Geographic practice preferences of graduating medical students pursuing careers in dermatology

Abstract

To the Editor: Geographic distribution disparities of US dermatologists result in poor access to dermatologic care in rural and underserved urban areas. Dermatologist density is significantly higher in metropolitan as compared to nonmetropolitan and rural counties,1Feng H. Berk-Krauss J. Feng P.W. Stein J.A. Comparison of dermatologist density between urban and rural counties in the United States.JAMA Dermatol. 2018; 154: 1265-1271Crossref PubMed Scopus (65) Google Scholar with most dermatologists practicing in metropolitan zip codes with high median incomes and low racial/ethnic diversity.2Vaidya T. Zubritsky L. Alikhan A. Housholder A. Socioeconomic and geographic barriers to dermatology care in urban and rural US populations.J Am Acad Dermatol. 2018; 78: 406-408Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar Limited access to care contributes to higher melanoma mortality rates in rural areas3Aneja S. Aneja S. Bordeaux J.S. Association of increased dermatologist density with lower melanoma mortality.Arch Dermatol. 2012; 148: 174-178Crossref PubMed Scopus (69) Google Scholar and decreased usage of dermatologic care for Hispanic, Black, and low-income patients.4Tripathi R. Knusel K.D. Ezaldein H.H. Scott J.F. Bordeaux J.S. Association of demographic and socioeconomic characteristics with differences in use of outpatient dermatology services in the United States.JAMA Dermatol. 2018; 154: 1286-1291Crossref PubMed Scopus (51) Google Scholar Few data exist on interest in rural or underserved urban practice in the dermatology workforce training pipeline. This study focused on the geographic preferences of medical students pursuing careers in dermatology. A cross-sectional study of graduating allopathic medical students was conducted using data from the 2016-2019 Association of American Medical Colleges Graduation Questionnaires (response rates: 80.5%, 81.1%, 83.0%, and 83.6%, respectively). The demographics of graduating medical students intending to practice dermatology (GMS-D) were compared to students pursuing other specialties (GMS-OS) using a Wilcoxon-type trend test or χ2 tests. The primary outcomes included intent-to-practice in a specific region (Northeast, South, Central, West, or non-United States) and urbanicity (urban, suburban, or rural). Unadjusted and adjusted odds ratios (aOR) were calculated for primary outcomes comparing GMS-D and GMS-OS, and among GMS-D, by sex, race/ethnicity, and sexual orientation, using logistic regression analyses controlling for age, sex, race/ethnicity, sexual orientation, and medical school region. All analyses were conducted using STATA version 16.1 with 2-sided α = 0.05 (Supplementary Methods, available via Mendeley at https://data.mendeley.com/datasets/mb9ygb9xcj/1). This study using deidentified, previously collected data was exempt from institutional review board review. Among 63,718 respondents, 5641 (8.9%) individuals with missing data on sex, sexual orientation, race/ethnicity, or specialty choice were excluded, resulting in 58,077 final study participants (1361 GMS-D and 56,716 GMS-OS) (Table I). GMS-D were more likely than GMS-OS to prefer to practice in suburban areas but less likely in urban or rural areas (Table II). Among GMS-D (Supplementary Tables I-III), there were no differences in geographic practice preferences by sex, but students identifying as a race/ethnicity underrepresented in medicine were more likely than non-underrepresented in medicine to prefer the South (44.1% vs 31.5%; aOR = 2.16; P = .01), and sexual minorities compared with heterosexuals were less likely to prefer the South (9.4% vs 34.5%; aOR = 0.31; P = .04) or suburban areas (16.9% vs 29.3%; aOR = 