiol 2013;68:e225-36. 16. Skaane P, Bandos AI, Gullien R, et al. Comparison of digital mammography alone and digital mammography plus tomosynthesis in a population-based screening program. Radiology 2013;267: 47-56. 17. Rafferty EA, Park JM, Philpotts LE, et al. Assessing radiologist performance using combined digital mammography and breast tomosynthesis compared with digital mammography alone: results of a multicenter, multireader trial. Radiology 2013;266:104-13. 18. Ciatto S, Houssami N, Bernardi D, et al. Integration of 3D digital mammography with tomosynthesis for population breast-cancer screening (STORM): a prospective comparison study. Lancet Oncol 2013;14:583-9. 19. Brandt KR, Craig DA, Hoskins TL, et al. Can digital breast tomosynthesis replace conventional diagnostic mammography views for screening recalls without calcifications? A comparison study in a simulated clinical setting. Am J Roentgenol 2013;200:291-8. 20. Haas BM, Kalra V, Geisel J, Raghu M, Durand M, Philpotts LE. Comparison of tomosynthesis plus digital mammography and digital mammography alone for breast cancer screening. Radiology 2013;269: 694-700.
21. Rose SL, Tidwell AL, Bujnoch LJ, Kushwaha AC, Nordmann AS, Sexton R Jr. Implementation of breast tomosynthesis in a routine screening practice: an observational study. AJR Am J Roentgenol 2013;200:1401-8. 22. Niklason LT, Christian BT, Niklason LE, et al. Digital tomosynthesis in breast imaging. Radiology 1997;205:399-406. 23. Niklason LT, Kopans DB, Hamberg LM. Digital breast imaging: tomosynthesis and digital subtraction mammography. Breast Dis 1998;10:151-64. 24. Breast Imaging and Intervention Guidelines - American College of Radiology [Internet]. Available from: http://www.acr.org/Quality-Safety/ Standards-Guidelines/Practice-Guidelines-by-Modality/Breast-Imaging. Accessed Oct 1, 2013. 25. SurveyMonkey. SurveyMonkey: Free online survey software & questionnaire tool [Internet]. Available from: http://www.surveymonkey. com/. Accessed August 14, 2013. 26. Society of Breast Imaging: About SBI [Internet]. Available from: http://www.sbi-online.org/displaycommon.cfm?an¼1&subarticlenbr¼ 9#Mission. Accessed September 5, 2013. 27. Fisher RA. On the interpretation of c2 from contingency tables, and the calculation of P. J R Stat Soc 1922;85(1):87-94. 28. Jr DWH, Lemeshow S, Sturdivant RX. Applied Logistic Regression (Wiley Series in Probability and Statistics). 3 edition. Wiley; 2013. 29. Team RC. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria; 2013. Available from: http://www.R-project.org/ 30. Inc SI. SAS 9.3 Software, Version 9.3 [Internet]. Cary, NC; 2012. Available from: http://www.sas.com/software/sas9/ 31. Press Announcements.FDA approves first 3-D mammography imaging system [Internet]. Available from: http://www.fda.gov/NewsEvents/ Newsroom/PressAnnouncements/ucm243072.htm. Accessed September 5, 2013. 32. Cavagnetto F, Taccini G, Rosasco R, Bampi R, Calabrese M, Tagliafico A. “In Vivo” average glandular dose evaluation: one-to-one comparison between digital breast tomosynthesis and full-field digital mammography. Radiat Prot Dosimetry 2013;157:53-61. 33. Feng SSJ, Sechopoulos I. Clinical digital breast tomosynthesis system: dosimetric characterization. Radiology 2012;263:35-42.
