Abstract
Background
Mammography is limited when analyzing dense breasts for 2 reasons: (1) breast density masks underlying cancers and (2) breast density is an independent risk factor for cancer. We undertook this study to assess whether there is a racial/ethnic difference in supplemental image ordering for women with dense breasts.
Methods
We conducted a retrospective, observational cohort study of women aged 50–75 from an academic medical center who had completed a screening mammogram between 2014 and 2016 that was read as BI-RADS 1 with heterogeneously or extremely dense breasts or BI-RADS 2 with extremely dense breasts. Data were abstracted on type, timing and frequency of supplemental imaging tests ordered within two years of an initial screening mammogram. Patient characteristics (age, race/ethnicity, insurance, and comorbidities) were also abstracted. We used bivariate and multivariate logistic regression to assess for differences in supplemental imaging ordered by race/ethnicity.
Results
Three hundred twenty-six women met inclusion criteria. Mean age was 58 years: 25% were non-Hispanic white, 30% were non-Hispanic black, 27% were Hispanic, 6% were Asian and 14% unknown. Seventy-nine (24%) women were ordered a supplemental breast ultrasound after the initial screening mammogram. Non-Hispanic black and Hispanic women were less likely to have supplemental imaging ordered compared to non-Hispanic white women (15% and 10%, respectively, vs. 45%, p < 0.0001). After controlling for patient age, ordering physician specialty, insurance, BI-RADS score, breast density, and family history of breast cancer, non-Hispanic black and Hispanic women remained less likely to be ordered supplemental imaging (OR 0.38 [95% CI 0.17–0.85] and OR 0.24 [95% CI 0.10–0.61], respectively, p < 0.0001).
Conclusion
Minority women with dense breasts are less likely to be ordered supplemental breast imaging. Further research should investigate physician and patient behaviors to determine barriers in supplemental imaging. Understanding these differences may help reduce disparities in breast cancer care and mortality.
Keywords: Cancer screening, Race and ethnicity, Women’s health
Introduction
Breast cancer is the most common non-skin cancer among women in the United States (US) and has the second highest cancer-related mortality rate [1]. Mammograms have allowed for early detection of breast cancer, and biennial screening mammograms for women between the ages of 50 and 74 (as recommended by the US Preventive Services Task Force) has resulted in a steady decrease in breast cancer mortality by 20–30% since the 1990s [2]. However, this decrease in mortality is not universal among all women: non-Hispanic black and Hispanic women are still being diagnosed with more advanced stage cancer and consequently suffer higher mortality rates [3]. Part of this disparity may be due to lower participation in screening mammograms among these minority groups [1].
Although an integral component of breast cancer screening, mammograms are limited in detecting cancer in women with dense breasts [4]. This is due to primarily two reasons: (1) dense breast tissue masks underlying cancers and (2) breast density has been found to be an independent risk factor for breast cancer [5]. Women with dense breasts thus experience higher rates of interval cancers that manifest within a year of a normal mammogram [6]. As of 2019, legislation in 38 states has mandated patient notification of breast density to identify women who may benefit from supplemental imaging following screening mammography [7]. With documented racial/ethnic disparities in breast cancer mortality, some studies have investigated whether there is a corresponding difference in breast density by race; many of these studies have concluded that breast density does not vary by race [8]. We undertook this study to assess whether there is a racial/ethnic difference in supplemental image ordering for breast cancer screening in women with dense breasts.
Methods
We conducted a retrospective, observational cohort study of women aged 50–75 from an urban academic medical center who completed a screening mammogram between 2014 and 2016 that was read as BI-RADS 1 and either heterogeneously or extremely dense breasts or BI-RADS 2 and extremely dense breasts. Data were abstracted from the electronic heath record on sociodemographic characteristics (age, race/ethnicity, insurance and family history of breast cancer) of women who met inclusion criteria and type and ordering provider for any supplemental imaging tests ordered. Women who reported being of Hispanic ethnicity were classified as Hispanic, regardless of race [9]. We categorized ordering provider type as either primary care (physicians from general internal medicine, family medicine, or geriatrics) or specialty care (physicians from obstetrics/gynecology or surgery). Insurance was dichotomized as public (e.g., Medicaid and/or Medicare) or private (e.g., commercial insurance, self-pay). The outcome variable, supplemental image ordering, was determined by whether there was a physician’s order for a supplementing imaging within 12 months after screening mammogram. The institutional review board at Icahn School of Medicine at Mount Sinai approved this study.
