Trends in Race/Ethnicity Among Applicants and Matriculants to US Surgical Specialties, 2010-2018

This cross-sectional study assesses trends in the percentage of applicants and matriculants to US surgical specialties who identified as part of a racial/ethnic group underrepresented in medicine from the 2010-2011 to 2018-2019 academic years


Introduction
A diverse physician workforce would not only better reflect the increasingly racially/ethnically diverse population of the US but also contribute toward addressing health care disparities, improving patient outcomes and satisfaction, and fostering greater innovation in medicine. [1][2][3] Despite these observations and numerous initiatives, the percentage of individuals who identify as belonging to racial/ethnic minority groups in medicine has continued to remain below that in the general US population. 4,5 For example, in 2019, the percentage of the population who identified as Black or African American, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, and Hispanic or Latino was more than 33%, 6 while these same groups constituted approximately 12% of US medical school graduates in the class of 2019, without including those who identified as multiracial. 7 Multiple studies have shown that underrepresentation based on race/ethnicity is also specifically present in surgical fields both among trainees and faculty. 3, [8][9][10] In response, surgical training programs across the US have engaged in initiatives aimed at increasing the number of individuals underrepresented in medicine (URM) based on the definition by the Association of American Medical Colleges (AAMC) as "those racial and ethnic populations that are underrepresented in the medical profession relative to their numbers in the general population." 11 However, the impact of these efforts remains unclear. [12][13][14][15] This study used data from the AAMC to assess trends in the race/ethnicity of applicants and matriculants to surgical specialties between 2010 and 2018.

Methods
We performed a cross-sectional study using data provided by the AAMC Accreditation Council of Graduate Medical Education residency programs. This study was determined to be exempt from review by the University of Washington institutional review board because it did not involve human participants. Data were deidentified. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. 16 Self-reported race/ethnicity (alone or in combination) for applicants and matriculants to surgical and nonsurgical specialties from 2010-2011 to 2018-2019 academic years was obtained and analyzed in April and July 2020. Academic years have been abbreviated to the year that the individual submitted their application to residency (eg, an applicant and matriculant who applied during 2010-2011 is denoted with 2010 rather than the calendar year of matriculation [2011]). Medical specialties were categorized as surgical or nonsurgical based on the definition provided by the American College of Surgeons. 17,18 Medical specialties considered surgical included general surgery (categorical), neurological surgery, obstetrics and gynecology, orthopedic surgery, otolaryngology, plastic surgery (integrated), thoracic surgery (integrated), urology, and vascular surgery (integrated). Medical specialties considered nonsurgical included anesthesiology, dermatology, emergency medicine, family medicine, internal medicine, neurology, pathology (anatomic and clinical), pediatrics, psychiatry, and radiology (diagnostic). The URM designation followed the AAMC definition and represents individuals who identify as American Indian or Alaska Native; Black or African American; Hispanic, Latino, or of Spanish origin; and Native Hawaiian or other Pacific Islander. Race/ethnicity data were only available for US citizens and permanent residents and include individuals who may have completed their medical education outside the US. White, Asian, other race/ethnicity, unknown race/ethnicity, and non-US citizen or nonpermanent US resident categories were considered non-URM.

Statistical Analysis
The χ 2 test for trend was performed to assess changes in proportions across all years between 2010 and 2018 for URM groups and individual races/ethnicities, and the t test was performed to evaluate a difference between the mean proportions of individuals who identified as URM in surgical vs nonsurgical specialties using Prism, version 8.4.2 (GraphPad Software). Results were considered statistically significant at a 2-tailed P < .05.