Purpose: To assess utilization of digital breast tomosynthesis (DBT) and examine criteria for offering DBT to patients. Methods: We created an online survey for physician members of the Society of Breast Imaging to assess their use of DBT. The questions covered availability of DBT at the participant’s practice, whether DBT was used for clinical care or research, clinical decision rules guiding patient selection for DBT, costs associated with DBT, plans to obtain DBT, and breast imaging practice characteristics. Fisher’s exact tests and logistic regression were used to compare DBT users and nonusers. Results: In all, 670 members responded (response rate ¼ 37%). Of these, 200 (30.0%) respondents reported using DBT, with 89% of these using DBT clinically. Participants were more likely to report DBT use if they worked at an academic practice (odds ratio [OR], 2.07; 95% confidence interval [CI], 1.41 to 3.03; P < .001), a practice with more than 3 breast imagers (OR, 2.36; 95% CI, 1.62 to 3.43; P < .001), or a practice with 7 or more mammography units (OR, 3.05; 95% CI, 2.11 to 4.39; P < .001). Criteria used to select patients to undergo DBT varied, with 107 (68.2%) using exam type (screening versus diagnostic), 25 (15.9%) using mammographic density, and 25 (15.9%) using breast cancer risk. Fees for DBT ranged from $25 to $250. In addition, 62.3% of nonusers planned to obtain DBT. Conclusion: DBT is becoming more common but remains a limited resource. Clinical guidelines would assist practices in deciding whether to adopt DBT and in standardizing which patients should receive DBT. Key Words: Digital breast tomosynthesis, mammography, breast cancer J Am Coll Radiol 2014;-:---. Copyright © 2014 American College of Radiology
INTRODUCTION
Digital breast tomosynthesis (DBT), used in combination with digital mammography, has been shown to improve the diagnostic accuracy of breast cancer screening and diagnosis [1-21]. However, little is known about how and where DBT is being used in the United States. Without understanding which radiologists have or have not chosen to employ DBT, it is difficult to identify barriers to adoption of this new technology. In addition, knowledge of current use of DBT will inform the development of clinical guidelines for DBT. DBT obtains multiple low-dose images of the breast from multiple angles and reconstructs those images into a
Department of Radiology, School of Medicine, University of Colorado Denver, Aurora, Colorado. b Department of Biostatistics, Colorado School of Public Health, University of Colorado Denver, Aurora, Colorado. There are no conflicts of interest to disclose. Corresponding author and reprints: Lara A. Hardesty, MD, Department of Radiology, University of Colorado Denver, Leprino Building, 12401 East 17th Avenue Mailstop L954, Aurora, CO 80045; e-mail: lara.hardesty@ ucdenver.edu. ª 2014 American College of Radiology 1546-1440/14/$36.00 http://dx.doi.org/10.1016/j.jacr.2013.11.025
a 3-dimensional data set that is processed into thin slices for interpretation [22]. The thin-slice interpretation helps minimize the interference of overlapping normal breast tissue when the radiologist is searching for breast cancer on a DBT study [2,15]. Existing literature on DBT has focused on the description of the technology [2,15,22,23]. Additional authors have described the diagnostic accuracy of DBT for cancer detection in observer performance studies [1,3-11,13,14] and the diagnostic accuracy of DBT for breast cancer screening [12,16-21]. No studies to date have described where and how DBT is being utilized. In this study, our goal was to describe the current use of DBT for clinical practice and research. We sought to identify practice characteristics that encouraged or inhibited adoption of DBT. In addition, we examined fees for DBT and whether these costs were passed on to the patient. Lastly, we reviewed breast imagers’ criteria for offering DBT to their patients to determine if a consensus existed regarding clinical indications for DBT. In other aspects of breast imaging, clinical indications for use are 1
2 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
specified clearly and are widely agreed upon [24]. Such guidelines do not exist for the clinical use of DBT. METHODS
An online survey was created to assess breast imagers’ use of DBT. The survey included a total of 23 questions. The questions covered several aspects of DBT use, including the availability of DBT at the participant’s clinic; whether DBT was used for clinical care, research purposes, or both; the clinical decision rules guiding patient selection for DBT; the patient costs associated with DBT; the practice’s future plans to obtain DBT; and general characteristics of the participants’ breast imaging practices. To limit the time burden for participants, branching logic was used. Thus, a survey respondent would only be presented with questions that were relevant based on previous responses. For example, a participant working at a clinic without DBT would not be asked to respond to questions about whether DBT was used for research or patient care. The survey was built using Survey Monkey’s “Select Plan” [25]. Written permission was obtained from the Society of Breast Imaging (SBI) to utilize its membership list for the study. An invitation to participate in the survey was sent by e-mail in November 2012 to all physician members of the SBI. These radiologists are experts in the field of breast imaging and are certified by either the ABR or the American Osteopathic Board of Radiology [26]. Radiologist members of the SBI were selected for 2 reasons: they are often the primary physicians interpreting breast imaging studies, and they commonly influence decisions regarding the adoption of new breast imaging technology into their practices. Only practicing members were invited to participate. Trainees and retired physicians were excluded. Descriptive statistics were calculated for practice demographics, both overall and stratified by DBT use. The size of a respondent’s practice was measured using 2 metrics: the number of radiologists at the practice reading breast imaging and the total number of mammography units at the practice. A map was produced showing the percentage of respondents using DBT in each geographic region of the United States. Frequencies were calculated for each of the survey responses. Separate Fisher’s exact tests [27] were used to examine the association between the odds of DBT use and practice type, and the association between DBT use and practice size. Logistic regression [28] was used to test the association between the odds of DBT use and the setting of the respondents’ breast imaging practice (rural, small town, small city, suburban, or major metropolitan area). For all hypothesis tests, odds ratios, 95% confidence intervals, and P values were produced. All hypothesis tests were performed with an alpha level of .05. Analyses were performed using R version 2.15.3 [29] and SAS version 9.3 [30].