Data analysis
Univariate analyses were conducted to compare baseline sociodemographic characteristics of the study cohort. Chi-square or t-test analyses were used, as appropriate, to assess for differences by race/ethnicity in age, insurance, ordering physician type, BI-RADS category, breast density, family history of breast cancer, and supplemental imaging ordered. Multiple logistic regression was performed to assess whether race/ethnicity was associated with ordering of supplemental imaging, controlling for patient age, ordering physician type, insurance, BI-RADS score, breast density, and family history of breast cancer. We report odds ratios (OR) and 95% confidence intervals (CI) significant at two-tailed, p < 0.05. All statistical analyses were conducted using SAS 9.4.
Results
Three hundred twenty-six women met inclusion criteria (Table 1). Mean age was 58 years: 25% were non-Hispanic white, 30% were non-Hispanic black, 27% were Hispanic, 6% were Asian and 14% unknown. One hundred forty-nine patients (46%) had private insurance and 177 (54%) had public insurance. Overall, 222 (68%) mammograms were read as BI-RADS 1 and 104 (32%) were BI-RADS 2. Of those which were read as BI-RADS 1, 136 (42%) mammograms were read as extremely dense and 190 (58%) as heterogeneously dense. One hundred eleven (34%) women had a family history of breast cancer.
Table 1.
Characteristics of study cohort (n = 326)
| N(%) | |
|---|---|
| Demographics | |
| Age (years), mean (SD) | 57.9(7.1) |
| Race | |
| Non-Hispanic White | 80 (24.5) |
| Non-Hispanic Black | 97 (29.8) |
| Hispanic | 86 (26.4) |
| Asian | 19(5.8) |
| Other | 44(13.5) |
| Clinical factors | |
| BI-RADS score | |
| 1 | 222(68.1) |
| 2 | 104(31.9) |
| Breast density | |
| Extremely dense | 136(41.7) |
| Heterogeneously dense | 190 (58.3) |
| Family history of breast cancer | 111 (34.1) |
| Health system factors | |
| Insurance type | |
| Private | 149 (45.7) |
| Public | 177 (54.3) |
| Ordering physician | |
| Primary care | 244 (74.9) |
| Specialty care | 82 (25.2) |
| Outcome variable | |
| Supplemental imaging ordered | 79 (24.2) |
There were several differences in demographic and clinical characteristics by race/ethnicity (Table 2). Non-Hispanic white women were more likely to have their initial screening mammogram ordered by a physician in specialty care (44% vs. 16% for non-Hispanic blacks and 14% for Hispanics [p < 0.0001]). Thirty percent of non-Hispanic white women had public insurance, compared to 64% of non-Hispanic blacks, 42% of Asians, and 76% of Hispanics (p < 0.0001). Forty-eight (60%) non-Hispanic whites were noted to have extremely dense breasts, compared to 31% of non-Hispanic black women, 58% of Asian women, and 31% of Hispanic women (p < 0.001) (Table 3).
Table 2.
