Journal Logo

EQUITY AND SOCIAL JUSTICE

Underrepresentation of Racial Minorities in Breast Surgery Literature

A Call for Increased Diversity and Inclusion

Cho, Daniel Y. MD, PhD; Kneib, Cameron J. MD; Shakir, Afaaf MD; Burns, Jacob R. MD; Lane, Megan MD§; Massie, Jonathan P. MD; Crowe, Christopher S. MD; Sobol, Danielle L. MD; Morrison, Shane D. MD, MS; Sousa, Janelle D. MD; Sabin, Janice PhD, MSW||

Author Information
doi: 10.1097/SLA.0000000000004481
  • Free

Abstract

Breast surgery comprises a significant portion of plastic surgery. By numbers, there are over 700,000 plastic surgery breast procedures performed on men and women annually.1,2 Breast augmentation is the highest volume surgical cosmetic procedure performed overall for women in the United States (329,914 in 2018) and breast reduction for gynecomastia is the second most common procedure in male cosmetic surgery patients.1,2 Additionally, there are nearly 200,000 reconstructive breast procedures performed by board certified plastic surgeons annually.1 Given the high volume of breast surgery procedures, it is a significant area of investigation in plastic surgery literature. The indications for breast surgery are many and include ptosis, macromastia, gynecomastia, congenital anomalies, gender confirmation, cosmesis, and oncologic reconstruction. Although there are no significant racial associations in the incidence and prevalence of most of these conditions, studies have confirmed that patient race is associated with a lower quality of clinical care. Specifically, in breast cancer, disparities have been identified in almost all aspects of patient care such as access to surveillance mammography, stage at diagnosis, genetic testing, duration from diagnosis to treatment, counseling for treatment decision making, surgical complications, access to breast reconstruction, disease free survival, and mortality.3–14

There has been a recent focus on race and equity in medicine with multiple studies demonstrating an underrepresentation of racial minorities in medical education and literature,15–21 including plastic surgery training,22–24 which contribute to inequities in healthcare.25–28 In plastic surgery, there are disparities in the access of minority patients to breast reconstruction,10–12,29,30 timely diagnosis of malignancy,31 replantation in hand injury,32 cleft and craniofacial care,33–35 and involvement in clinical trials.36 Compared to other fields in medicine, plastic surgery still remains behind in studies that evaluate the critical role of race and equity in patient care and outcomes.

The US population is 76.5% White, 13.4% African American, 5.9% Asian, and 4.2% other, of which 18.3% identify as Hispanic or Latino. Ethnic representation for all plastic surgery procedures is 70% White, 13% Hispanic, 10% African American and 6% Asian with some variability based on specific procedure; for example, the demographics for breast reconstruction are 71% White, 13% African America, 11% Hispanic, 4% Asian and 1% other (Table 1). The aim of this study was to use established photogrammetric analysis techniques to evaluate the visual representation of racial diversity in the published breast surgery literature and to understand how it compares to the general and plastic surgery patient population. The goal of this work is to identify disparities in the visual representation of race that may exist in plastic surgery research and to increase awareness of this issue to improve equity and diversity in the field, which can reduce implicit bias to allow for more equitable and just care for all patients.

TABLE 1 - Demographic Information for the Population of the United States41 and the Patients Undergoing Cosmetic and Breast Surgical Procedures1,2
2018 United States Census Bureau Data
White (%) African-American (%) Hispanic (%) Asian (%) Other (%)
US Population 76.5 13.4 18.3 5.9 4.2
2018 ASPS Plastic Surgery Statistics Report
White (%) African-American (%) Hispanic (%) Asian (%) Other (%)
Cosmetic procedures 70 9 11 7 3
Breast augmentation 74 7 11 5 3
Breast reconstruction 71 13 11 4 1
2018 ASAPS Cosmetic Surgery National Data Bank Statistics
White (%) African-American (%) Hispanic (%) Asian (%) Other (%)
Cosmetic procedures 70 9 13 6 1

METHODS

Sample Selection

Leading journals in the field of plastic surgery were selected for analysis including Annals of Plastic Surgery (APS), Aesthetic Surgery Journal (ASJ), Journal of Plastic, Reconstructive, and Aesthetic Surgery (JPRAS), and Plastic and Reconstructive Surgery (PRS). All articles published during the full first calendar year in which photographs and graphics were published in color of each journal as well as 2000, 2010, and 2016 were selected for inclusion (Fig. 1). These years were selected a priori as a means to cover a representative sample from each journal over a given decade (2000 and 2010) as well as the most recent year of publication at the time of study initiation. Because of the initial date each journal started publishing in color, some of the years may be different. Only articles pertaining to aesthetic or reconstructive breast surgery were included in our analysis. The title, corresponding author(s), as well as the countries of origin of the corresponding author(s) were noted for each article.

