LETTERS
Annals of Internal Medicine COMMENTS AND RESPONSES Supplemental Ultrasonography Screening for Women With Dense Breasts TO THE EDITOR: We read Sprague and colleagues' (1) article
with great interest. Legislation about breast density notification is motivated by the low sensitivity of mammography in women with dense breasts. Women diagnosed with advanced-stage cancer after adhering to annual screening commonly wonder how their disease was not identified during screening. Models depend on the values of input parameters. For example, Sprague and colleagues used a lenient definition of “sensitivity” for mammography; results were considered to be false-negative only when cancer was detected clinically within 12 months after a screening mammography. However, many cases of late-stage cancer are detected on mammography after being masked on earlier screening rounds. Screening failures are more appropriately any diagnoses of advanced disease. When supplemental magnetic resonance imaging screening is used, mammographic sensitivity averages less than 40% (2), far lower than the estimates used by Sprague and colleagues. Overestimating mammographic sensitivity artificially reduces the potential contribution of any supplemental screening. The sensitivity of mammography in the ACRIN 6666 (American College of Radiology Imaging Network 6666) trial was 52% (59 of 111), and ultrasonography was not limited to women with negative results on mammography (3). Sensitivity for invasive cancer was 46% (41 of 89) and 60% (53 of 89) for mammography and ultrasonography, respectively, with 34% (30 of 89) of cases of invasive cancer seen only on ultrasonography. Parallel sensitivity inputs should be used for all considered methods. The combination of mammography and ultrasonography produces a low interval cancer rate of 1.1 to 1.2 per 1000 women (3, 4). Although supplemental ultrasonography nearly doubles detection of invasive cancer compared with mammography alone, the modeled-estimated life savings for annual ultrasonography was only 0.43 per 1000 in women with dense breasts compared with 10.2 per 1000 for annual mammography alone in women aged 40 to 74 years (1). False-positive results for any screening method decrease with subsequent screening rounds. Sprague and colleagues (1) estimated a 6% false-positive biopsy rate for screening ultrasonography. The observed false-positive biopsy rate was 4.5% in the ACRIN 6666 trial in years 2 and 3 (3). Weigert similarly reported a 3.3% (135 of 4128) false-positive biopsy rate in year 3 and only 1.3% (42 of 3330) in year 4 while maintaining an additional cancer detection rate due to ultrasonography of 3.0 per 1000 examinations (5). In summary, the modeling of Sprague and colleagues (1) likely substantially underestimates the benefit and overestimates the harms and costs of supplemental screening ultrasonography, in particular during repeated screening and with increasing experience. A reduction in the detection of advanced disease is the primary goal of screening. Ultrasonography is effective in that role. Of note, only 11 of 111 women
(9.9%) in the ACRIN 6666 trial were diagnosed with stage IIB or higher disease. Further reduction of avoidable falsepositive results remains a desirable goal for any screening method. Wendie A. Berg, MD, PhD David Gur, ScD University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L15-0182. References 1. Sprague BL, Stout NK, Schechter C, van Ravesteyn NT, Cevik M, Alagoz O, et al. Benefits, harms, and cost-effectiveness of supplemental ultrasonography screening for women with dense breasts. Ann Intern Med. 2015;162:157-66. [PMID: 25486550] doi:10.7326/M14-0692 2. Berg WA. Tailored supplemental screening for breast cancer: what now and what next? AJR Am J Roentgenol. 2009;192:390-9. [PMID: 19155400] doi:10 .2214/AJR.08.1706 3. Berg WA, Zhang Z, Lehrer D, Jong RA, Pisano ED, Barr RG, et al; ACRIN 6666 Investigators. Detection of breast cancer with addition of annual screening ultrasound or a single screening MRI to mammography in women with elevated breast cancer risk. JAMA. 2012;307:1394-404. [PMID: 22474203] doi: 10.1001/jama.2012.388 4. Corsetti V, Houssami N, Ghirardi M, Ferrari A, Speziani M, Bellarosa S, et al. Evidence of the effect of adjunct ultrasound screening in women with mammography-negative dense breasts: interval breast cancers at 1 year follow-up. Eur J Cancer. 2011;47:1021-6. [PMID: 21211962] doi:10.1016/j.ejca .2010.12.002 5. Weigert JM. The Connecticut experiment continues: ultrasound in the screening of women with dense breasts years 3 and 4 [Abstract]. Presented at Radiologic Society of North America 2014, Chicago, Illinois, 30 November–5 December 2014. Abstract no. SSA01-03.
