Strategies to Enhance Diversity, Equity, and Inclusion in Pathology Training Programs: A Comprehensive Review of the Literature

There is overwhelming evidence demonstrating how DEI enrich trainee experiences and improve patient outcomes. Accordingly, we sought to review the extant literature on DEI within the field of pathology, with an emphasis on graduate medical education training programs. As there are limited resources on DEI specific to the field of pathology, we also sought to characterize the efforts of other medical specialties. We recognize that although pathology is a unique specialty, many of the DEI efforts and interventions implemented in other specialties may be applicable to our field.

A PubMed search was conducted of articles published prior to June 2020 using the following search terms either alone or in combination: diversity and inclusion, equity, pathology training programs, pipeline programs, pathology post-sophomore fellowship, residency training, underrepresented minority, underrepresented in medicine. This was extended to a general Google search of similar combinations of the same terms. The search terms were developed to focus on resources on the topics of DEI in pathology and in medical specialty training programs. Although this literature review focuses on diversity within pathology, studies pertaining to other medical specialties or to research fields were included in the literature review. We recognize that there are shared challenges to DEI in nonmedical disciplines; however, this review is focused on medical specialty training programs specifically. The search was restricted to English-language publications and publications that had been previously translated to English. Additional content was obtained through review of the gray literature, including the Web sites of the AAMC, the Electronic Residency Application Service, and the National Resident Matching Program. A total of 87 publications met the search criteria and were included.

During the past several decades, medical student interest in pathology residency training has declined substantially.^3,23,24  Many possible causes have been cited, including limited patient interaction, decreasing exposure to the discipline of pathology in medical school curricula, a perceived lack of jobs, and lack of medical student identification with practicing pathologists.^3,24  Programs like yearlong pathology postsophomore fellowships have been shown to increase medical student interest in pathology through an immersive experience that typically mirrors the first year of pathology residency training.^25–28  Additionally, 1- to 2-month elective programs have proven effective in recruiting medical students to the field of pathology^25,29  (Table 1). Many medical schools also offer pathology interest groups to attract students to the field.³  Despite these efforts, there has recently been a significant decrease in the number of US medical graduates applying to pathology residency programs,^3,23,30,31  with the AAMC reporting a decrease from 563 to 390 US allopathic applicants per year from 2008 to 2019.

Along with the decrease in overall interest in the discipline of pathology, there has been a lag of representation of UIM trainees within the field.^32–34  Overall UIM pathology resident representation has increased at a yearly rate of approximately 0.29%,³⁴  mostly because of increased Hispanic resident representation; however, representation of black/African American and Native American/American Indian/Alaska native/native Hawaiian trainees has showed no significant rate of change since 1995.^32,34  Data from the AAMC and the Electronic Residency Application Service show that from 2015 to 2019 the number of black/African American and Hispanic/Latinx pathology residency applicants has remained stable, with black/African American and Hispanic/Latinx applicants comprising 146 (6.2%) and 133 (5.7%) of 2341 total pathology applicants, respectively, in 2015 versus 134 (6.7%) and 98 (4.9%) of 2013 total applicants in 2019, and the number of Native American/American Indian/Alaska native/native Hawaiian applicants has remained below 0.01% (12 of 2341 total pathology applicants in 2015 versus 4 of 2013 applicants in 2019).³⁰  In a recent analysis by Mortensen et al,⁴  the number of UIM applicants was significantly lower for their pathology training program when compared with US UIM medical student applicants to all AAMC medical specialties from 2015 through 2017.