FIGURE 1
FIGURE 1:
Flow diagram of journals and years selected for analysis with the total number of articles published during the calendar year, articles pertaining to aesthetic or reconstructive breast surgery, and articles containing figures for analysis; first year of color figure publication by journal.

Every published figure was classified as a photograph or graphic, which includes drawings, computer-rendered images, and other nonphotographic depictions. All figures depicting human subjects with visible skin were selected for inclusion; exclusion criteria included images of nonhuman subjects, implants, bone, muscle, fat, nerve, internal organs, or insufficient skin for accurate analysis. Each individual image or illustration in multi-panel figures was considered a separate entity.

Photogrammetric Analysis

Each image and graphic was visually analyzed by authors D.Y.C., C.J.K., A.S., J.R.B., M.L., J.P.M., C.S.C., or D.L.S. using the Fitzpatrick scale as a guide with input from observable phenotypes including hair texture and color, eye color, and facial features following previously established methods.15,18,37–39 To ensure consistency and reduce bias, all figures were analyzed with a standardized approach without consideration of any written descriptions of the images or graphics within the article. Each image was categorized as “White” or “nonWhite,” which included Asian, non-White Hispanic, Native American, African American, and multiracial individuals based on the above criteria. Black and white images, as well as those with poor quality or insufficient phenotypic clues, were noted to be “unidentifiable” and excluded from the final analysis.

Inter-Rater Reliability

To ensure high fidelity data, the variability in the analysis of the images and graphics among the authors was assessed. Twenty-two images representing different skin tones were selected from the sample and were categorized by each author following the method outlined above. All responses were blinded and pooled for analysis using Cohen's kappa (κ) coefficient. Interpretation was based on the Landis and Koch cutoffs for correlation reliability: κ < 0.20 was interpreted as slight agreement, 0.21 to 0.40 as fair agreement, 0.40 to 0.60 as moderate agreement, 0.61 to 0.80 as substantial agreement, and >0.8 as almost perfect agreement.40

Statistical Analysis

The average number of White and non-White images and graphics per article were calculated and pairwise comparisons were made using a 2-tailed unpaired Student t-test. Proportional data of White versus non-White images and graphics were also calculated and compared per article. Univariate regression analyses comparing the average number of White and non-White images over time were performed. In addition, univariate regression of the proportion of non-White images over time was performed. Pearson correlation coefficient (r) was reported for all univariate regression analyses. For all statistical analyses, significance allowed for a type I error of α = 0.05.

RESULTS

A total of 914 published articles with 2774 images and 353 graphics met inclusion criteria based on publication date and subject matter of breast surgery. Interrater reliability for grading skin type was determined to have a κ coefficient of 0.65 (P < 0.001), indicating substantial agreement between raters. All data are summarized in Table 2. There were 392 (42.9%) articles that included figures depicting human skin of which only 35 (3.8%) articles contained images of non-White subjects; 2065 (91.7%) images were of White skin while 184 (8.3%) images were of non-White skin. This corresponded to 2.26 average White photos per article compared to 0.20 non-White photos per article (P < 0.001). There were 5 articles (0.6%) which contained graphics and a total of 9 graphics with non-White skin (6.3%) compared to 133 graphics with Wite skin (93.7%) (0.15 White graphics per article versus 0.01 non-White graphics per article, P < 0.001).