TO THE EDITOR: Sprague and colleagues (1) nicely describe
the limited benefit of screening breast ultrasonography and the high cost and harm of false-positive results leading to unnecessary biopsies. However, they did not mention the most serious harm of screening: overdiagnosis (2, 3) leading to unnecessary treatment of cases of indolent cancer that would never become clinically apparent. Our experience in Connecticut since enactment of the Breast Density Insurance and Notification Law of 2009 leads us to believe that screening breast ultrasonography greatly increases the risk for overdiagnosis. A total of 46% (17 of 37) of cases of breast cancer treated at Yale that were diagnosed by screening ultrasonography were histologic grade I, compared with 12% (78 of 649) diagnosed by self-examination (P < 0.001) and 29% (295 of 1002) diagnosed by screening mammography (P < 0.050). Because fewer than 5% of all breast cancer deaths in the Surveillance, Epidemiology, and End Results Program database are due to grade I cancer, many of these ultrasonography-detected cases probably represent overdiagnosis. The models in Sprague and colleagues' article calculate the number of breast cancer deaths averted by ultrasonography assuming that each case of cancer diagnosed by this method has the same biologic characteristics as cancer diagnosed by selfexamination or mammography. However, the actual reduction in mortality caused by the screening is unknown and may be less than that predicted by the models. It is clear that screen© 2015 American College of Physicians 801
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
LETTERS ing breast ultrasonography can detect cases of cancer not detected by mammography, but whether this is good or bad is unclear.
Reference 1. Sprague BL, Stout NK, Schechter C, van Ravesteyn NT, Cevik M, Alagoz O, et al. Benefits, harms, and cost-effectiveness of supplemental ultrasonography screening for women with dense breasts. Ann Intern Med. 2015;162:157-66.
Brandon Hayse, BS Donald R. Lannin, MD Yale University School of Medicine New Haven, Connecticut Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterest Forms.do?msNum=L15-0183.
References 1. Sprague BL, Stout NK, Schechter C, van Ravesteyn NT, Cevik M, Alagoz O, et al. Benefits, harms, and cost-effectiveness of supplemental ultrasonography screening for women with dense breasts. Ann Intern Med. 2015;162:157-66. [PMID: 25486550] doi:10.7326/M14-0692 2. Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med. 2012;367:1998-2005. [PMID: 23171096] doi:10.1056/NEJMoa1206809 3. Miller AB, Wall C, Baines CJ, Sun P, To T, Narod SA. Twenty five year follow-up for breast cancer incidence and mortality of the Canadian National Breast Screening Study: randomised screening trial. BMJ. 2014;348:g366. [PMID: 24519768] doi:10.1136/bmj.g366
TO THE EDITOR: Sprague and colleagues (1) used 3 simulations to assess the benefits, harms, and cost-effectiveness of supplemental ultrasonography screening in women with dense breasts. Each calculation incorporates assumptions that decrease the validity of the findings. The primary analysis involves a hypothetical population of women aged 50 to 74 years, but dense breasts predominantly affect younger women. Extremely dense breasts—the condition for which the authors' models consistently found the greatest benefit—are unusual after menopause. For this reason, among women for whom breast cancer screening might be advisable, the most relevant cohort is those aged 40 to 55 years. By including women up to age 74 years and neglecting those younger than 50 years in the primary analysis, the authors simultaneously diluted the benefit of supplemental screening ultrasonography and exaggerated the costs. In a secondary analysis, the authors considered ultrasonography in women aged 40 to 74 years who were screened annually. This strategy doubles the attributable risks and costs compared with models for biennial screening while maintaining the larger, dilutive age bracket of their primary analysis. Perhaps a more useful set of models would evaluate the number of tumors found with ultrasonography, the expenses of early detection versus treatment of later disease, and the quality-adjusted life-years saved by adding ultrasonography screening for women with dense breasts aged 40 to 55 years.
Elaine J. Schattner, MD Weill Cornell Medical College New York, New York Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L15-0184.
