Trends in Internal Medicine Faculty by Sex and Race/Ethnicity, 1980-2018

Key Points Question Did sex and racial/ethnic diversity in the academic internal medicine workforce between 1980 and 2018 reflect the changing demographic composition of the general population (in 2017) and medical student body? Findings This cross-sectional study used data on full-time medical school faculty appointed to US internal medicine departments from 1980 to 2018, matriculants to US medical schools from 1980 to 2018, and the general US population from 1980 to 2017. The study found that progress was made in diversifying the academic workforce; however, it may not yet reflect the diversity of US medical students or the US population. Meaning These findings suggest that continued improvement of recruitment and retention efforts for women and groups who are underrepresented in medicine as medical students and faculty in internal medicine is needed.


Introduction
By 2060 the US population will have undergone significant demographic changes. The US Census Bureau projects that between 2014 and 2020 the population will have increased by nearly 100 million people, reaching 417 million, and the number of adults 65 years and older is expected to increase by nearly 75% between 2020 and 2060 (from 56.4 million to 98.2 million). 1 The racial and ethnic makeup of the US population will change as it becomes a minority-majority nation, with racial/ ethnic minorities constituting 56.4% of the population by 2060. 1 Another anticipated change is an increase in the number of adults with multiple chronic medical conditions, 2,3 and patients with multiple chronic conditions use more health services than other individuals and account for approximately 83% of all health care spending. 4,5 Taken together, these projected changes will bear directly on the health and economy of future generations, and a larger, more diverse physician workforce focused on the prevention and management of complex chronic disease (such as internal medicine [IM] and its subspecialties) could help meet the nation's increasing health care needs and lead to improved patient outcomes. [6][7][8][9] Because IM faculty serve as teachers and direct role models for medical students, they play an instrumental role in recruitment into this field. Indeed, diversity among faculty is associated with diversity and cultural competence among students and a reduction of health-related disparities. [10][11][12][13] However, it is unclear to what extent diversity among IM faculty reflects the diversity of the nation's population or the diversity of the medical student body. Our aim in conducting this study was to add to existing research [14][15][16] on demographic trends among faculty in other specialties by analyzing the sex and race/ethnicity composition of IM faculty, other medical school faculty, medical school matriculants, and the general US population. We also examined these trends through the intersection of sex and race/ethnicity.

Methods
This secondary analysis of a cross-sectional study used data on full-time medical school faculty appointed to IM departments in the US from January 1, 1980, to December 31, 2018, matriculants to US medical schools that granted doctor of medicine (MD) degrees from January 1, 1980, to December 31, 2018, and the general US population from January 1, 1980, to December 31, 2017 (the most recent data available) to describe and compare sex-and race/ethnicity-specific trends in these groups over time. We obtained medical school faculty data from the Association of American Medical Colleges (AAMC) Faculty Roster, a comprehensive national database of full-time US medical school faculty. 17 Faculty were classified based on department, not medical specialties; thus, IM faculty included general IM and other subspecialties. We obtained medical school matriculant data from the AAMC Applicant Matriculant File. 18 We obtained general population data from the US Census Bureau national population data sets. [19][20][21][22] For US Census Bureau and medical student data from the AAMC Applicant Matriculant File, race/ethnicity and sex are self-reported; however, faculty race/ethnicity and sex data from the AAMC Faculty Roster are not considered self-reported because they are reported to the Faculty Roster by the faculties' institutions and then supplemented by other AAMC data sources. Additional information regarding the reporting and classification of race/ethnicity can be found in the eAppendix in the Supplement. This secondary analysis used deidentified data; however, in the original data collection, data were identifiable, and written informed consent was obtained. This secondary analysis was deemed exempt from guidelines for research involving human participants by the institutional review board of the American Institutes for Research. combination with any other race or ethnicity), Black (Black or African American, alone only), Native Hawaiian or other Pacific Islander (alone only), and American Indian or Alaska Native (alone only). We then compared these proportions with the proportion of female and URM matriculants in US MD-granting medical schools and the proportion of women and individuals from traditionally underrepresented groups in the general population. To allow for comparability, we defined underrepresented minorities as the same groups who compose URM faculty and matriculants (ie, Hispanic, Black, Native Hawaiian or other Pacific Islander, and American Indian or Alaska Native).
The χ 2 test was used to compare IM faculty with all other clinical faculty in 2018. To assess trends in sex and URM status between 1980 and 2018, we estimated the slope and associated probability values for each group by using simple linear regression models in which year was used as an independent variable. A 2-sided P < .05 was considered to be statistically significant. Statistical analyses were conducted using SAS software, version 9.4 (SAS Institute Inc).

Results
From 1980 to 2018 the absolute number of full-time IM faculty increased almost 4-fold (from 10 964 to 42 547) ( Table), and IM retained its broader status as the clinical practice specialty with the highest number of physicians during those years. 17

Comparison by URM Status
The prevalence of URM individuals among faculty and medical students increased, although not to the same extent as female representation. When changes in the percentage of URM individuals over time were examined, all trends (slopes) were positive and statistically significant (slopes of 0.17 for IM faculty, 0.13 for other clinical faculty, and 0.14 for matriculants; P < .001 for all).
For IM faculty, the percentage of URM individuals has more than doubled during the past 38 years (4.1% vs 9.7%) (Figure 2). However, in 2018, URM groups still made up a small portion of IM faculty, which was nearly identical to the URM representation in all other clinical departments (9.7% vs 9.8%, χ 2 = 0.120, P = .73). For matriculants, the URM percentage increased in the early 1990s,