RESULTS
We identified 1,930 eligible physicians from the SBI membership list. For 127 members, delivery of the e-mail invitation failed. Survey invitations were e-mailed successfully to 1,803 SBI members. A total of 670 physicians responded to the survey (response rate ¼ 37%). Number of Digital Breast Tomosynthesis Units
Of the 670 respondents, 200 (29.9%) reported using DBT, 102 (51%) of DBT users had only a single DBT unit at their practice, and 12 (6%) worked in practices with 7 or more DBT units. Only 11 (5.5%) DBT users worked in practices where all mammography units were DBT units. Demographics of Practices Offering DBT
Practice demographics are summarized in Table 1. Figure 1 shows the percentage of respondents using DBT by region of the United States. Academics versus Private Practice. Of respondents working in an academic institution, 38.8% (73 of 188) had DBT available in their practices, whereas 23.5% (104 of 443) of those working in private practice had DBT. Clinicians working in an academic environment were 2.07 times more likely to report using DBT (95% confidence interval [CI], 1.41 to 3.03; P < .001) than those working in private practice. Table 1. Practice demographics Overall DBT Users Characteristic (N [ 670) (n [ 200) Type of Practice Academic Private No response Practice setting Major metropolitan area Small city Suburban Small town Rural No response US region Northeast North central South West Other No response Number of breast imaging radiologists 3 or fewer More than 3 No response Number of mammography units 6 or fewer 7 or more No response
Nonusers (n [ 470)
188 (28.1%) 443 (66.1%) 39 (5.8%)
73 (36.5%) 104 (52.0%) 23 (11.5%)
115 (24.5%) 339 (72.1%) 16 (3.4%)
304 (45.4%)
99 (49.5%)
205 (43.6%)
152 146 21 8 39
(22.7%) (21.8%) (3.1%) (1.2%) (5.8%)
34 39 3 2 23
(17.0%) (19.5%) (1.5%) (1.0%) (11.5%)
118 107 18 6 16
(25.1%) (22.8%) (3.8%) (1.3%) (3.4%)
157 140 196 122 16 39
(23.4%) (20.9%) (29.3%) (18.2%) (2.4%) (5.8%)
58 35 36 40 8 23
(29.0%) (17.5%) (18.0%) (20.0%) (4.0%) (11.5%)
99 105 160 82 8 16
(21.1%) (22.3%) (34.0%) (17.4%) (1.7%) (3.4%)
409 (61.0%) 222 (33.1%) 39 (5.8%)
89 (44.5%) 88 (44.0%) 23 (11.5%)
320 (68.1%) 134 (28.5%) 16 (3.4%)
434 (64.8%) 215 (32.1%) 21 (3.1%)
95 (47.5%) 99 (49.5%) 6 (3.0%)
339 (72.1%) 116 (24.7%) 15 (3.2%)
Data presented are n (%). DBT ¼ digital breast tomosynthesis.
Hardesty et al/DBT Utilization in the US 3
Fig 1. Percentage of respondents using DBT by region of the United States.