Trends in racial and ethnic diversity within pathology are similar to those observed broadly within undergraduate and graduate medical education and the broader physician workforce. In 2019, Lett et al³⁵  published a long-term analysis of trends in UIM medical student representation from 2002 to 2012. For their analysis, they assessed for changes in representation quotients (RQs) in undergraduate medical education, defined as “the ratio of proportion of a particular subgroup among the total population of applicants or matriculants relative to the corresponding estimated proportion of that subgroup in the US population.” In their study an RQ greater than 1 indicated overrepresentation, and an RQ less than 1 indicated underrepresentation. They described a trend in decreasing representation of black/African American female medical school matriculants (RQ decrease from 0.65 to 0.53, P = .001), with no significant change in the persistently low number of black/African American male medical school matriculants (RQ stable at approximately 0.36, P = .76).³⁵  Medical students who identify as UIM are more likely to choose primary care fields such as internal medicine, pediatrics, and family medicine.³⁶  Furthermore, representation of UIM physicians practicing in academic settings, including pathologists, is declining.³²  Physicians who identify as UIM, and specifically black/African American and Hispanic/Latinx physicians, comprise only approximately 4% (approximately 40 500 and 43 700 of approximately 956 500 physicians, respectively) of the US physician workforce^37,38  despite comprising approximately 13% and 18% of the US population (43 984 096 and 60 724 311 of 328 239 523 people), respectively.³⁹  The limited pool of potential UIM applicants and lack of diversity among practicing pathologists further strain the recruitment of a diverse trainee population in pathology.

Holistic review, as defined by the AAMC, includes “mission-aligned admissions or selection processes that consider a broad range of factors—experiences, attributes, and academic metrics—when reviewing applications.”⁴⁰  Studies have demonstrated that when implemented effectively, holistic review of residency applicants improves representation of UIM interviewees and ultimately improves diversity within a program.³⁷  A core element of the holistic review is equitable assessment not only of applicant attributes (personal attributes and demographics such as race, ethnicity, socioeconomic status, gender, first-generation college student, etc), but also experiences and academic metrics.⁴⁰  There is an expanding body of literature assessing and emphasizing how the unequitable assessment of applicant academic metrics and experience such as USMLE scores, AOA status, medical school clerkship grades, research experience, number of publications, and school ranking as initial screening tools for residency applicants may disproportionately impact UIM applicants.^{4,7–12,20,37}  Additionally, some studies have demonstrated that implicit bias training of faculty reviewing residency applications may lead to a modification of behavior and a reduction in bias-based actions to further support successful implementation of a holistic review.^{14,16,17,41,42} 

Several studies have revealed disparities in US medical student performances on the USMLE, with UIM students disproportionately scoring lower on USMLE step 1 and step 2 exams when compared with students who do not identify as UIM^4,5,9,11,12  (Table 2). Similar disparities in examination performance exist for the Medical College Admission Test.^43–46  Although higher Medical College Admission Test and USMLE scores have been shown to correlate with higher performance on other written examinations, such as resident in-service examinations and written specialty board examinations,^10,47  performances on standardized exams have not been shown to be an accurate predictor of a student's future ability to provide high-quality patient care or to correlate with interpersonal communication skills, faculty evaluations, or professionalism.^10,44,48,49  In some studies, eliminating the requirement for minimum USMLE scores or lowering the required score increased the number of UIM applicants invited to interview with a given residency program without compromising applicant quality, and may ultimately lead to an increase in program diversity.^10,37 

Similar to their likelihood of achieving higher USMLE scores, students who do not identify as UIM are more likely to be elected for AOA membership^8,11,12,20  (Table 3). Competitive residency programs and medical specialties, such as those in dermatology, plastic surgery, and orthopedics, have strongly associated AOA membership with successful match rates, but have more recently proposed that this metric may inadvertently contribute to a lack of diversity in their respective fields.^8,12,20,50  This disparity may in part be because of the AOA election process. Despite standardized AOA nomination guidelines (only students in the top quartile of their medical school class are eligible for AOA membership; each medical school chapter can select only up to 16% of medical students to be members of AOA; and the students elected into AOA are chosen based on scholastic achievement, professionalism, leadership, community service, and research), AOA election criteria are not standardized. Specifically, AOA chapters may develop chapter-specific metrics and rubrics, possibly furthering the disparities in UIM student representation.^8,51 