TABLE 2 - Geographic Region of Origin and Distribution of White vs Non-White Representation in Articles, Images, and Graphics
Origin Articles Images Graphics
Journal US (%) International (%) Multinational (%) Figures (%) Non-White Images (%) Non-White Graphics (%) White (%) Non-White (%) White (%) Non-White (%)
APS 59 (66.3) 27 (30.3) 3 (3.4) 40 (44.9) 9 (10.1) 0 (0) 383 (88.3) 51 (11.8) 10 (100) 0 (0)
ASJ 53 (63.1) 31 (36.9) 0 (0) 52 (61.9) 3 (3.6) 0 (0) 487 (95.9) 21 (4.1) 49 (100) 0 (0)
JPRAS 18 (10.3) 151 (86.3) 6 (3.4) 86 (49.1) 16 (9.1) 0 (0) 363 (80.0) 91 (20.0) 4 (100) 0 (0)
PRS 300 (53.0) 234 (41.3) 32 (5.7) 214 (37.8) 7 (1.2) 5 (0.9) 832 (97.2) 24 (2.8) 70 (88.6) 9 (11.4)
TOTAL 430 (47.0) 443 (48.5) 41 (4.5) 392 (42.9) 35 (3.8) 5 (0.6) 2065 (91.7) 187 (8.3) 133 (93.7) 9 (6.3)

Comparison of APS, ASJ, JPRAS, and PRS revealed variable degrees of racial representation (Fig. 2). The greatest diversity was seen in JPRAS, a Europe based journal, where 49.1% (86) of articles in JPRAS contained images and 9.1% (16) of articles contained non-White photos. Of all images analyzed in JPRAS, 20.0% (91) of images were of non-white subjects, whereas all graphics analyzed depicted White skin. A total of 44.9% (40), 61.9% (52), and 37.8% (214) of articles in APS, ASJ, and PRS respectively contained photographs or images. In APS, 10.1% (9) of articles contained non-White photos or graphics with a total of 11.8% (51) of all images showing non-White skin tones and no graphics depicting non-White skin tones. In ASJ, 3.6% (3) of articles contained non-White figures with 4.1% (21) of images showing non-White skin tone and no non-White graphics. The only non-White graphics in the entire series were seen in PRS with a total of 9 graphics (11.4%); however, PRS had the least diversity in images with 2.8% (24) non-White images represented in the 1.2% (7) of articles with non-White figures (Fig. 3). Analysis by average number of images and graphics per article confirmed statistically significant under representation of non-White skin across all journals (P < 0.001).

FIGURE 2
FIGURE 2:
Distribution of country of origin for all published articles (ALL, top) and for each individual journal (bottom) by geographic region (US, white bars; international, black bars; multinational, grey bars).
FIGURE 3
FIGURE 3:
Distribution of white (white bars) vs non-white (black bars) images (A) and graphics (B) in all journals (top) and for each individual journal (bottom).

Temporal analysis revealed that there has been an increase in diversity over time from 0% of articles including non-White photos in the years analyzed before 2010 to 7.3% to 10.3% after 2010 across all the journals included in this study. A similar trend was seen with graphics, which revealed 5.6% and 12.2% non-White graphics in 2010 and 2016, respectively.

Given the small number of articles with non-White figures and graphics, a separate analysis was performed exclusively on these papers. A total of 35 out of 914 (3.8%) articles contained 187 images of non-White skin as well as 195 images of White skin. Of articles containing non-White images, the majority were seen in JPRAS with 45.7% (16) of the articles followed by APS, PRS, and ASJ with 25.7% (9), 20.0% (7), and 8.6% (3), respectively. By article, there was not a statistically significant difference in the frequency of White and non-White images with 5.6 average Wite images per article compared to 5.3 non-White (P = 0.830). Of articles depicting non-White skin, 51.4% of these articles were published from the United States whereas 42.9% were from international research groups and 5.7% represented multi-national collaborations.

A total of 5 articles out of 913 (0.6%) of articles contained non-White graphics. All of these were published in PRS. A total of 9 (75%) non-White graphics and 3 (25%) White graphics were identified in this cohort. There was no statistical significance in frequency seen with 1.8 average non-White graphics per article compared to 0.6 White graphics (P = 0.145). These papers also included 28 (63.6%) White images and 4 (9.1%) non-White images representing a statistically significant under representation of non-White images despite the inclusion of non-White graphics (5.6 average White images per article vs. 0.8 non-White, P = 0.018). All of these articles included US authors, and 2 were part of multi-national collaborations.

