Underrepresentation of Underrepresented Minorities in Academic Medicine: The Need to Enhance the Pipeline and the Pipe

Why bother? 

If the current US population continues to grow at a similar pace, nearly 50% of the 2050 US census will be non-white or a “person of color.”5, 6 Although with each successive generation (about 40 years or so), we clearly witness and document substantive changes ushering us toward the utopian goal of a color-blind society, the numbers bear witness to the fact that significant changes need to occur to approach this goal. With respect to health care, minority populations, which currently comprise approximately 35% of the US population (∼27% are Hispanic or Black; Figure 1A), are likely to receive a higher percent of uncompensated health care (“Uncompensated care [UC] is health care that is delivered, but not paid for by either a patient or a third party payer. Most UC is delivered to the very ill during or after a visit to an emergency room. In 2004, UC was ∼$41 billion dollars”).6, 7, 8 Assuming that those likely to receive uncompensated care will comprise a larger share of minority physician practices provides some impetus for increasing the number of URM practitioners and academicians. Black physicians care for significantly more black patients (25 percentage points more versus other ethnic group physicians; P < .001) and for more patients on Medicaid (on average 45% of their patients were insured by Medicaid; P ≤ .001); Hispanic physicians care for more Hispanic patients (21% percentage points more; P < .001), and more uninsured patients (P < .03) than do other physicians.9 In addition, there is evidence that black and Hispanic patients seek physicians of their own race because of personal preference or language, and not only because of geographic proximity.10 Thus, simply from an economic necessity, the health care delivery system must factor in how to increase the number of physicians who will deliver health care, as well as increase the number of physician scientists, clinical investigators, and scientists who will analyze the ailments that plague these populations.

View Large Image Download to PowerPoint

Figure 1. US census data and distribution of matriculated medical students, gastroenterology and total physician trainees among US ethnic and race backgrounds. (A) The distribution of the different US subpopulations based on the 2000 US latest official census and the estimated 2008 US census24 (rounded off to the nearest million) is shown. The population total also includes the categories of Native Hawaiian and other Pacific Islander and those with ≥2 races that are not included in the race/ethnic background analysis. The Hispanic (Latino) designation includes Mexican American, Other Hispanic, Cuban, Puerto Rican, and Multiple Hispanic. The US Census Bureau defines 5 categories of race (white, black or African American, American Indian and Alaskan native, Asian, and native Hawaiian/other Pacific Islander) and 2 categories of ethnic background (Hispanic/Latino and non-Hispanic/Latino).25 NANA, Native American, Native Alaskan; NHOPI, Native Hawaiian and other Pacific Islander. (B) Details of the numbers used to generate the graph are shown in Supplemental Table 1. The numbers were collected from the 2008–2009 graduate medical education data.15

It is a reasonable assumption that most health care recipients do not care what “color” or ethnic group their physician belongs to as long as they are well trained and care about them as patients. However, black respondents with black physicians were more likely than those with non-black physicians to rate their physicians as excellent and as providers of preventative care. Similarly, Hispanic patients treated by Hispanic physicians were more likely than those treated by non-Hispanic physicians to be satisfied with their health care.11 Although these conclusions are derived from data obtained ≥10 years ago, these conclusions are likely to still hold. Thus, URM providers are more likely to serve patient populations that are a reflection of themselves, which in turn are the same populations typically listed as “the underserved.” However, statistically a majority of URMs in the United States will have a non-URM physician because URMs only comprise about 7% of practicing physicians.5 Therefore, from an educational perspective, academic medical faculty who are training the next generation of physicians as well as those delivering health care should reflect the diverse populations they will be serving.

Enhancing the pool of URM trainees and academic physicians will likely alleviate the present disparities in the quality of health care that relate to specific measures of health care delivery and URM populations. For example, the 2008 National Health Care Disparities Report indicates that blacks were more likely to be diagnosed at an advanced stage with colorectal cancer than whites (104 vs 80 per 100,000, respectively).12 In addition, Hispanics were less likely than non-Hispanic whites in 2005 to receive colorectal cancer screening (37.3% vs 58.5%).12 However, URM disparities in health care are complicated and relate to several other benchmarks such as income, level of education, and access to health care. Regardless, the number of matriculated black and Hispanic medical students is also not representative of these 2 minority groups in the US population (Figure 1B), which is an additional impetus for promoting predoctoral URM students to pursue health-related disciplines.

In addition to the societal needs to increase the pool of URM trainees and academic physicians, there are several other tangible benefits. These benefits include providing a pool of mentors for students, to better serve patients, to make the medical center a more diverse and interesting place, to bring different points of view to debates, and to provide a pool of researchers and clinical investigators who bring a different perspective to their scholarly activities. Diversity in medicine is a win–win goal that improves the institution and the educational experience.13 Of note, Project 3000 by 2000, which began in 1990 as a AAMC initiative, intended to enroll 3000 URM students in US medical schools by 2000.14 This is indeed a laudable effort, but it remains to be met; even in 2007 the total number of matriculated URM students was approximately 2500 (Supplemental Table 1 [vs 1470 URM enrollees in 1990]).