[PMID: 25486550] doi:10.7326/M14-0692
IN RESPONSE: Drs. Berg and Gur state that we overestimated the sensitivity of mammography screening and underestimated the specificity of ultrasonography screening in our study. We used the definition of sensitivity published by the American College of Radiology Breast Imaging Reporting and Data System (1), which declares that a result is false-negative when “there is a tissue diagnosis of cancer within one year of a negative examination.” We used Breast Cancer Surveillance Consortium data on the sensitivity of digital mammography among women with dense breasts who are not otherwise at high risk to inform our comparative effectiveness study of supplemental ultrasonography for the general population of women with dense breasts (2). Our models were calibrated such that the sensitivity of digital mammography in simulated scenarios with no supplemental ultrasonography screening matched the observed sensitivity of digital mammography in community practice. Low estimates of mammography sensitivity have been reported when supplemental ultrasonography or magnetic resonance imaging is used in high-risk populations, such as in the ACRIN 6666 trial. Indeed, more than one half of women in that trial had a personal history of breast cancer, among whom the sensitivity of mammography to detect subsequent breast cancer is lower than in the general population (3). It also included a mix of film and digital mammography, resulting in lower sensitivity because film is less effective than digital mammography in women with dense breasts (4). Although the ACRIN 6666 trial suggests that specificity may be improved in subsequent screening rounds, data remain sparse about the specificity of ultrasonography screening in community practice. In sensitivity analyses, we found that— even if ultrasonography screening achieved 98% specificity—this supplemental screening for women with dense breasts who have digital mammography would still not achieve reasonable comparative effectiveness benchmarks. We agree with Mr. Hayse and Dr. Lannin that overdiagnosis is an important harm to consider because the introduction of any supplemental screening method may increase overdiagnosis. Given the already unfavorable results for supplemental ultrasonography screening and the complexity of estimating overdiagnosis, we chose not to address this harm in our study. Hence, our study likely underestimates the harms of supplemental ultrasonography screening. We agree with Dr. Schattner that alternative supplemental ultrasonography screening regimens that more specifically target women for whom the benefit– harm profile will be improved should be considered. Alternative regimens likely should consider risk for cancer and for masking on mammography rather than simply density or age. Strategies that focus on women aged 40 to 55 years will result in low cancer detection rates and elevated harms due to false-positive results because of the generally low breast cancer incidence in this age group. Furthermore, the prevalence of dense breasts as defined by state laws as heterogeneously or extremely dense
802 Annals of Internal Medicine • Vol. 162 No. 11 • 2 June 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
www.annals.org
LETTERS exceeds 50% at ages 40 to 49 years and remains greater than 25% up to age 74 years (5). Finally, our study does not address the use of diagnostic ultrasonography after inconclusive mammography as Dr. Schattner describes. Brian L. Sprague, PhD University of Vermont Burlington, Vermont Constance D. Lehman, MD, PhD University of Washington School of Medicine Seattle, Washington Anna N.A. Tosteson, ScD Giesel School of Medicine at Dartmouth Lebanon, New Hampshire Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=M14-0692. References 1. Sickles EA, D’Orsi CJ, Bassett LW, Appleton CA, Berg WA, Burnside ES, et al. ACR BI-RADS Mammography. In: ACR BI-RADS Atlas, Breast Imaging Reporting and Data System. Reston, VA: American College of Radiology; 2013. 2. Stout NK, Lee SJ, Schechter CB, Kerlikowske K, Alagoz O, Berry D, et al. Benefits, harms, and costs for breast cancer screening after US implementation of digital mammography. J Natl Cancer Inst. 2014;106:dju092. [PMID: 24872543] doi:10.1093/jnci/dju092 3. Houssami N, Abraham LA, Miglioretti DL, Sickles EA, Kerlikowske K, Buist DS, et al. Accuracy and outcomes of screening mammography in women with a personal history of early-stage breast cancer. JAMA. 2011;305:790-9. [PMID: 21343578] doi:10.1001/jama.2011.188 4. Kerlikowske K, Hubbard RA, Miglioretti DL, Geller BM, Yankaskas BC, Lehman CD, et al; 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. 2011;155:493-502. [PMID: 22007043] doi:10.7326/0003-4819-155-8-201110180-00005 5. Sprague BL, Gangnon RE, Burt V, Trentham-Dietz A, Hampton JM, Wellman RD, et al. Prevalence of mammographically dense breasts in the United States. J Natl Cancer Inst. 2014;106. [PMID: 25217577] doi:10.1093/jnci/dju255
Can the United States Buy Better Advance Care Planning? TO THE EDITOR: We found Halpern and Emanuel's article (1) provocative but not representative of the recommendations of the recent Institute of Medicine report, “Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life” (2). Halpern and Emanuel state that “one of the [Institute of Medicine] Committee's central recommendations, to reimburse physicians for engaging patients in advance care planning, merits careful consideration.” To clarify, the specific recommendations around clinician–patient communication and advance care planning focus on establishing quality standards for these concepts that are measurable, actionable, and evidence-based and recommend tying “such standards to reimbursement” (2). Improving advance care planning and end-of-life communication is complex and cannot be addressed with a single central recommendation. Several changes must occur synergistically to move the needle on this issue. The 507-page rewww.annals.org