DISCUSSION

This study is the first known investigation of racial diversity and inclusion in breast imagery. Overall, 8.3% of images in the breast-related plastic surgery literature represented non-White subjects. The United States Census Bureau data reveals that 23.5% of the U.S. population is non-White, with 18.3% reporting Hispanic ethnicity41 (Table 1). Data from the American Society of Plastic Surgeons in the 2018 Plastic Surgery Statistics Report revealed that 26% of patients undergoing breast augmentation and 29% of patients undergoing breast reconstruction were non-White1 (Table 1). This data is consistent with figures from the American Society of Aesthetic Plastic Surgery Cosmetic (Aesthetic) Surgery National Data Bank Statistics that shows 30% of patients undergoing cosmetic procedures in 2018 were non-White2 (Table 1). Compared to these percentages, there is an over three-fold degree of racial underrepresentation seen in the literature compared to relevant patient populations. This limits the exposure of surgeons in both training and practice to the diversity that exists in the diagnosis, treatment, outcomes, and experiences of breast surgery patients, especially in patient populations that providers may not see frequently. This lack of exposure can result in diagnostic biases, inequitable patient treatment plans, poor understanding of the variability of outcomes with different skin tones, and ultimately inferior clinical outcomes.

Although the data from international and multi-national authors show a greater degree of diversity, the overwhelming overrepresentation of White skin found from American authors demonstrates that there is significant bias, likely unintentional and implicit, in the visual representation found in breast surgery literature. The data is so skewed in the breast surgery literature that there are no clear temporal trends of increasing diversity as has been seen in the general plastic surgery38 and aesthetic surgery literature39 despite the continued increasing diversity of the U.S. and plastic surgery population. It is estimated that by 2044, Whites will be a minority in the United States42 which highlights the urgent need to represent diversity across plastic surgery literature, particularly in breast surgery, to ensure optimal care for all patients.

Uneven racial representation has been recognized as an important contributor to racial inequality in the health care experience, treatment, and understanding of outcomes.18 However, there is a paucity of research on how physicians and surgeons use and interpret imagery, as well as how published visual materials impact the delivery of patient care.43 When the overwhelming majority of images that surgeons encounter is White, the images will create the perception and bias that this is the normative population, perpetuate further racial bias or even racism in medical literature, and has the potential to negatively impact the care of medium and darker skinned individuals.18 Importantly, all of these factors can occur without the awareness of the provider due to implicit bias and a lack of adequate exposure.17 Furthermore, minority patients may perceive this lack of provider knowledge as a lack of expertise or experience with their specific race and as a sign that they are not welcome at certain practices, which can undermine the patient-doctor relationship. This can affect not only the ability of a patient to obtain necessary care but also limit the volume and exposure of plastic surgeons to diverse patient populations.

There are several limitations to this study. The data collection was limited to major plastic surgery journals based in the United States and Europe, which tend to have a higher White population; however, our data demonstrate that there exists statistically significant over-representation of the White population in the literature. The study included select years of analysis due to the sheer volume of images published in plastic surgery journals; however, the data are consistent amongst the time points, which reduces the likelihood of selection bias and supports the generalizability of the results. The use of skin tone and visual cues to identify the subject as being White or non-White is based on previously established methodology.15,18,37 This technique has been modified to enhance accuracy but is limited as the binary classification does not represent the true diversity of subcategories of race (such as African American, Asian, Latino, Native American, Pacific Islander, etc). This analysis does not allow for identification of specific causes for the lack of diversity seen in breast plastic surgery images. However, this does allow for the assessment of a large volume of images to better understand trends in the visual representation of race.

CONCLUSIONS

This study demonstrates the insufficient racial diversity visually represented in breast-related plastic surgery literature across multiple journals. There has been limited progress made towards more equitable imagery over time; the degree of racial representation does not approach representation of the US population or plastic surgery patient population. Additionally, breast surgery literature has significantly less representation than that of general plastic surgery and aesthetic surgery literature. Increasing awareness of image content, existing implicit bias or racism, and the need for equitable visual representation will allow for improved racial diversity in the plastic surgery literature as well as improved care and outcomes for all patients. Increasing diversity representation in the medical literature needs to extend to medical textbooks, course materials, and patient educational guides across all of medicine. As surgeons, we have a responsibility to ensure that our trainees are not only technically but also culturally competent and that all our patients feel welcome and well represented by our specialties.