Practice Size. Respondents from practices with more than 3 breast imagers were 2.36 times more likely to report DBT use (95% CI,1.62 to 3.43; P < .001) when compared with respondents from practices with 3 or fewer breast imagers. Similarly, respondents from practices with 7 or more mammography units were 3.05 times more likely to report using DBT (95% CI, 2.11 to 4.39; P < .001) when compared with radiologists in practices with 6 or fewer mammography units. Practice Setting. Practice setting was significantly associated with the odds of DBT use (P ¼ .02). Radiologists working in small cities were 0.52 times less likely to report DBT use (95% CI, 0.34 to .81; P ¼ .004) than breast imagers working in major metropolitan areas. The odds of reporting DBT use were similar between clinicians working in rural (odds ratio [OR], 0.6; 95% CI; 0.12 to 3.04; P ¼ .54), suburban (OR, 0.66, 95% CI, 0.43 to 1.01; P ¼ .06), and small town (OR, 0.3; 95% CI, 0.09 to 1.05; P ¼ .06) settings when compared with those working in a major metropolitan area. DBT Usage
Clinical Versus Research Use. Of the responders, 141 (70.5%) breast imagers using DBT reported using it for clinical patient care, with 36 (18%) using DBT for both research and patient care, and 17 (8.5%) using DBT for research only. Among respondents using DBT for both research and clinical care, only 3 (8.1%) reported having specific DBT units in their practices dedicated to research
(2 had a single dedicated unit, and 1 had 2 dedicated units). Among respondents using DBT for research only, 13 (76.5%) had a single dedicated research unit and 4 (23.5%) had 2 dedicated research units). Criteria for Offering DBT to Patients. Of 177 physicians using DBT for clinical care, only 20 (11.3%) reported that all clinical mammograms were performed using DBT. Among the 157 physicians who did not use DBT for all clinical mammograms, 107 (68.2%) offered DBT to patients based on the type of exam required (screening or diagnostic), 25 (15.9%) offered DBT to patients based on the patients’ breast tissue density, and 25 (15.9%) offered DBT to patients based on other risk factors, such as family history of breast cancer. By Screening Versus Diagnostic Mammogram. Of 103 physicians offering DBT to patients based on exam type, 26 (25.2%) offered DBT to all patients undergoing screening mammography, 15 (14.6%) offered DBT to all patients undergoing screening mammography at the main office only, 41 (39.8%) offered DBT to some, but not all, patients undergoing screening mammography, and 19 (18.4%) did not offer DBT to any screening mammography patients. Two survey participants did not respond to the question about screening mammography. Twenty (19.4%) offered DBT to all patients undergoing diagnostic mammography, 7 (6.8%) offered DBT to all patients undergoing diagnostic mammography at
4 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
the main office, 64 (62.1%) offered DBT to some, but not all, patients undergoing diagnostic mammography, and 10 (9.7%) did not offer DBT to any diagnostic mammography patients. Two survey participants did not respond to the question about diagnostic mammography. By Breast Density. There were 25 breast imagers offering DBT to patients based on breast tissue density. Radiologists were allowed to select one rationale or multiple rationales for their decision to offer DBT to patients based on breast tissue density. Twenty-four of these 25 (96%) offered DBT to patients with extremely dense breasts, 24 (96%) offered DBT to patients with heterogeneously dense breasts, 11 (44%) offered DBT to patients with scattered fibroglandular tissue, and 6 (24%) offered DBT to patients with fatty breasts. By Risk Factors Other Than Mammographic Density. Among the 25 physicians offering DBT based on the patient’s risk factors, 12 (48.0%) offered DBT to patients with a personal history of breast cancer and 12 (48.0%) offered DBT to patients with a high risk of developing breast cancer in the future based upon family history or upon a personal history of atypical ductal hyperplasia or lobular carcinoma in situ. Patient Costs
Of the 177 respondents using DBT clinically, 47 (26.6%) charged patients an upfront fee for the service, 122 (68.9%) did not charge a fee, and 8 (4.5%) declined to answer. Fees for DBT ranged from $25.00 to $250.00. The average fee for DBT services was $69.07 (standard deviation ¼ $40.92). Of the respondents using DBT clinically, 43 (24.3%) reported that patients were required to sign a waiver accepting costs for DBT that were not covered by insurance and 36 (83.7 %) of these reported that non-covered costs were passed to the patient. Plans to Obtain DBT