port recommends a multidimensional approach to improving advance care planning, including enhancements in professional training and education, system and policy change, and public engagement. It defines advance care planning as a process of planning and discussion that occurs over time, recognizes the limitations of the living will, and distinguishes advance directives from actionable medical orders. We direct readers to the life-cycle model of advance care planning presented in the report (2), which emphasizes a wellness–illness approach toward planning and normalizing conversations around end-of-life care. Such a model has been developed and implemented in New York (3–5). Rather than acute care– based reactive models, advance care planning should occur long before the patient arrives in the intensive care unit, a time when many patients do not have the capacity to make medical decisions. Clinicians in upstate New York identified inadequate reimbursement as a barrier to end-of-life discussions. In response, Excellus BlueCross BlueShield developed a comprehensive reimbursement model for thoughtful discussions with seriously ill persons who might die in the next year coupled with wellinformed, shared decision making which has led to completion of New York's Medical Orders for Life-Sustaining Treatment (MOLST). This reimbursement model includes face-to-face time with the patient, discussions with the patient's family that may not take place in person, and care planning to support MOLST orders. Eligible clinicians (physicians, nurse practitioners, and physician assistants) are trained to conduct these discussions and enrolled in electronic MOLST (eMOLST), New York's electronic form and discussion completion system that serves as the state's registry. Use of eMOLST ensures that all clinicians who access this electronic registry in an emergency retrieve MOLST and documentation of the discussion, ensuring end-of-life care for patients and families that is consistent with their preferences. Patricia A. Bomba, MD Excellus BlueCross BlueShield Rochester, New York Salimah H. Meghani, PhD, MBE, RN University of Pennsylvania Philadelphia, Pennsylvania Disclosures: Drs. Bomba and Meghani served on the Institute of Medicine Committee on “Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life.” Authors have disclosed no conflicts of interest. Forms can be viewed at www .acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum =L15-0135. References 1. Halpern SD, Emanuel EJ. Can the United States buy better advance care planning? Ann Intern Med. 2015;162:224-5. [PMID: 25486099] doi:10.7326 /M14-2476 2. Institute of Medicine. Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. Washington, DC: National Academies Pr; 2014. 3. Belluck P. Coverage for end-of-life talks gaining ground. The New York Times. 30 August 2014. Accessed at www.nytimes.com/2014/08/31/health /end-of-life-talks-may-finally-overcome-politics.html on 2 December 2014. Annals of Internal Medicine • Vol. 162 No. 11 • 2 June 2015 803
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
LETTERS 4. Bomba PA. Landmark legislation in New York affirms benefits of a two-step approach to advance care planning including MOLST: a model of shared, informed medical decision-making and honoring patient preferences for care at the end of life. Widener Law Review. 2011;17:475-500. 5. Compassion and Support at the End of Life. MOLST provider training: beyond the health care proxy: advance care planning for patients with serious illness. 2009. Accessed at www.compassionandsupport.org/index.php/for _professionals/molst_training_center/provider_training on 2 December 2014.
The Doctor: For Life and at the End of Life TO THE EDITOR: I share Pizzo's concerns (1) about handing
off patients and, in a sense, wiping our hands free as these patients enter a different phase of their life with their illness. I am a medical oncologist and have noticed a change in the past few years where the hospice programs seem to pressure me to hand the care off to the hospice palliative care physician. I have always told patients that hospice will be my eyes and ears, but I will still be the captain of the ship. I now feel like this is a lie. Hospice services in my area do not seem to want me involved anymore, most certainly because of changes in reimbursement and the desire to control the utilization of palliative care. I am concerned that my training that stressed continuity has been abandoned. We now hand patients off to hospitalists when they are sick and to hospice when they are dying—2 times when it seems that continuity would be most important. Joseph Kash, MD Edward Hospital Naperville, Illinois
Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterest Forms.do?msNum=L15-0136. Reference 1. Pizzo PA. The Doctor: for life and at the end of life. Ann Intern Med. 2015; 162:228-9. [PMID: 25486216] doi:10.7326/M14-2399
TO THE EDITOR: Most persons in the United States will now die of multiple chronic conditions, with the timing of death being unclear until the last few days. Persons alive at 70 years of age face nearly 3 years of self-care disability in the last phase of their lives, with a series of private arrangements and health care supports. It seems that “palliative care” should not be mostly the province of specialists but a highly valued skill of virtually all physicians. Furthermore, it seems that Pizzo's comments (1) on the last part of life should focus more on nursing homes, home care, long-term care financing, continuity for persons living with disabilities associated with aging, and related topics. The Institute of Medicine report actually does address these issues, but they have not been discussed in the professional or public press.