Results
A total of 737 034 and 265 365 matriculants during the 9 years of the study period were included in the sample (Figure 1). Of these, 107 851 applicants (14.6%) and 33 544 matriculants (12.6%) identified as URM. A total of 29 724 of 134 158 applicants (22.1%) and 1295 of 41 347 matriculants (3.2%) to US surgical specialties were categorized as other race/ethnicity (self-identified), unknown race/ethnicity, and non-US citizen or nonpermanent US resident. A total of 134 158 applicants (18.2%) applied to surgical specialties, and 602 876 (81.8%) to nonsurgical specialties.    in 2018; P = .99). As with applicants, thoracic surgery was the only surgical specialty with an increase

Differences in Representation Among Matriculants Compared With Applicants
In a comparison of the difference between the mean percentages of matriculants and applicants who

Discussion
Despite efforts to diversify the workforce, our study found no significant change in the percentage of individuals who identified as URM applying or matriculating into the collective 9 American College Several national surgical societies have committed to promote diversity through the development of diversity committees and societies for the advancement of specific groups that are URM to increase racial/ethnic representation. [27][28][29][30] The American College of Surgeons has underscored the importance of moving beyond superficial efforts toward diversity by urging inclusivity, the concept of creating a welcoming space for all health care participants from faculty to patients as individuals and members of the group. 28,31 Framing recruitment and exposure initiatives in terms of value added over rote diversity metrics will be a necessity to overcoming the inertia of structural bias in creating sustainable growth in the proportion of surgical subspecialty training matriculants from URM backgrounds. 32 The difference in representation of matriculants compared with applicants who identify as URM in surgical specialties increased from 2010 to 2018 (Table). Individuals who did not identify as URM were also overrepresented by a mean of 20.1% compared with their proportion of the applicant pool.
This was likely attributable to the 22.0% mean difference in representation between White matriculants and applicants because the differences for Asian, other race/ethnicity, unknown race/ ethnicity, and non-US citizen or non-permanent US resident were less than 1.0% and often negative for each cycle. Orthopedic, plastic, and vascular surgery were the only specialties in which there was a smaller difference in URM representation for most of the 8 years after 2010 owing to an increase in matriculation of individuals who identified as URM; however, these specialties had more applicants who identified as URM than matriculants in this group, as has been reported previously for plastic surgery. 19 The difference in representation affects the entire training pathway: 14 There are also steps during the application process that should be studied further and potentially restructured. Websites are an underused way of communicating a program's commitment to diversity and could be enhanced to further increase diversity among applicants. 44 Receiving an interview offer has been recognized as an important step of gatekeeping the number of applicants who identified as URM. 20 A study assessing general surgery programs found that having more residents and faculty who identify as URM was not associated with an increased likelihood of applicants who identified as URM being offered an interview. 45 Creative and new strategies should be developed to encourage more applicants, make the interview process more equitable, and decrease the overall attrition of individuals who identify as URM who matriculate. 12,46-48 Continued active evaluation in a reiterative way will be needed to identify strategies that are or are not helping to diversify the workforce.
Addressing the work environment experienced by residents who identify as URM may improve the resident experience as well as influence the encounters of medical students who identify as URM during their clinical rotations and their perception of surgery. 49,50 Studies about women identified that the environment of training affects why students choose to pursue surgery. 51,52 Recent reports detail the discrimination and harassment that individuals who identify as URM and women experience in surgical training, which has been disavowed by the American Surgical Assocation. 53,54 Further change in hierarchical and social dynamics is likely necessary to increase the number of individuals who identify as URM who ultimately pursue a surgical field.

Limitations
This study has limitations. We used self-report data, which relies on participants' willingness to accurately share personal information. Race is a complex social category with evolving meaning, shown by how AAMC permitted students to select more than 1 race/ethnicity after 2013. This change means that data after 2013 with race/ethnicity alone and in combination counts multiracial individuals in more than 1 category, potentially overestimating the percentage of students who identify as URM. We did not have access to identify which applicants who identified as URM attended osteopathic or non-US medical schools, which may be a confounder because these students tend to have lower matriculation rates into US residency programs, including surgical ones. 55,56 Because our focus was on surgical specialties, we did not investigate the trends for individual nonsurgical specialties; perhaps there is insight that can be gained from that analysis in future studies.