Comparison of the Intersection of Sex and URM Status
As  Comparing the percentages of URM female and male IM faculty, starting in the late 1980s and continuing through the 1990s, URM proportions among female IM faculty appeared to increase faster than those among male IM faculty, but beginning around 1996, the increase in the percentage of URM female faculty slowed, whereas growth remained steadier among male faculty (Figure 4).
The percentage of female IM faculty who were URM remained at least 1 percentage point above that for male IM faculty who were URM, a relationship that remained for the almost 30 years of the study period (1.3% higher in 1980 and 1.9% higher in 2018) (Figure 4).
Among URM female and male matriculants, from 1980 to 1994 there were also notable trends.
There was a sharp increase in URM female medical school matriculants that peaked at 19.4%. This number decreased after 1994, and then in 2006 the proportion began to increase again, reaching 19.1% in 2018. The percentage of URM male matriculants decreased slightly after 1994 but increased to 17.1% by 2018.

Discussion
In this cross-sectional study, despite improved representation, women, URM individuals, and perhaps most especially those who share these dual identities, continued to face challenges in terms of true representation in academic medicine. According to current literature, these challenges are associated with a lack of parity in compensation, 23,24 promotion, leadership opportunities, 25 and, at their worst, frank discrimination and harassment. [26][27][28][29] Equally concerning are the increased levels of isolation, burnout, and attrition in academic medicine. 30    37 The present study saw a corresponding peak in the percentage of URM matriculants during this time; however, the campaign was ultimately unsuccessful in meeting its goals, likely hindered, in part, by state ballot initiatives to ban affirmative action. In subsequent years, our study found, there was a decrease in the percentage of URM matriculants.

JAMA Network Open | Medical Education
Although legislative challenges related to affirmative action may have played a role, the disparities seen in this study are too stark to attribute them solely to those conflicts. Possible causes of the differences observed in the percentages of IM clinical faculty and medical school matriculants who are URM warrant further investigation. Prior work 39 found that although URM minority matriculation was increasing for male and female Hispanic students and Black female students, Black males were faring less well. This finding may explain why the present study showed overall trends toward increases in the percentages of female matriculants and IM faculty who were URM but less pronounced trends for URM men. The data in the present study suggest that although progress has been made, a need to focus recruitment efforts in academic IM remains .
A lack of diverse role models among IM faculty also may amplify the difficulties associated with increasing IM faculty diversity. Almost one-third of the US population are underrepresented minorities, whereas only 18.1% of matriculants and 9.7% of clinical IM faculty in this study were individuals from URM groups. A lack of representation of women and URM individuals in academic IM may contribute to a medical student body unable to see a future in academic IM. 40 As an increasingly diverse medical student population moves through the pipeline and completes training, it remains to be seen how many will move into faculty positions.
Renewed attention and innovation around diversity and inclusion are needed to improve representation and retention of women and underrepresented groups in academic IM. 35,41,42 Efforts in line with this mission include incentives to reduce loan burdens, 7 changing traditional promotion criteria to include activities to which women and URM individuals may disproportionately contribute (eg, cultural competence curriculum development, diversity committee representation, and social science research), 43 hiring and promoting diverse faculty to serve as mentors to medical students and junior faculty 44,45 (without unduly burdening them with the sole responsibility for mentoring female and URM students), reinvigorating the pipeline to medical school through sustained community outreach to ensure access to resources and opportunities to attend and succeed in medical school, 46 supporting race-conscious admissions policies and challenging anti-affirmative action legislature that can hamper efforts toward sustained underrepresented minority recruitment, 37 experimenting with technology and other practices that can enhance job flexibility, 36 and getting organizations to think of innovative

JAMA Network Open | Medical Education
Trends in Internal Medicine Faculty by Sex and Race/Ethnicity, 1980-2018 ways to promote general IM from a national level. 7 However, the most effective evidence-based strategies appear to be those focused on shifting culture, which requires change that is supported by leadership and paired with accountability mechanisms. 41,[47][48][49] The nation needs physicians who understand the importance of disease prevention, excel at chronic disease management, and routinely integrate social and economic determinants of health into the practice of medicine. 50 Internal medicine physicians represent the largest specialty in medicine, and they focus on the comprehensive care of adults with complex medical problems, making them ideally suited to play a significant role in managing the challenges that lie ahead. 7 Therefore, a special emphasis must be placed on increasing diversity among IM faculty to avoid perpetuating disparities in the field and in the health of the population. 51

Limitations
This study has limitations. First, there were variations in the way that race/ethnicity was reported (self-report vs institutional report; see the Methods) and classified. ). In addition, given the cross-sectional nature of these data and a lack of information regarding faculty and student experiences, we cannot infer causality from any of the results presented.

Conclusions
This cross-sectional study found that although progress was made in diversifying the demographic makeup of IM clinical faculty from 1980 to 2018, it may not yet reflect the diversity of the current population or of medical school matriculants. As the health care system is faced with caring for an increasingly diverse, aging, chronic disease-laden patient population, the urgency to meet the needs of patients is escalating, and the importance of securing a diverse physician workforce that can best serve them is crucial. Internal medicine continues to be the largest specialty, and internists have a broad effect on health care, playing a key role in providing primary and subspecialty care. 7 Continued improvement of recruitment and retention efforts for female and URM medical students and faculty in IM has the potential to significantly increase the size and diversity of the IM physician workforce and improve the quality of, and access to, comprehensive and equitable care.

JAMA Network Open | Medical Education
Trends in Internal Medicine Faculty by Sex and Race/Ethnicity, 1980-2018