Joanne Lynn, MD, MA, MS Altarum Institute, Center for Elder Care and Advanced Illness Washington, DC Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterest Forms.do?msNum=L15-0137. Reference 1. Pizzo PA. The Doctor: for life and at the end of life. Ann Intern Med. 2015; 162:228-9. [PMID: 25486216] doi:10.7326/M14-2399
804 Annals of Internal Medicine • Vol. 162 No. 11 • 2 June 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
www.annals.org
Annals of Internal Medicine COMMENTS AND RESPONSES Supplemental Ultrasonography Screening for Women With Dense Breasts TO THE EDITOR: We read Sprague and colleagues' (1) article
with great interest. Legislation about breast density notification is motivated by the low sensitivity of mammography in women with dense breasts. Women diagnosed with advanced-stage cancer after adhering to annual screening commonly wonder how their disease was not identified during screening. Models depend on the values of input parameters. For example, Sprague and colleagues used a lenient definition of “sensitivity” for mammography; results were considered to be false-negative only when cancer was detected clinically within 12 months after a screening mammography. However, many cases of late-stage cancer are detected on mammography after being masked on earlier screening rounds. Screening failures are more appropriately any diagnoses of advanced disease. When supplemental magnetic resonance imaging screening is used, mammographic sensitivity averages less than 40% (2), far lower than the estimates used by Sprague and colleagues. Overestimating mammographic sensitivity artificially reduces the potential contribution of any supplemental screening. The sensitivity of mammography in the ACRIN 6666 (American College of Radiology Imaging Network 6666) trial was 52% (59 of 111), and ultrasonography was not limited to women with negative results on mammography (3). Sensitivity for invasive cancer was 46% (41 of 89) and 60% (53 of 89) for mammography and ultrasonography, respectively, with 34% (30 of 89) of cases of invasive cancer seen only on ultrasonography. Parallel sensitivity inputs should be used for all considered methods. The combination of mammography and ultrasonography produces a low interval cancer rate of 1.1 to 1.2 per 1000 women (3, 4). Although supplemental ultrasonography nearly doubles detection of invasive cancer compared with mammography alone, the modeled-estimated life savings for annual ultrasonography was only 0.43 per 1000 in women with dense breasts compared with 10.2 per 1000 for annual mammography alone in women aged 40 to 74 years (1). False-positive results for any screening method decrease with subsequent screening rounds. Sprague and colleagues (1) estimated a 6% false-positive biopsy rate for screening ultrasonography. The observed false-positive biopsy rate was 4.5% in the ACRIN 6666 trial in years 2 and 3 (3). Weigert similarly reported a 3.3% (135 of 4128) false-positive biopsy rate in year 3 and only 1.3% (42 of 3330) in year 4 while maintaining an additional cancer detection rate due to ultrasonography of 3.0 per 1000 examinations (5). In summary, the modeling of Sprague and colleagues (1) likely substantially underestimates the benefit and overestimates the harms and costs of supplemental screening ultrasonography, in particular during repeated screening and with increasing experience. A reduction in the detection of advanced disease is the primary goal of screening. Ultrasonography is effective in that role. Of note, only 11 of 111 women
(9.9%) in the ACRIN 6666 trial were diagnosed with stage IIB or higher disease. Further reduction of avoidable falsepositive results remains a desirable goal for any screening method. Wendie A. Berg, MD, PhD David Gur, ScD University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L15-0182. References 1. Sprague BL, Stout NK, Schechter C, van Ravesteyn NT, Cevik M, Alagoz O, et al. Benefits, harms, and cost-effectiveness of supplemental ultrasonography screening for women with dense breasts. Ann Intern Med. 2015;162:157-66. [PMID: 25486550] doi:10.7326/M14-0692 2. Berg WA. Tailored supplemental screening for breast cancer: what now and what next? AJR Am J Roentgenol. 2009;192:390-9. [PMID: 19155400] doi:10 .2214/AJR.08.1706 3. Berg WA, Zhang Z, Lehrer D, Jong RA, Pisano ED, Barr RG, et al; ACRIN 6666 Investigators. Detection of breast cancer with addition of annual screening ultrasound or a single screening MRI to mammography in women with elevated breast cancer risk. JAMA. 2012;307:1394-404. [PMID: 22474203] doi: 10.1001/jama.2012.388 4. Corsetti V, Houssami N, Ghirardi M, Ferrari A, Speziani M, Bellarosa S, et al. Evidence of the effect of adjunct ultrasound screening in women with mammography-negative dense breasts: interval breast cancers at 1 year follow-up. Eur J Cancer. 2011;47:1021-6. [PMID: 21211962] doi:10.1016/j.ejca .2010.12.002 5. Weigert JM. The Connecticut experiment continues: ultrasound in the screening of women with dense breasts years 3 and 4 [Abstract]. Presented at Radiologic Society of North America 2014, Chicago, Illinois, 30 November–5 December 2014. Abstract no. SSA01-03.