Of the 470 respondents who did not use DBT in their practice, 293 (62.3%) indicated that they had considered obtaining DBT. Among these considering purchase of DBT, 37 (12.6%) planned to do so within 6 months, 57 (19.5%) planned within 6 to 12 months, 85 (29.0%) planned within 1 to 2 years, and 73 (24.9%) did not know when they planned to obtain DBT. Finally, 41 (14.0%) reported that they would obtain DBT once a valid Current Procedural Terminology (CPT) code was available. DISCUSSION
There are several limitations to the survey. First, the study targeted only radiologist members of the SBI. Nonmembers may also play an important role in the adoption of new breast imaging technologies. Therefore, the sample may not reflect DBT use in the population of breast imagers as a whole. In addition, some respondents chose not to answer all of the questions that they were asked. It is unknown if DBT users were more
inclined to respond to the survey than non-DBT users, which may have created a nonresponse bias. Finally, e-mail invitations to the survey were sent to radiologists, rather than to breast imaging practices. Therefore, practices with greater numbers of physician members of the SBI are more likely to be represented in the survey results than practices with fewer SBI members. Our results indicate that DBT has crossed from a research technique to a clinically used technology. The data show that 70.5% of breast imagers who reported use of DBT used it for clinical patient care, and 18% used DBT for both research and patient care. A majority of practices (62.3%) without DBT plan to obtain it soon. However, DBT is far from being the accepted standard of care. Even among those using DBT for patient care, only 11.3% reported performing all clinical mammograms using DBT. Several barriers block adoption. The cost of converting all of a practice’s existing mammography units to DBT units at one time is likely prohibitive and has probably slowed adoption. Lack of a CPT code for DBT may be a barrier to implementation, because 14.0% of radiologists report that they will obtain DBT once a valid CPT code is available. Without a CPT code, radiologists may be reluctant to adopt the new technology and pass the costs on to patients. With no published guidelines regarding DBT use, criteria for offering DBT to patients varied among radiologists in our survey. Exam type (65.2%) was the most common criteria for offering patients DBT. Among those who selected patients for DBT by type of mammogram, 25.2% offered DBT to all screening mammography patients and 19.4% offered DBT to all diagnostic mammography patients. There seems to be an emerging consensus among radiologists who use tissue density as their rationale for offering DBT to their patients. Physicians overwhelmingly (96%) reported offering DBT to patients with extremely dense or heterogeneously dense breasts. The rationale is likely that DBT use may reduce superpositioning errors in some dense or heterogeneously dense breasts. Physicians probably choose DBT for these women in hopes of improving detection of cancers that are occult on 2D mammography, while simultaneously reducing false-positive diagnoses. The literature does, in fact, suggest that DBT may perform better clinically than 2D mammography. Reader performance studies on case sets containing a higher number of cancers than would be found in a clinical setting showed that DBT increased the rate of cancer detection [12,17]. One such study found a 22% increase in sensitivity for invasive cancers and a 3% increase for in situ cancers [17]. Another showed an 8% increase in cancer detection with DBT [12]. Reader performance studies also showed that DBT increased sensitivity by decreasing recall rates, with one study
Hardesty et al/DBT Utilization in the US 5
demonstrating a 40% decrease in the recall rate [1,4]. DBT was also shown to be useful in a diagnostic mammographic setting. One reader performance study compared the evaluation of masses using DBT versus using mammographic spot compression views and found similar visibility ratings, reader performance, and BI-RADS assessments, suggesting that spot compression views may not be necessary for mass characterization if a woman has undergone DBT [9]. Another showed that cancer size assessed with DBT was significantly more similar to tumor size at pathology than was size assessed by digital mammography alone [6]. These and other reader performance studies [7] led the FDA to approve DBT for clinical use in the United States on February 11, 2011 [31]. This FDA approval was for DBT to be used in combination with conventional mammography. In current clinical usage in the United States, “DBT” refers to the combined use of DBT plus conventional 2D digital mammography. Not surprisingly, use of DBT in combination with conventional 2D digital