REFERENCES

1. American Society of Plastic Surgeons. ASPS 2018 Plastic Surgery Statistics Report. ASPS National Clearinghouse of Plastic Surgery Procedural Statistics; 2018.
2. The American Society for Aesthetic Plastic Surgery's Cosmetic Surgery National Data Bank: statistics 2018. Aesthet Surg J 2019; 39:1–27.
3. Hung M-C, Ekwueme DU, Rim SH, et al. Racial/ethnicity disparities in invasive breast cancer among younger and older women: an analysis using multiple measures of population health. Cancer Epidemiol 2016; 45:112–118.
4. Teysir J, Gegechkori N, Wisnivesky JP, et al. Racial disparities in surveillance mammography among older breast cancer survivors. Breast Cancer Res Tr 2019; 176:461–467.
5. Katz SJ, Wallner LP, Abrahamse PH, et al. Treatment experiences of Latinas after diagnosis of breast cancer. Cancer 2017; 123:3022–3030.
6. Doe S, Petersen S, Buekers T, et al. Does a multidisciplinary approach to invasive breast cancer care improve time to treatment and patient compliance? J Natl Med Assoc 2020; 112:268–274.
7. DeSantis CE, Ma J, Sauer AG, et al. Breast cancer statistics, 2017, racial disparity in mortality by state. Ca Cancer J Clin 2017; 67:439–448.
8. White-Means SI, Osmani AR. Affordable care act and disparities in health services utilization among ethnic minority breast cancer survivors: evidence from longitudinal medical expenditure panel surveys 2008-2015. Int J Environ Res Pu 2018; 15:1860.
9. Ren J-X, Gong Y, Ling H, et al. Racial/ethnic differences in the outcomes of patients with metastatic breast cancer: contributions of demographic, socioeconomic, tumor and metastatic characteristics. Breast Cancer Res Tr 2019; 173:225–237.
10. Sharma K, Grant D, Parikh R, et al. Race and breast cancer reconstruction. Plast Reconstr Surg 2016; 138:354–361.
11. Kamali P, Zettervall SL, Wu W, et al. Differences in the reporting of racial and socioeconomic disparities among three large national databases for breast reconstruction. Plast Reconstr Surg 2017; 139:795–807.
12. Mets EJ, Chouairi FK, Gabrick KS, et al. Persistent disparities in breast cancer surgical outcomes among hispanic and African American patients. Eur J Surg Oncol 2019; 45:584–590.
13. Gómez-Trillos S, Sheppard VB, Graves KD, et al. Latinas’ knowledge of and experiences with genetic cancer risk assessment: barriers and facilitators. J Genet Couns 2020; 29:505–517.
14. Yedjou CG, Tchounwou PB, Payton M, et al. Assessing the racial and ethnic disparities in breast cancer mortality in the United States. Int J Environ Res Pu 2017; 14:486.
15. Martin GC, Kirgis J, Sid E, et al. Equitable imagery in the preclinical medical school curriculum. Acad Med 2016; 91:1002–1006.
16. Tsai J, Ucik L, Baldwin N, et al. Race matters?; examining and rethinking race portrayal in preclinical medical education. Acad Med 2016; 91:916–920.
17. Turbes S, Krebs E, Axtell S. The hidden curriculum in multicultural medical education. Acad Med 2002; 77:209–216.
18. Louie P, Wilkes R. Representations of race and skin tone in medical textbook imagery. Soc Sci Med 2018; 202:38–42.
19. Ripp K, Braun L. Race/ethnicity in medical education: an analysis of a question bank for step 1 of the United States medical licensing examination. Teach Learn Med 2017; 29:1–8.
20. Sabin JA, Nosek BA, Greenwald AG, et al. Physicians’ implicit and explicit attitudes about race by MD race, ethnicity, and gender. J Health Care Poor U 2009; 20:896–913.
21. Braun L, Saunders B. Avoiding racial essentialism in medical science curricula. Ama J Ethics 2017; 19:518–527.
22. Rhemtulla IA, Vonderhaar RJ, Mauch JT, et al. Improvement in racial disparity among patients undergoing panniculectomy after bariatric surgery. Am J Surg 2019; 218:37–41.