0.46; P = .01), but more likely to prefer the Northeast (34.0% vs 19.5%; aOR = 2.38; P = .04) and urban areas (83.1% vs 64.3%; aOR = 2.88; P < .001), with no sexual minority GMS-D intending rural practice.Table IDemographic characteristics of GMS-OS and GMS-DDemographicsAll participants (N = 58,077)P value∗P values calculated using a χ2 test or a Wilcoxon-type trend test (age only).GMS-OS (N = 56,716)GMS-D (N = 1361)Sex, n (%)Male29,056 (51.2)532 (39.1)<.001Female27,660 (48.8)829 (60.9)Age (y), n (%) <2623,172 (40.9)590 (43.4).02 for trend 27-2923,698 (41.8)566 (41.6) 30-326628 (11.7)148 (10.8) >333218 (5.7)57 (4.2)Race/ethnicity underrepresented in medicine (URiM), n (%)†Students were categorized being from racial/ethnic minority groups underrepresented in medicine if they reported identifying with at least one of the following: American Indian or Alaska Native; Black or African American; Hispanic, Latino, or of Spanish origin; or Native Hawaiian or other Pacific Islander. Non-URiM47,990 (84.6)12,204 (88.5)<.001 URiM†Students were categorized being from racial/ethnic minority groups underrepresented in medicine if they reported identifying with at least one of the following: American Indian or Alaska Native; Black or African American; Hispanic, Latino, or of Spanish origin; or Native Hawaiian or other Pacific Islander.8726 (15.4)157 (11.5)Sexual orientation, n (%)‡Sexual minority students were defined as those students selecting “bisexual” or “gay or lesbian” and heterosexual students were defined as those students selecting “heterosexual or straight” in response to the question, “How do you self-identify?” Heterosexual53,155 (93.7)1281 (94.1).55 Sexual minority3561 (6.3)80 (5.9)Medical school region, n (%)§AAMC-defined region of medical school attended. Northeast16,239 (28.6)375 (27.6).42 South18,815 (33.2)481 (35.3) Central15,070 (26.6)352 (25.9) West6592 (11.6)153 (11.2)AAMC, Association of American Medical Colleges; GMS-D, graduating medical students intending to practice dermatology; GMS-OS, graduate medical students intending to practice other specialties.∗ P values calculated using a χ2 test or a Wilcoxon-type trend test (age only).† Students were categorized being from racial/ethnic minority groups underrepresented in medicine if they reported identifying with at least one of the following: American Indian or Alaska Native; Black or African American; Hispanic, Latino, or of Spanish origin; or Native Hawaiian or other Pacific Islander.‡ Sexual minority students were defined as those students selecting “bisexual” or “gay or lesbian” and heterosexual students were defined as those students selecting “heterosexual or straight” in response to the question, “How do you self-identify?”§ AAMC-defined region of medical school attended. Open table in a new tab Table IIRegion and urbanicity of intended practice location between GMS-OS and GMS-Dn (%)P valueUnadjusted odds ratio∗Unadjusted odds ratio for the logistic regression model controlling for sexual orientation only.P valueAdjusted odds ratio†Adjusted odds ratio for the logistic regression model controlling for sexual orientation, sex, age, underrepresented in medicine identity, and medical school region.P valueRegion of intended practice location‡Students who selected “Undecided or No Preference” (11,566/58,077; 19.9%) or had missing data (8689/58,077; 15.0%) were excluded.Northeast GMS-OS (N = 36,878)8201 (22.2).171.0 (Ref)1.0 (Ref) GMS-D (N = 944)192 (20.3)0.89 (0.77-1.05).170.86 (0.69-1.05).13South GMS-OS (N = 36,878)10,679 (29.0).0061.0 (Ref)1.0 (Ref) GMS-D (N = 944)312 (33.1)1.21 (1.06-1.39).0061.16 (0.95-1.42).14Central GMS-OS (N = 36,878)6711 (18.2).571.0 (Ref)1.0 (Ref) GMS-D (N = 944)165 (17.5)0.95 (0.80-1.13).570.98 (0.78-1.23).88West GMS-OS (N = 36,878)10,385 (29.4).601.0 (Ref)1.0 (Ref) GMS-D (N = 944)270 (28.6)0.96 (0.83-1.11).601.04 (0.88-1.23).66Non-US GMS-OS (N = 36,878)452 (1.2).041.0 (Ref)1.0 (Ref) GMS-D (N = 944)5 (0.5)0.42 (0.18-0.99).040.43 (0.18-0.99).04Urbanicity of intended practice location§Students who selected “Undecided or No Preference” (7628/58,077; 13.1%) or had missing data (126/58,077; 02%) were excluded.Urban GMS-OS (N = 49,117)34,633 (70.5)<.0011.0 (Ref)1.0 (Ref) GMS-D (N = 1206)789 (65.4)0.79 (0.70-0.89)<.0010.80 (0.70-0.90)<.001Suburban GMS-OS (N = 49,117)10,194 (20.8)<.0011.0 (Ref)1.0 (Ref) GMS-D (N = 1206)344 (28.5)1.52 (1.34-1.73)<.0011.54 (1.36-1.76)<.001Rural GMS-OS (N = 49,117)4290 (8.7)<.0011.0 (Ref)1.0 (Ref) GMS-D (N = 1206)73 (6.1)0.67 (0.53-0.85)<.0010.65 (0.51-0.83)<.001GMS-D, graduating medical students intending to practice dermatology; GMS-OS, graduate medical students intending to practice other specialties; Ref, reference.∗ Unadjusted odds ratio for the logistic regression model controlling for sexual orientation only.† Adjusted odds ratio for the logistic regression model controlling for sexual orientation, sex, age, underrepresented in medicine identity, and medical school region.‡ Students who selected “Undecided or No Preference” (11,566/58,077; 19.9%) or had missing data (8689/58,077; 15.0%) were excluded.§ Students who selected “Undecided or No Preference” (7628/58,077; 13.1%) or had missing data (126/58,077; 02%) were excluded. Open table in a new tab AAMC, Association of American Medical Colleges; GMS-D, graduating medical students intending to practice dermatology; GMS-OS, graduate medical students intending to practice other specialties. GMS-D, graduating medical students intending to practice dermatology; GMS-OS, graduate medical students intending to practice other specialties; Ref, reference. This novel study identifies that GMS-D are less likely to pursue rural or urban practice and that race/ethnicity and sexual orientation influence geographic practice preferences. Study limitations include the significant portion of undecided students and that the preferences may not reflect actual practice location. Efforts to cultivate interest in practicing in underserved areas in the dermatology workforce pipeline may include mentoring premedical and medical students, rural or underserved urban residency training tracks, recruiting residents and faculty from underserved areas with a desire to return to serve their communities, expanding telemedicine, and funding incentives for practice in underserved areas.5Streifel A. Wessman L.L. Farah R.S. et al.Rural residency curricula: potential target for improved access to care?.Cutis. 2021; 107: 54-55;E2Crossref PubMed Scopus (2) Google Scholar Further studies assessing motivations influencing geographic practice preferences will inform efforts to reduce access disparities to dermatologic care in the United States. Dr Mansh is supported by the National Institute of Environmental Health Sciences (U01ES029603). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Dr Brodell is a principal investigator for a clinical trial (Novartis) and for the CorEvitas psoriasis biologic registry. He serves on editorial boards of American Medical Student Research (faculty advisor); Practice Update Dermatology (Editor-in-Chief); Journal of the American Academy of Dermatology (Associate Editor); Practical Dermatology; Journal of the Mississippi State Medical Society; SKIN: The Journal of Cutaneous Medicine; and Archives of Dermatological Research. Drs Fulk, Wessman, Farah, Firkin Smith, Gaddis, and Gupta have no conflicts to report. This material is based upon data provided by the Association of American Medical Colleges (“AAMC”). The views expressed herein are those of the authors and do not necessarily reflect the position or policy of the AAMC.