mammography increases the radiation dose that a patient receives relative to digital mammography alone [31,32]. However, the mean glandular dose that a patient receives from undergoing DBT and digital mammography in combination is less than the Mammography Quality Standards Act limit for a 2-view mammogram [33]. A more recent reader performance study found that compared with additional mammographic views, DBT significantly improves diagnostic accuracy for noncalcified lesions, including masses, distortions, and asymmetries [14]. Another showed that DBT can replace additional mammographic views obtained to evaluate noncalcified findings recalled from screening mammography, while maintaining similar sensitivity and specificity [19], and a third showed that the diagnostic accuracy of DBT is at least equal to digital mammographic spot compression views [10]. Four recent publications have demonstrated that clinical use of DBT in women undergoing screening mammography led to increased cancer detection rate (increased sensitivity) while simultaneously decreasing recall rates (improved specificity) [16,18,20,21]. One of these showed a 27% increase in cancer detection rate and a 15% decrease in recall rate with the use DBT relative to conventional digital mammography [16]. Another found that cancer detection rate increased from 4.0 per 1,000 without DBT to 5.4 per 1,000 with DBT while recall rates decreased from 8.7% to 5.5% [21]. A third study found that the cancer detection rate increased, but not significantly, from 5.2 per 1,000 without DBT to 5.7 per 1,000 with DBT, while recall rates decreased significantly from 12.0% to 8.4% [20]. The fourth study found that DBT significantly improved cancer detection rates from 5.3 cancers per 1,000 screens without DBT to 8.1 cancers per 1,000 screens when DBT plus digital mammography were used [18]. It is unusual for a new technology to improve
both sensitivity and specificity. These studies suggest that DBT will be beneficial in clinical use. CONCLUSIONS
There is increasing evidence of the clinical efficacy of DBT for breast cancer screening and diagnosis. DBT is a promising new technology used by 30% of respondents, but it remains a limited resource with only 11% of users performing all mammograms with DBT, and methods to select which patients receive DBT vary as do costs to patients. Clinical guidelines for DBT usage would help practices determine whether to adopt DBT and what patients would benefit the most from receiving DBT. Standardization of reimbursement for DBT via the adoption of a CPT code would help practitioners make adoption decisions based on science, rather than cost. TAKE-HOME POINTS
DBT use is growing rapidly, with 29.9% of survey respondents using DBT, 88.5% of whom use it clinically. DBT usage is more likely in academic practices and in larger practices. The majority (62.3%) of respondents who do not have DBT plan to obtain it. Access to DBT units is limited, with only 11.3% of DBT users performing all mammograms using DBT. To determine who should receive DBT, respondents use varying decision points: screening versus diagnostic mammogram, breast density, and risk factors for breast cancer. REFERENCES 1. Poplack SP, Tosteson TD, Kogel CA, Nagy HM. Digital breast tomosynthesis: initial experience in 98 women with abnormal digital screening mammography. Am J Roentgenol 2007;189:616-23. 2. Park JM, Franken EA, Garg M, Fajardo LL, Niklason LT. Breast tomosynthesis: present considerations and future applications. Radiographics 2007;27(suppl 1):S231-40. 3. Good WF, Abrams GS, Catullo VJ, et al. Digital breast tomosynthesis: a pilot observer study. AJR Am J Roentgenol 2008;190:865-9. 4. Gur D, Abrams GS, Chough DM, et al. Digital breast tomosynthesis: observer performance study. Am J Roentgenol 2009;193:586-91. 5. Gennaro G, Toledano A, di Maggio C, et al. Digital breast tomosynthesis versus digital mammography: a clinical performance study. Eur Radiol 2010;20:1545-53. 6. Förnvik D, Zackrisson S, Ljungberg O, et al. Breast tomosynthesis: Accuracy of tumor measurement compared with digital mammography and ultrasonography. Acta Radiol Stockh Swed 1987 2010;51:240-7. 7. Gur D, Bandos AI, Rockette HE, et al. Localized detection and classification of abnormalities on FFDM and tomosynthesis examinations rated under an FROC paradigm. Am J Roentgenol 2011;196:737-41. 8. Spangler ML, Zuley ML, Sumkin JH, et al. Detection and classification of calcifications on digital breast tomosynthesis and 2D digital mammography: a comparison. AJR Am J Roentgenol 2011;196:320-4. 9. Noroozian M, Hadjiiski L, Rahnama-Moghadam S, et al. Digital breast tomosynthesis is comparable to mammographic spot views for mass characterization. Radiology 2012;1(262):61-8.