Current statistics and issues pertaining to academic physicians 

The US population shows ethnic/race fluctuations over time (Figure 1A), with a drop between the 2000 census and the estimated 2008 census in the percent of whites (69.1% vs 65.6%) compared with an increase in Hispanics/Latinos (12.5% vs 15.4%), and the limited increase in blacks (12% vs 12.2%). However, the percent of faculty has not changed significantly between 1990 and 2008 for blacks (now ∼3%) or Hispanics (now ∼4%; Figure 2), which reflects in large part the persistent issue of underrepresentation of these populations and the Native American/Alaskan group as compared with their US population distribution.

View Large Image Download to PowerPoint

Figure 2. Faculty ethnicity and race distribution in the United States. Details of the numbers used to generate the graph are shown in Supplemental Table 1, and represent the numbers for 1980, 1990, 2000, 2003, 2005, and 2008 for the sum of all faculty (professors, associate professors, assistant professors, and instructors). The data were collected from AAMC Faculty Roster3 and publications.26

The decrease in the relative population of whites (Figure 1A) parallels the relative decrease in white medical school faculty (Figure 2), whereas Asian faculty has increased representation within medical school faculties as compared with their relative population in the United States. Although there is an overall increase in total black and Hispanic faculty when comparing 1980 with 2008, the disturbing trend for blacks is that, since 2000, there has not been a change in the percent of total black faculty in the United States (Figure 2). The one slight improvement is that the number of black assistant professors has increased from 2% of total assistant professors in 1980 to 4.1% in 2000 (data not shown); however, this is unchanged in 2008 (Figure 3). Within the Hispanic faculty group, the “Other Hispanic” category is the largest (67.5%) and Cuban Hispanics represent the smallest group (1.2%), which does not reflect their representation in the US census (Figure 3). An accurate breakdown with respect to ethnic background and academic versus community practice affiliation is presently not available from the major gastroenterology societies including the American Gastroenterological Association, the American Association for the Study of Liver Diseases, the American College of Gastroenterology, or the American Society for Gastrointestinal Endoscopy.

View Large Image Download to PowerPoint

Figure 3. Faculty rank distribution as related to ethnicity and race. Details of the numbers used to generate the graph are shown in Supplemental Table 1 and correspond to AAMC Faculty Roster3 and publications for 2008.26 Within the Hispanic group of 5,048 total faculty in 2008, the subgroup percentages are 1.2% Cuban, 13% Mexican, 18.3% Puerto Rican, and 67.5% other Hispanic. In terms of the latest official US census for 2000, the distribution of the 35.3 million Hispanics is 3.2% Cuban, 58.4% Mexican, 9.6% Puerto Rican, and 28.8% other Hispanic.24

Another important statistic to highlight is that, for all ethnic groups during 2008 except for blacks, the total percentage of male outnumber female faculty by 1.7- to 2.1-fold, with whites having the greatest male-to-female ratio of medical school faculty among all other ethnic groups (Figure 4A; Supplemental Table 2). In addition, and equally striking, is that the percentage of females at the rank of professor ranged between 16.2% and 17.7% in 2006 and 2008, respectively (Supplemental Table 3), and this disproportionate number of women at the professor rank cuts across all ethnicities and races (Figure 4B). Notably, the percent of total female faculty shows a progressive decline from the instructor to the professor rank (Figure 4B). Also important to highlight is the selective lower ratio of male-to-female total black physician trainees (0.75 in 2008–2009 for blacks vs 1.16–1.31 for the remaining groups; Supplemental Table 4). The ratio of male-to-female trainees has shifted significantly toward equalization across all ethnic backgrounds during the past 10 years (eg, it was 1.62–1.81 male-to-female during 1998–1999), except for black trainees (Supplemental Table 4). The potential reasons for the selectively low male-to-female black trainee ratio merits further assessment.

View Large Image Download to PowerPoint

Figure 4. Distribution of female and male faculty by rank and ethnicity/race. Details of the numbers used to generate the graphs are shown in Supplemental Tables 2 (for panel A) and 3 (for panel B, year 2008). (A) Percent female faculty for each of the race/ethnic backgrounds. The data were obtained from the AAMC Faculty Roster and excludes faculty with missing gender data.3, 7 NANA corresponds to Native American, Native Alaskan; and NHOPI represents Native Hawaiian and other Pacific Islander.