TO THE EDITOR: Sprague and colleagues (1) nicely describe
the limited benefit of screening breast ultrasonography and the high cost and harm of false-positive results leading to unnecessary biopsies. However, they did not mention the most serious harm of screening: overdiagnosis (2, 3) leading to unnecessary treatment of cases of indolent cancer that would never become clinically apparent. Our experience in Connecticut since enactment of the Breast Density Insurance and Notification Law of 2009 leads us to believe that screening breast ultrasonography greatly increases the risk for overdiagnosis. A total of 46% (17 of 37) of cases of breast cancer treated at Yale that were diagnosed by screening ultrasonography were histologic grade I, compared with 12% (78 of 649) diagnosed by self-examination (P < 0.001) and 29% (295 of 1002) diagnosed by screening mammography (P < 0.050). Because fewer than 5% of all breast cancer deaths in the Surveillance, Epidemiology, and End Results Program database are due to grade I cancer, many of these ultrasonography-detected cases probably represent overdiagnosis. The models in Sprague and colleagues' article calculate the number of breast cancer deaths averted by ultrasonography assuming that each case of cancer diagnosed by this method has the same biologic characteristics as cancer diagnosed by selfexamination or mammography. However, the actual reduction in mortality caused by the screening is unknown and may be less than that predicted by the models. It is clear that screen© 2015 American College of Physicians 801
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
LETTERS ing breast ultrasonography can detect cases of cancer not detected by mammography, but whether this is good or bad is unclear.
Reference 1. Sprague BL, Stout NK, Schechter C, van Ravesteyn NT, Cevik M, Alagoz O, et al. Benefits, harms, and cost-effectiveness of supplemental ultrasonography screening for women with dense breasts. Ann Intern Med. 2015;162:157-66.
Brandon Hayse, BS Donald R. Lannin, MD Yale University School of Medicine New Haven, Connecticut Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterest Forms.do?msNum=L15-0183.
References 1. Sprague BL, Stout NK, Schechter C, van Ravesteyn NT, Cevik M, Alagoz O, et al. Benefits, harms, and cost-effectiveness of supplemental ultrasonography screening for women with dense breasts. Ann Intern Med. 2015;162:157-66. [PMID: 25486550] doi:10.7326/M14-0692 2. Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med. 2012;367:1998-2005. [PMID: 23171096] doi:10.1056/NEJMoa1206809 3. Miller AB, Wall C, Baines CJ, Sun P, To T, Narod SA. Twenty five year follow-up for breast cancer incidence and mortality of the Canadian National Breast Screening Study: randomised screening trial. BMJ. 2014;348:g366. [PMID: 24519768] doi:10.1136/bmj.g366
TO THE EDITOR: Sprague and colleagues (1) used 3 simulations to assess the benefits, harms, and cost-effectiveness of supplemental ultrasonography screening in women with dense breasts. Each calculation incorporates assumptions that decrease the validity of the findings. The primary analysis involves a hypothetical population of women aged 50 to 74 years, but dense breasts predominantly affect younger women. Extremely dense breasts—the condition for which the authors' models consistently found the greatest benefit—are unusual after menopause. For this reason, among women for whom breast cancer screening might be advisable, the most relevant cohort is those aged 40 to 55 years. By including women up to age 74 years and neglecting those younger than 50 years in the primary analysis, the authors simultaneously diluted the benefit of supplemental screening ultrasonography and exaggerated the costs. In a secondary analysis, the authors considered ultrasonography in women aged 40 to 74 years who were screened annually. This strategy doubles the attributable risks and costs compared with models for biennial screening while maintaining the larger, dilutive age bracket of their primary analysis. Perhaps a more useful set of models would evaluate the number of tumors found with ultrasonography, the expenses of early detection versus treatment of later disease, and the quality-adjusted life-years saved by adding ultrasonography screening for women with dense breasts aged 40 to 55 years.
Elaine J. Schattner, MD Weill Cornell Medical College New York, New York Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L15-0184.