Characteristics of the study cohort by race/ethnicity
| Characteristic, N (%) | Non-Hispanic | Non-Hispanic | Hispanic (N= 86) | Asian (N= 19) | Other (N=44) | Total (N=326) |
|---|---|---|---|---|---|---|
| White (N= 80) | Black(N=97) | |||||
| Age, mean (SD) | 58 | 57 | 57 | 59 | 44 | 56 |
| Private insurance | 56 (70.0) | 35 (36.1) | 21 (24.4) | 11 (57.9) | 26(59.1) | 149 (45.7)**** |
| BI-RADS 2 | 37 (46.3) | 23 (23.7) | 20 (23.3) | 8(42.1) | 16 (36.4) | 104(31.9)** |
| Extremely dense breasts | 48 (60.0) | 30 (30.9) | 27(31.4) | 11 (57.9) | 20 (45.5) | 136 (41.7)*** |
| Family history of breast cancer | 38 (47.5) | 29 (29.9) | 25 (29.1) | 5 (26.3) | 14(31.8) | 111 (34.1) |
| Ordering physician is specialist | 35 (43.8) | 15 (15.5) | 12(13.9) | 4(21.1) | 16 (4.91) | 82 (25.2)**** |
| Supplemental imaging completed | 36 (45.0) | 15(15.5) | 9(10.5) | 5 (26.3) | 14(31.8) | 79 (24.2)**** |
p < 0.05,
p < 0.01,
p < 0.001,
p < 0.0001
Table 3.
Factors associated with ordering of supplemental imaging
| Variables | O.R | C.I | |
|---|---|---|---|
| Age | 0.99 | 0.94 | 1.03 |
| Race (reference = white) | |||
| Non-Hispanic Black | 0.38 | 0.17 | 0.85* |
| Hispanic | 0.24 | 0.10 | 0.61** |
| Asian | 0.57 | 0.17 | 1.90 |
| Other | 0.72 | 0.31 | 1.70 |
| Private insurance | 1.00 | 0.51 | 1.96 |
| BI-RADS 2 | 2.64 | 1.02 | 6.84* |
| Extremely dense | 1.60 | 0.61 | 4.21 |
| Family history of breast cancer | 1.57 | 0.86 | 2.85 |
| Ordering physician is specialist | 2.23 | 1.18 | 4.18* |
p < 0.05,
p < 0.01
Seventy-nine (24%) women were ordered a supplemental breast ultrasound after the initial screening mammogram. Non-Hispanic black and Hispanic women were less likely to have supplemental imaging ordered compared to non-Hispanic white women (15% and 10%, respectively, vs. 45%, p < 0.0001). After controlling for patient age, ordering physician specialty, insurance, BI-RADS score, breast density, and family history of breast cancer, non-Hispanic black and Hispanic women remained less likely to be ordered supplemental imaging (OR 0.38 [95% CI 0.17–0.85] and OR 0.24 [95% CI 0.10–0.61], respectively, p < 0.0001).
Discussion
We found that among women with findings of dense breasts on screening mammography, there was a racial/ethnic difference in rates of supplemental image ordering. Compared to non-Hispanic white women, non-Hispanic black and Hispanic women were a third and a quarter less likely to have had supplemental ultrasound ordered. Understanding factors that may explain this difference in breast cancer screening practices is critical for ensuring equal care for all women.
Racial/ethnic disparities in breast cancer diagnoses and treatment have been widely documented. Ahmed et al., in meta-analysis of 39 studies, determined that non-Hispanic black and Hispanic women exhibited significantly lower utilization of screening mammography compared to white women.1 Furthermore, Elmore et al. found that non-Hispanic black women were half as likely to use screening mammography for cancer detection and more likely to have longer delays between diagnosis and treatment compared to white women (23 days compared to 18 days) [10]. Hoppe et al. also showed that compared to white women, non-Hispanic black women had longer times to first treatment, surgery, chemotherapy, radiation, and endocrine therapy [11]. Due to decreased use of screening mammography and delays in treatment after diagnosis, it is not surprising that non-Hispanic black [12] and Hispanic women [13] experience increased rates of aggressive, later stage diagnoses and poorer prognoses. Overall, though age-standardized incidence rates of breast cancer are lower among non-Hispanic black and Hispanic women compared to non-Hispanic white women, minority women continue to experience higher mortality rates, even when controlling for disease stage and tumor characteristics [12]. Our findings similarly demonstrate there is less aggressive screening for minority women with dense breasts.