Therefore, the major issues that need to be addressed include increasing the pipeline of URMs, promoting the success and retention of junior URM faculty, enhancing the support of senior URM faculty to serve as the needed mentors, and expanding the pool of URM and non-URM mentors for URMs. One statistic that exemplifies the apparent lack of retention of black assistant professors is their low ratio of associate to assistant professors (32%–34%) in 2000–2008, which may be taken as a reflection of lack of retention, as compared with that of whites (65%–66%) and Hispanics (38%–43%; Table 1). Similar trends are found when comparing the ratio of professors to associate professors. Unfortunately, the pipeline based on the most recent 2008–2009 figures15 does not seem promising when analyzing the number of adult or pediatric gastroenterology trainees or total resident physicians in ACGME-accredited and in combined specialty graduate medical education programs (Figure 1B; Supplemental Table 1).

Table 1. Relative Estimates of Promotions When Comparing Transitions from 2000 to 2008⁎

		White			Hispanic			Black
		Assistant	Associate	Professor	Assistant	Associate	Professor	Assistant	Associate	Professor
	2000	28,485	18,672	22,139	1,859	800	678	1,648	564	311
	2008	30,175	19,676	24,803	2,636	1,011	912	2,054	674	394
	Percent ratio of Assoc to Assis
	2000	66%		—	43%		—	34%		—
	2008	65%		—	38%		—	32%		—
	Percent ratio of Prof to Assoc
	2000	—	119%		—	85%		—	55%
	2008	—	126%		—	90%		—	59%

Assoc, Associate Professors; Assis, Assistant Professors; Prof, Professors.

⁎ Numbers are derived from AAMC institutional member access to Faculty Roster data.

As with any dataset, there are potential caveats. The data shown in Figure 1, Figure 2, Figure 3, Figure 4 represent statistics of full-time faculty and does not include volunteer or part-time faculty. Another potential limitation is the possibility of incomplete or biased reporting. Regardless, the numbers are striking enough such that the conclusions are likely to hold.

Recommendations to the system at large 

The low representation and the stagnation of the numbers of Black and Hispanic faculty in US medical schools, which is mirrored in adult and pediatric gastroenterology and matriculated medical students (Figure 1), are troubling. Significant efforts by the National Institutes of Health (NIH) and medical schools to promote diversity have been in place but the impact has not been as rewarding as one would have hoped or anticipated. This conclusion is based on the trends and data that are collected by the AAMC and the American Medical Association.3, 15 Some of the issues were articulated by black high school juniors in a Milwaukee pubic high school where 89% of the students are black. Students stated that, from their perspective, the major barriers to becoming a physician include financial constraints, lack of knowledge about the medical field, limited encouragement at home or school, negative peer pressure, lack of black role models, racism in medicine, and better alternatives for securing a high income.16 These sentiments provide clear aspects that need to be addressed. We envision several specific suggestions that might contribute to reversing the current trend (Figure 5) as follows.

View Large Image Download to PowerPoint

Figure 5. Two major goals are required to address underrepresentation of URMs in academic medicine. Two areas need to be addressed in order to enhance the number of URM academic physicians and biomedical scientists. First is to increase the pool of URMs who pursue health-related sciences by attracting graduate/medical, undergraduate and high school students. Second is to develop strategies to retain junior trainees and faculty members.

Recommendations to the URMs 

The URMs themselves play a critical role in promoting diversity, and in that capacity need to proactively undertake several measures. First, URMs should seek out mentors and role models early on, and it is important to highlight that these mentors need not come from an URM background. One concrete example is reflected by one of the authors of this commentary (J.L.M.) who sought out mentors (Fred Gorelick, Tadataka Yamada) who were not URMs, but who were instrumental in helping promote J.L.M.'s career as would any outstanding mentor irrespective of ethnicity or race. A second important reminder to the URMs is to network and to take it upon themselves to strive to be role models and mentors to those junior to them. The latter is a responsibility that brings the joy and fulfillment of witnessing the success of a trainee. Third, URMs need to support and participate in diversity-promoting programs and increase their visibility within their own institutions. However, this may spill into what may be termed the “minority tax” (eg, “Black tax,” “Hispanic tax”), which reflects on the appointment of URM faculty to more committees than their non-URM colleagues. This is a challenge that URMs need to balance with the need for adequate protected time for their own academic progress. Fourth, the recognition by URMs and non-URMs alike that promoting diversity, while doing so not at the expense of underserving any trainee no matter what color or creed, is ultimately uplifting to our society at large. Fifth, the need by URMs to exercise and cultivate their resilience to enhance their academic productivity. Examples of important resilience measures (which of course help any minority or majority) include clarity of goals and priorities, spirituality, family support, having a sense of humor, hard work, learning to organize and multitask, being able to say no, and assertiveness.23 Fifth, the realization by the URMs that their abilities to achieve are limitless, and that many of the overt barriers that may have interfered with their success in the past are now less evident.