[PMID: 25486550] doi:10.7326/M14-0692
IN RESPONSE: Drs. Berg and Gur state that we overestimated the sensitivity of mammography screening and underestimated the specificity of ultrasonography screening in our study. We used the definition of sensitivity published by the American College of Radiology Breast Imaging Reporting and Data System (1), which declares that a result is false-negative when “there is a tissue diagnosis of cancer within one year of a negative examination.” We used Breast Cancer Surveillance Consortium data on the sensitivity of digital mammography among women with dense breasts who are not otherwise at high risk to inform our comparative effectiveness study of supplemental ultrasonography for the general population of women with dense breasts (2). Our models were calibrated such that the sensitivity of digital mammography in simulated scenarios with no supplemental ultrasonography screening matched the observed sensitivity of digital mammography in community practice. Low estimates of mammography sensitivity have been reported when supplemental ultrasonography or magnetic resonance imaging is used in high-risk populations, such as in the ACRIN 6666 trial. Indeed, more than one half of women in that trial had a personal history of breast cancer, among whom the sensitivity of mammography to detect subsequent breast cancer is lower than in the general population (3). It also included a mix of film and digital mammography, resulting in lower sensitivity because film is less effective than digital mammography in women with dense breasts (4). Although the ACRIN 6666 trial suggests that specificity may be improved in subsequent screening rounds, data remain sparse about the specificity of ultrasonography screening in community practice. In sensitivity analyses, we found that— even if ultrasonography screening achieved 98% specificity—this supplemental screening for women with dense breasts who have digital mammography would still not achieve reasonable comparative effectiveness benchmarks. We agree with Mr. Hayse and Dr. Lannin that overdiagnosis is an important harm to consider because the introduction of any supplemental screening method may increase overdiagnosis. Given the already unfavorable results for supplemental ultrasonography screening and the complexity of estimating overdiagnosis, we chose not to address this harm in our study. Hence, our study likely underestimates the harms of supplemental ultrasonography screening. We agree with Dr. Schattner that alternative supplemental ultrasonography screening regimens that more specifically target women for whom the benefit– harm profile will be improved should be considered. Alternative regimens likely should consider risk for cancer and for masking on mammography rather than simply density or age. Strategies that focus on women aged 40 to 55 years will result in low cancer detection rates and elevated harms due to false-positive results because of the generally low breast cancer incidence in this age group. Furthermore, the prevalence of dense breasts as defined by state laws as heterogeneously or extremely dense
802 Annals of Internal Medicine • Vol. 162 No. 11 • 2 June 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
www.annals.org
LETTERS exceeds 50% at ages 40 to 49 years and remains greater than 25% up to age 74 years (5). Finally, our study does not address the use of diagnostic ultrasonography after inconclusive mammography as Dr. Schattner describes. Brian L. Sprague, PhD University of Vermont Burlington, Vermont Constance D. Lehman, MD, PhD University of Washington School of Medicine Seattle, Washington Anna N.A. Tosteson, ScD Giesel School of Medicine at Dartmouth Lebanon, New Hampshire Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=M14-0692. References 1. Sickles EA, D’Orsi CJ, Bassett LW, Appleton CA, Berg WA, Burnside ES, et al. ACR BI-RADS Mammography. In: ACR BI-RADS Atlas, Breast Imaging Reporting and Data System. Reston, VA: American College of Radiology; 2013. 2. Stout NK, Lee SJ, Schechter CB, Kerlikowske K, Alagoz O, Berry D, et al. Benefits, harms, and costs for breast cancer screening after US implementation of digital mammography. J Natl Cancer Inst. 2014;106:dju092. [PMID: 24872543] doi:10.1093/jnci/dju092 3. Houssami N, Abraham LA, Miglioretti DL, Sickles EA, Kerlikowske K, Buist DS, et al. Accuracy and outcomes of screening mammography in women with a personal history of early-stage breast cancer. JAMA. 2011;305:790-9. [PMID: 21343578] doi:10.1001/jama.2011.188 4. Kerlikowske K, Hubbard RA, Miglioretti DL, Geller BM, Yankaskas BC, Lehman CD, et al; 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. 2011;155:493-502. [PMID: 22007043] doi:10.7326/0003-4819-155-8-201110180-00005 5. Sprague BL, Gangnon RE, Burt V, Trentham-Dietz A, Hampton JM, Wellman RD, et al. Prevalence of mammographically dense breasts in the United States. J Natl Cancer Inst. 2014;106. [PMID: 25217577] doi:10.1093/jnci/dju255
Can the United States Buy Better Advance Care Planning? TO THE EDITOR: We found Halpern and Emanuel's article (1) provocative but not representative of the recommendations of the recent Institute of Medicine report, “Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life” (2). Halpern and Emanuel state that “one of the [Institute of Medicine] Committee's central recommendations, to reimburse physicians for engaging patients in advance care planning, merits careful consideration.” To clarify, the specific recommendations around clinician–patient communication and advance care planning focus on establishing quality standards for these concepts that are measurable, actionable, and evidence-based and recommend tying “such standards to reimbursement” (2). Improving advance care planning and end-of-life communication is complex and cannot be addressed with a single central recommendation. Several changes must occur synergistically to move the needle on this issue. The 507-page rewww.annals.org