The long-term and mortality benefits of supplemental imaging for women with dense breasts are still being studied. There is evidence that supports the benefits of supplemental imaging, particularly in terms of detection of small node-negative invasive cancers and reducing interval cancers in women with dense breasts [14]. For example, ultrasound has been found to be helpful in detecting node-negative invasive cancers not seen on mammography [15], and magnetic resonance imaging (MRI) in detecting ductal carcinoma in situ as well as reducing late stage disease and metastases [14]. Berg et al. investigated outcomes following supplemental ultrasound or MRI to annual mammograms for 2662 women with dense breasts; supplemental imaging in this population was found to increase both cancer detection yield as well as false-positive results [16]. This trade-off of improved cancer detection with increased false-positive rates was further highlighted in the ASTOUND trial, which used adjunct screening with tomosynthesis or ultrasound in 5300 women with mammography-negative dense breasts; adjunct screening detected 29 additional tumors, mostly invasive node-negative cancers, however, caused 64 false-positive screens [17]. Thus, although the benefits and risks of supplemental imaging are still being evaluated, our study highlights the racial/ethnic difference in supplemental screening among women with dense breasts.
There are several factors that may explain our findings. First, it is well understood that systemic factors, such as lower socioeconomic status and lack of health insurance, are often barriers to care and may ultimately lead to both lower utilization of screening mammography as well as use of supplemental imaging [18]. However, even after we controlled for insurance and assessed for supplemental imaging tests that were ordered but not completed, racial/ethnic differences in supplemental imaging remained. Second, it is possible that specialists are more likely to order more tests. We found that specialty physicians were more likely to order supplemental imaging as compared to their generalist counterparts, and in our cohort, non-Hispanic white women were more likely to have specialist physicians order their initial screening mammogram. This may have partially explained the increased supplemental imaging for non-Hispanic white women. However, even after controlling for physician specialty, we found that non-Hispanic black and Hispanic women were still less likely to be ordered supplemental imaging. Third, patient health literacy and care seeking behaviors could have accounted for the observed differences. Though breast density notification laws are becoming more widespread, patient lack of understanding about breast density and subsequent increased risk for breast cancer could contribute to decreased seeking of medical care by the non-Hispanic black and Hispanic populations in our study [7]. Studies have further found that patient pain and embarrassment may lead to generally decreased utilization of screening mammography [18]. Lastly, physicians’ personal biases and prejudices should be considered as well to possibly explain our findings.
Our study has a few limitations. The data were collected from a single urban academic institution so our results may not be generalizable to other institutions or other settings. Furthermore, we were not able to capture whether supplemental images were ordered by physicians outside of our institution, although given that the initial screening mammogram was ordered by physicians within our institution, it is likely that these physicians would be the ones who would order any supplemental imaging. We were also not able to assess whether physicians ordered supplemental imaging for other facilities outside our institution. Finally, we were unable to ascertain whether patients declined further testing after being recommended supplemental imaging by their physicians, as we were only able to capture whether supplemental imaging was ordered.
In conclusion, we found that non-Hispanic black and Hispanic women with dense breasts were less likely to be ordered supplemental imaging after screening mammogram. While we were not able to determine whether this difference was due to a system-level, physician and/or patient-level cause, some of the difference may be due to the lack of current recommendations or guidelines for supplemental imaging for women with dense breasts. Further research should investigate physician and patient behaviors to determine barriers in supplemental imaging for non-Hispanic black and Hispanic women with dense breasts. Understanding these differences may allow for the development of interventions in policy and health care access, physician clinical training, and patient education. These initiatives may ultimately help reduce disparities in breast cancer care and mortality.
Footnotes
Conflict of interest We have no conflicts of interest to disclose and confirm that this manuscript has not been published elsewhere, nor is it under consideration by another journal. This manuscript has also not been previously submitted to Breast Cancer Research and Treatment. All authors have contributed to, reviewed, and approved the enclosed manuscript. Dr. Lin was supported by a National Cancer Institute Cancer Prevention and Control Career Development Award (1K07CA166462–01). The institutional review board at Icahn School of Medicine at Mount Sinai approved this study.