23. DeSanti RL, Nair L, Adetayo OA. Increasing diversity in plastic surgery. Plast Reconstr Surg 2018; 142:813e–814e.
24. Silvestre J, Serletti JM, Chang B. Racial and ethnic diversity of U.S. plastic surgery trainees. J Surg Educ 2017; 74:117–123.
25. Hall WJ, Chapman MV, Lee KM, et al. Implicit racial/ethnic bias among health care professionals and its influence on health care outcomes: a systematic review. Am J Public Health 2015; 105:e60–e76.
26. Karani R, Varpio L, May W, et al. Commentary: racism and bias in health professions education: how educators, faculty developers, and researchers can make a difference. Acad Med 2017; 92:S1–S6.
27. Metzl JM, Roberts DE. Structural competency meets structural racism: race, politics, and the structure of medical knowledge. Virtual Mentor 2014; 16:674–690.
28. Massie JP, Cho DY, Kneib CJ, et al. It's not just black and white: a call for skin color diversity representation in plastic surgery literature. J Craniofac Surg 2019; 30:2659.
29. Berlin NL, Momoh AO, Qi J, et al. Racial and ethnic variations in one-year clinical and patient-reported outcomes following breast reconstruction. Am J Surg 2017; 214:312–317.
30. Epstein S, Tran BN, Cohen JB, et al. Racial disparities in postmastectomy breast reconstruction: National trends in utilization from 2005 to 2014. Cancer 2018; 124:2774–2784.
31. Hu S, Soza-Vento RM, Parker DF, et al. Comparison of stage at diagnosis of melanoma among Hispanic, Black, and White patients in Miami-Dade County, Florida. Arch Dermatol 2006; 142:704–708.
32. Mahmoudi E, Swiatek PR, Chung KC, et al. Racial variation in treatment of traumatic finger/thumb amputation: a national comparative study of replantation and revision amputation. Plast Reconstr Surg 2016; 137:576e–585e.
33. Sacks GN, Skolnick GB, Trachtenberg A, et al. The impact of ethnicity on craniosynostosis in the United States. J Craniofac Surg 2019; 30:2526–2529.
34. Wu RT, Peck CJ, Shultz BN, et al. Racial disparities in cleft palate repair. Plast Reconstr Surg 2019; 143:1738–1745.
35. Hanba C, Svider PF, Chen FS, et al. Race and sex differences in adult facial fracture risk. JAMA Facial Plast Su 2016; 18:441–448.
36. Silvestre J, Abbatematteo JM, Serletti JM, et al. Racial and ethnic diversity is limited for plastic surgery clinical trials in the United States. Plast Reconstr Surg 2016; 137:910e–911e.
37. Telles E, Flores RD, Urrea-Giraldo F. Pigmentocracies: educational inequality, skin color and census ethnoracial identification in eight Latin American countries. Res Soc Strat Mobil 2015; 40:39–58.
38. Massie JP, Cho DY, Kneib CJ, et al. Patient representation in medical literature: are we appropriately depicting diversity? Plastic Reconstr Surg - Global Open 2019; 7:e2563.
39. Cho DY, Kneib CJ, Massie JP, et al. Visual representation of racial diversity in aesthetic surgery literature. J Plast Reconstr Aesthet Surg 2020; S1748-6815(20)30239-4. DOI: 10. 1016/j. bjps. 2020. 05. 048; Epub ahead of print.
40. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33:159–174.
41. 2013; Bureau USC. ACS Demographic and Housing Estimates (2009-2013 American Community Survey 5-Year Estimates). https://www.census.gov/history/pdf/2013armonkpop.pdf
42. Colby SL, Ortman JM. Projections of the Size and Composition of the U.S. Population: 2014 to 2060 Population Estimates and Projections. US Census Bureau Current Population Reports.
43. Goff PA, Eberhardt JL, Williams MJ, et al. Not yet human: implicit knowledge, historical dehumanization, and contemporary consequences. J Pers Soc Psychol 2008; 94:292–306.
Keywords:

breast surgery; diversity; health equity; plastic surgery; racial minorities

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.