6 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
10. Tagliafico A, Astengo D, Cavagnetto F, et al. One-to-one comparison between digital spot compression view and digital breast tomosynthesis. Eur Radiol 2012;1(22):539-44. 11. Wallis MG, Moa E, Zanca F, Leifland K, Danielsson M. Two-view and single-view tomosynthesis versus full-field digital mammography: high-resolution X-ray imaging observer study. Radiology 2012;262: 788-96. 12. Skaane P, Gullien R, Bjørndal H, et al. Digital breast tomosynthesis (DBT): initial experience in a clinical setting. Acta Radiol 2012;53:524-9. 13. Thibault F, Dromain C, Breucq C, et al. Digital breast tomosynthesis versus mammography and breast ultrasound: a multireader performance study. Eur Radiol 2013;23:2441-9. 14. Zuley ML, Bandos AI, Ganott MA, et al. Digital breast tomosynthesis versus supplemental diagnostic mammographic views for evaluation of noncalcified breast lesions. Radiology 2013;266:89-95. 15. Alakhras M, Bourne R, Rickard M, Ng KH, Pietrzyk M, Brennan PC. Digital tomosynthesis: A new future for breast imaging? Clin Radiol 2013;68:e225-36. 16. Skaane P, Bandos AI, Gullien R, et al. Comparison of digital mammography alone and digital mammography plus tomosynthesis in a population-based screening program. Radiology 2013;267: 47-56. 17. Rafferty EA, Park JM, Philpotts LE, et al. Assessing radiologist performance using combined digital mammography and breast tomosynthesis compared with digital mammography alone: results of a multicenter, multireader trial. Radiology 2013;266:104-13. 18. Ciatto S, Houssami N, Bernardi D, et al. Integration of 3D digital mammography with tomosynthesis for population breast-cancer screening (STORM): a prospective comparison study. Lancet Oncol 2013;14:583-9. 19. Brandt KR, Craig DA, Hoskins TL, et al. Can digital breast tomosynthesis replace conventional diagnostic mammography views for screening recalls without calcifications? A comparison study in a simulated clinical setting. Am J Roentgenol 2013;200:291-8. 20. Haas BM, Kalra V, Geisel J, Raghu M, Durand M, Philpotts LE. Comparison of tomosynthesis plus digital mammography and digital mammography alone for breast cancer screening. Radiology 2013;269: 694-700.
21. Rose SL, Tidwell AL, Bujnoch LJ, Kushwaha AC, Nordmann AS, Sexton R Jr. Implementation of breast tomosynthesis in a routine screening practice: an observational study. AJR Am J Roentgenol 2013;200:1401-8. 22. Niklason LT, Christian BT, Niklason LE, et al. Digital tomosynthesis in breast imaging. Radiology 1997;205:399-406. 23. Niklason LT, Kopans DB, Hamberg LM. Digital breast imaging: tomosynthesis and digital subtraction mammography. Breast Dis 1998;10:151-64. 24. Breast Imaging and Intervention Guidelines - American College of Radiology [Internet]. Available from: http://www.acr.org/Quality-Safety/ Standards-Guidelines/Practice-Guidelines-by-Modality/Breast-Imaging. Accessed Oct 1, 2013. 25. SurveyMonkey. SurveyMonkey: Free online survey software & questionnaire tool [Internet]. Available from: http://www.surveymonkey. com/. Accessed August 14, 2013. 26. Society of Breast Imaging: About SBI [Internet]. Available from: http://www.sbi-online.org/displaycommon.cfm?an¼1&subarticlenbr¼ 9#Mission. Accessed September 5, 2013. 27. Fisher RA. On the interpretation of c2 from contingency tables, and the calculation of P. J R Stat Soc 1922;85(1):87-94. 28. Jr DWH, Lemeshow S, Sturdivant RX. Applied Logistic Regression (Wiley Series in Probability and Statistics). 3 edition. Wiley; 2013. 29. Team RC. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria; 2013. Available from: http://www.R-project.org/ 30. Inc SI. SAS 9.3 Software, Version 9.3 [Internet]. Cary, NC; 2012. Available from: http://www.sas.com/software/sas9/ 31. Press Announcements.FDA approves first 3-D mammography imaging system [Internet]. Available from: http://www.fda.gov/NewsEvents/ Newsroom/PressAnnouncements/ucm243072.htm. Accessed September 5, 2013. 32. Cavagnetto F, Taccini G, Rosasco R, Bampi R, Calabrese M, Tagliafico A. “In Vivo” average glandular dose evaluation: one-to-one comparison between digital breast tomosynthesis and full-field digital mammography. Radiat Prot Dosimetry 2013;157:53-61. 33. Feng SSJ, Sechopoulos I. Clinical digital breast tomosynthesis system: dosimetric characterization. Radiology 2012;263:35-42.