References
- 1.Ahmed AT, Welch BT, Brinjikji W et al. (2017) Racial disparities in screening mammography in the United States: a systematic review and meta-analysis. J Am Coll Radiol 14(2):157–165.e159 [DOI] [PubMed] [Google Scholar]
- 2.Davis C, Cadet TJ, Moore M, Darby K (2017) A comparison of compliance and noncompliance in breast cancer screening among African American women. Health Soc Work 42(3):159–166 [DOI] [PubMed] [Google Scholar]
- 3.Bhargava S, Moen K, Qureshi SA, Hofvind S (2018) Mammographic screening attendance among immigrant and minority women: a systematic review and meta-analysis. Acta Radiol 59(11):1285–1291 [DOI] [PubMed] [Google Scholar]
- 4.Freer PE (2015) Mammographic breast density: impact on breast cancer risk and implications for screening. Radiographics 35(2):302–315 [DOI] [PubMed] [Google Scholar]
- 5.Berg WA (2016) Current status of supplemental screening in dense breasts. Journal Clin Oncol 34(16):1840–1843 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Mandelson MT, Oestreicher N, Porter PL et al. (2000) Breast density as a predictor of mammographic detection: comparison of interval- and screen-detected cancers. J Natl Cancer Inst 92(13):1081–1087 [DOI] [PubMed] [Google Scholar]
- 7.DenseBreast-info, Inc. (2018) Legislation and Regulations—What is Required? https://densebreast-info.org/legislation.aspx.
- 8.El-Bastawissi AY, White E, Mandelson MT, Taplin S (2001) Variation in mammographic breast density by race. Ann Epidemiol 11(4):257–263 [DOI] [PubMed] [Google Scholar]
- 9.Hitlin S, Brown JS, Elder JGH (2007) Measuring Latinos: racial vs ethnic classification and self-understandings. Soc Forces 86(2):587–611 [Google Scholar]
- 10.Elmore JG, Nakano CY, Linden HM, Reisch LM, Ayanian JZ, Larson EB (2005) Racial inequities in the timing of breast cancer detection, diagnosis, and initiation of treatment. Med Care 43(2):141–148 [DOI] [PubMed] [Google Scholar]
- 11.Hoppe EJ, Hussain LR, Grannan KJ, Dunki-Jacobs EM, Lee DY, Wexelman BA (2018) Racial disparities in breast cancer persist despite early detection: analysis of treatment of stage 1 breast cancer and effect of insurance status on disparities. Breast Cancer Res Treat [DOI] [PubMed] [Google Scholar]
- 12.Coughlin SS (2014) Intervention approaches for addressing breast cancer disparities among African American Women. Ann Trans Med Epidemiol 1(1). [PMC free article] [PubMed] [Google Scholar]
- 13.Livaudais JC, Coronado GD, Espinoza N, Islas I, Ibarra G, Thompson B (2010) Educating Hispanic Women about breast cancer prevention: evaluation of a home-based promotora-led intervention. J Women’s Health 19(11):2049–2056 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Berg WA (2016) Supplemental breast cancer screening in women with dense breasts should be offered with simultaneous collection of outcomes data. Ann Intern Med 164(4):299–300 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Berg WA, Blume JD, Cormack JB et al. (2008) Combined screening with ultrasound and mammography vs mammography alone in women at elevated risk of breast cancer. JAMA 299(18):2151–2163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Berg WA, Zhang Z, Lehrer D et al. (2012) 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 307(13):1394–1404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Tagliafico AS, Mariscotti G, Valdora F et al. (2018) A prospective comparative trial of adjunct screening with tomosynthesis or ultrasound in women with mammography-negative dense breasts (ASTOUND-2). Eur J Cancer. 104:39–46 [DOI] [PubMed] [Google Scholar]
- 18.Alexandraki I, Mooradian AD (2010) Barriers related to mammography use for breast cancer screening among minority women. J Natl Med Assoc 102(3):206–218 [DOI] [PubMed] [Google Scholar]