port recommends a multidimensional approach to improving advance care planning, including enhancements in professional training and education, system and policy change, and public engagement. It defines advance care planning as a process of planning and discussion that occurs over time, recognizes the limitations of the living will, and distinguishes advance directives from actionable medical orders. We direct readers to the life-cycle model of advance care planning presented in the report (2), which emphasizes a wellness–illness approach toward planning and normalizing conversations around end-of-life care. Such a model has been developed and implemented in New York (3–5). Rather than acute care– based reactive models, advance care planning should occur long before the patient arrives in the intensive care unit, a time when many patients do not have the capacity to make medical decisions. Clinicians in upstate New York identified inadequate reimbursement as a barrier to end-of-life discussions. In response, Excellus BlueCross BlueShield developed a comprehensive reimbursement model for thoughtful discussions with seriously ill persons who might die in the next year coupled with wellinformed, shared decision making which has led to completion of New York's Medical Orders for Life-Sustaining Treatment (MOLST). This reimbursement model includes face-to-face time with the patient, discussions with the patient's family that may not take place in person, and care planning to support MOLST orders. Eligible clinicians (physicians, nurse practitioners, and physician assistants) are trained to conduct these discussions and enrolled in electronic MOLST (eMOLST), New York's electronic form and discussion completion system that serves as the state's registry. Use of eMOLST ensures that all clinicians who access this electronic registry in an emergency retrieve MOLST and documentation of the discussion, ensuring end-of-life care for patients and families that is consistent with their preferences. Patricia A. Bomba, MD Excellus BlueCross BlueShield Rochester, New York Salimah H. Meghani, PhD, MBE, RN University of Pennsylvania Philadelphia, Pennsylvania Disclosures: Drs. Bomba and Meghani served on the Institute of Medicine Committee on “Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life.” Authors have disclosed no conflicts of interest. Forms can be viewed at www .acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum =L15-0135. References 1. Halpern SD, Emanuel EJ. Can the United States buy better advance care planning? Ann Intern Med. 2015;162:224-5. [PMID: 25486099] doi:10.7326 /M14-2476 2. Institute of Medicine. Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. Washington, DC: National Academies Pr; 2014. 3. Belluck P. Coverage for end-of-life talks gaining ground. The New York Times. 30 August 2014. Accessed at www.nytimes.com/2014/08/31/health /end-of-life-talks-may-finally-overcome-politics.html on 2 December 2014. Annals of Internal Medicine • Vol. 162 No. 11 • 2 June 2015 803
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
LETTERS 4. Bomba PA. Landmark legislation in New York affirms benefits of a two-step approach to advance care planning including MOLST: a model of shared, informed medical decision-making and honoring patient preferences for care at the end of life. Widener Law Review. 2011;17:475-500. 5. Compassion and Support at the End of Life. MOLST provider training: beyond the health care proxy: advance care planning for patients with serious illness. 2009. Accessed at www.compassionandsupport.org/index.php/for _professionals/molst_training_center/provider_training on 2 December 2014.
The Doctor: For Life and at the End of Life TO THE EDITOR: I share Pizzo's concerns (1) about handing
off patients and, in a sense, wiping our hands free as these patients enter a different phase of their life with their illness. I am a medical oncologist and have noticed a change in the past few years where the hospice programs seem to pressure me to hand the care off to the hospice palliative care physician. I have always told patients that hospice will be my eyes and ears, but I will still be the captain of the ship. I now feel like this is a lie. Hospice services in my area do not seem to want me involved anymore, most certainly because of changes in reimbursement and the desire to control the utilization of palliative care. I am concerned that my training that stressed continuity has been abandoned. We now hand patients off to hospitalists when they are sick and to hospice when they are dying—2 times when it seems that continuity would be most important. Joseph Kash, MD Edward Hospital Naperville, Illinois
Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterest Forms.do?msNum=L15-0136. Reference 1. Pizzo PA. The Doctor: for life and at the end of life. Ann Intern Med. 2015; 162:228-9. [PMID: 25486216] doi:10.7326/M14-2399
TO THE EDITOR: Most persons in the United States will now die of multiple chronic conditions, with the timing of death being unclear until the last few days. Persons alive at 70 years of age face nearly 3 years of self-care disability in the last phase of their lives, with a series of private arrangements and health care supports. It seems that “palliative care” should not be mostly the province of specialists but a highly valued skill of virtually all physicians. Furthermore, it seems that Pizzo's comments (1) on the last part of life should focus more on nursing homes, home care, long-term care financing, continuity for persons living with disabilities associated with aging, and related topics. The Institute of Medicine report actually does address these issues, but they have not been discussed in the professional or public press.
Joanne Lynn, MD, MA, MS Altarum Institute, Center for Elder Care and Advanced Illness Washington, DC Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterest Forms.do?msNum=L15-0137. Reference 1. Pizzo PA. The Doctor: for life and at the end of life. Ann Intern Med. 2015; 162:228-9. [PMID: 25486216] doi:10.7326/M14-2399
804 Annals of Internal Medicine • Vol. 162 No. 11 • 2 June 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/934098/ by a Universite Laval Biblioteque User on 07/28/2017
www.annals.org
