15-Year Diversity Trends Among Dermatology Resident Trainees Compared With Other Specialties

Abstract

This cross-sectional study examines racial and ethnic diversity trends among dermatology resident trainees and applicants compared with other specialties.

Dermatology, the second-least diverse specialty, is striving to increase workforce diversity.¹ We examined racial and ethnic diversity trends among dermatology resident trainees and applicants compared with other specialties.

Methods

For this retrospective, cross-sectional study, we acquired a special report from the Association of American Medical Colleges (AAMC) of trainees’ and applicants’ self-reported race and ethnicity by specialty from 2005 to 2020. Rate ratios (RR) were calculated to contextualize resident demographics with US Census demographics, where RR = % demographic (trainees)/% demographic (US population).

When the RR equaled 1, this indicated demographic parity between the trainee and US populations, whereas an RR of less than 1 indicated underrepresentation. We defined ideal representation as proportionate to the US population because of the well-established benefits of achieving a physician workforce reflective of patient diversity (eg, improved patient outcomes and satisfaction, heightened attention to mitigating disparities, and increased practice in underserved settings).¹ We used 2-sample t tests to compare means and linear regression to assess representation trends (annual RR by race and ethnicity). Stata statistical software, version 17.0 (StataCorp LLC) was used, and statistical significance was assessed at P < .05. University of California San Francisco Institutional Review Board exempted this study from review because it used previously collected and deidentified data. The STROBE reporting guideline was followed.

Results

Black and Latinx trainees were less represented in dermatology than in other specialties (mean annual RR [SD], Black trainees, dermatology, 0.32 [0.04]; primary care, 0.54 [0.04], P < .001; specialty care, 0.39 [0.02], P < .001; Latinx trainees, dermatology, 0.14 [0.09]; primary care, 0.23 [0.08], P < .01; specialty care, 0.18 [0.07], P = .14). Between 2005 and 2020, the annual representation of Black trainees remained unchanged in dermatology (regression-coefficient, β = −0.001; P = .66), but down-trended for primary (β = −0.008; P < .001) and specialty care (β = −0.004; P = .001). Meanwhile, the representation of Latinx trainees remained unchanged in dermatology (β = −0.003; P = .61) and specialty care (β = 0.005; P = .24), but increased in primary care (β = 0.013; P < .001) (Figure).

Figure. Trends of Black and Latinx Representation Among Trainees by Specialty, 2005-2020^a,b.

A, Compared to the US population, where a rate ratio of 1 indicates proportionate representation, Black and Latinx resident trainees were underrepresented across all specialties, but more so in dermatology. There were no statistically significant changes in representation of Black or Latinx dermatology trainees over the course of the study period. B, Latinx representation declined to an inflection point from 2005 to 2008. This phenomenon coincides with changes in the Association of American Medical Colleges (AAMC) race and ethnicity reporting methodology that allowed survey respondents to select multiple race and ethnicity categories. While Black and Latinx representation shows an upward trend between 2017 and 2020, this finding was not statistically significant (P = .05, each). Primary care specialties include family medicine, internal medicine, pediatrics, and combined internal medicine and pediatrics (AAMC State Physician Workforce Data Report, 2022). All other specialties accredited by the Accreditation Council for Graduate Medical Education were classified as specialty care.

^aTrainee demographics were obtained from AAMC by request. US demographics were obtained from the US Census.

^bAmerican Indian/Alaska Native or Native Hawaiian/Pacific Islander representation trends were not analyzed due to the small overall number of trainees.

^cThe slope statistically significantly differs from 0.

The proportions of each racial and ethnic demographic among dermatology residency applicants and matriculating trainees (postgraduate year [PGY]-2s) from 2005 to 2020 are presented in the Table. Black and Latinx race and ethnicity comprised a lower mean proportion of PGY-2s compared with applicants annually (Black, 4.01% [1.09%] vs 5.97% [0.69%], P < .001; Latinx, 2.06% [1.47%] vs 6.37% [1.62%], P < .001). Conversely, the mean proportion of White race among PGY-2s trainees was larger than that of applicants annually (65.09% [3.91%] vs 51.71% [2.89%], P < .001). Lastly, Asian race was represented proportionately across PGY-2s and applicants (19.26% [1.84%] vs 19.21% [1.77%], P = .93).

Table. Racial and Ethnic Representation Among Dermatology Applicants to Matriculating Trainees (Postgraduate Year [PGY]-2s), 2005-2020^a.

Race/ethnicity	No. (%)				Mean annual difference (% PGY-2s − % Applicants)
	Applicants, 2005	PGY-2s, 2006	Applicants, 2019	PGY-2s, 2020	2005-2020	2010-2020	2015-2020
Totalb	851	342	1170	456	NA	NA	NA
Asian	171 (20.1)	60 (17.5)	253 (21.6)	105 (23.0)	+0.1	−0.4	0.0
Black	48 (5.6)	12 (3.5)	71 (6.1)	20 (4.4)	−2.0	−1.9	−1.5
Latinx	57 (6.7)	6 (1.8)	65 (5.6)	18 (3.9)	−4.3	−3.7	−1.7
Native American/Alaska Native	1 (0.1)	1 (0.3)	4 (0.3)	2 (0.4)	0.0	0.0	0.0
Native Hawaiian/Pacific Islander	1 (0.1)	0	0	0	0.0	0.0	0.0
White	449 (52.8)	240 (70.2)	556 (47.5)	259 (56.8)	+13.4	+13.6	+11.5

Abbreviation: NA, not applicable.

^a Trainee demographics were obtained from the Association of American Medical Colleges by request. US demographics were obtained from the US Census.

^b Multiracial (non-Latinx), unknown, and non-US citizen/permanent resident categories are included in annual dermatology applicant and entering dermatology resident totals.

Discussion

Despite calls to increase diversity,^1,2 racial and ethnic minority representation among trainees in dermatology has largely remained stagnant, even compared with other specialties. Much of these disparities can be attributed to the leaky pipeline—the disproportionate, stepwise reduction in racial and ethnic minority representation along the path to medicine.^2,3 This leaky pipeline is the direct result of structural racism, which includes, but is not limited to, historical and contemporary economic disinvestment from majority-minority schools, kindergarten through grade 12.^3,4

Furthermore, this study’s finding that disproportionately fewer Black and Latinx applicants entered dermatology residency compared with nonunderrepresented groups highlights the opportunity for intervention in residency selection. Historically, dermatology residencies compared applicants using biased metrics^2,5,6; however, recent changes to US Medical Licensing Examination Step 1 scoring will necessitate adjustments to selection criteria. Holistic review—a mission-aligned process that considers academic metrics alongside desired experiences and attributes—has shown potential promise to reduce racial disparities in admissions.² Dermatologists committed to increasing workforce diversity can advocate for equitable school policies and mentor underrepresented students of all ages to promote interest in dermatology and advancement throughout the educational pathway.^1,2,4

This study’s strengths include the use of consistently collected and self-reported race and ethnicity data. The AAMC surveys are voluntary; thus, these findings may be limited by missing data. Match rates could not be ascertained from AAMC data as the number of Electronic Residency Application Service applicants does not equal the number of applicants who participate in the Match; the applying/matriculating ratio is an approximation.

As these findings illustrate, dermatologists must intervene throughout the educational pipeline, including residency selection and mentorship, to effectively increase diversity.