Some studies suggest that racial and ethnic disparities may exist in medical school clerkship grading at some institutions, and that students from UIM groups are more likely to report lower clerkship grades^7,11  (Table 4). In one analysis by Lee et al,⁷  80% (110 of 137) of UIM students at their institution received lower grades (P < .001), whereas a second analysis by Low et al¹¹  revealed only 3% (7 of 243) of outstanding Medical Student Performance Evaluation summary words were received by UIM students (P < .001). Low et al¹¹  additionally reported that higher USMLE step 1 examination scores and female gender were associated with higher clinical clerkship grades, and furthermore demonstrated that racial/ethnic disparities in grades received persisted in 4 of the 6 clerkships examined even when adjusting for USMLE step 1 and final written clerkship examination scores. The authors suggested that instructor implicit biases, students' personal circumstances, a racialized learning environment, and instructor characteristics may contribute to the disparities in clinical clerkship grading.¹¹  However, it is important to highlight that both studies noted disparities in grading of students from non-UIM minority groups and clerkship-specific gender discrepancies.^7,11  Although grading trends favoring white students have been suggested in some studies, further studies are needed to determine whether significant racial/ethnic disparities in grading exist in other academic institutions.

Prior research experience is often highly cited among residency programs when selecting residency interview candidates. Data from the 2018 National Resident Matching Program program director survey⁵²  highlighted that 41% (505 of 1233) of residency training programs and 50% (20 of 40) of pathology residency training programs responded that participation and involvement in research is considered when selecting applicants to interview, with a mean importance rating of 3.7 out of 5 and 3.5 out of 5, respectively. Several studies have sought to determine whether racial and ethnic differences exist in research experiences for medical students, graduate medical education program trainees, and practicing physicians. To date, data from several studies demonstrate that physicians from UIM groups are less likely to receive research awards from the National Institutes of Health.^1,5,6,53  These studies demonstrate how inequitable research exposure often begins early in UIM undergraduate education, and is further exacerbated in medical school, where UIM students are less likely to attend a research-intensive medical school (top-40 ranked for National Institutes of Health funding),⁵  are less likely to participate in research electives and/or dedicated research years during medical school, ^5,54  and are less likely to graduate from an MD/PhD program. Interestingly, Andriole et al⁵  also noted that graduates with more than $50 000 in debt were slightly less likely to receive mentored research career development (ie, mentored K) awards; however, this association was not found to be statistically significant.

Letters of recommendation (LORs) are essentially universally required for residency applications. As such, selection committees often rely on letter writers' comments to help determine which applicants to invite for an interview. Few studies exist on examining for potential racial bias in LORs. One such publication by Grimm et al⁵⁵  found that letter writers for radiology residency applicants use less agentic descriptors to describe UIM applicants compared with white and Asian applicants.

There are more published studies assessing for potential gender bias in LORs for residency and fellowship applicants (Table 5). Studies have found that, when compared with letters for female applicants, letters for male applicants had a more authentic tone,^56,57  used more achievement words⁵⁸  and more agentic words^59–61  (ie, superb, intelligent, exceptional),⁵⁹  contained more possessive language,⁶¹  were more likely to contain the words “future leader”⁵⁹  or “future success,”⁶¹  and contained more references to personal drive, knowledge, leadership, and power.^56,58  Letters for female applicants tended to use more terms such as “hardworking,” “compassionate,” and “teacher”^56,62  and more socio-communal phrases (ie, teamwork, helpfulness)⁶²  and were more likely to reference the applicant's spouse's accomplishments.⁶¹  Additionally, a comparison of narrative versus standardized LORs by Friedman et al⁶³  revealed females were less likely to be described as “bright” in narrative LORs. Friedman et al⁶³  also demonstrated females were more likely to have comments on their appearance in narrative LORs, similar to other studies. In contrast, Kobayashi et al⁶⁴  found that letters for female applicants to their orthopedic surgery residency program were slightly longer than those for male applicants and used slightly more “achieve” words, whereas letters for male applicants showed more use of the applicant's first name and used more “young” words. Alternatively, French et al⁶⁵  found that letters for female applicants used more “standout” terms (ie, amazing, outstanding) than those for male applicants, but otherwise found no gender-based differences in LORs, and Messner and Shimahara⁶⁶  found no significant difference in LORs for male and female applicants.

The reality is that we all harbor implicit, often unconscious, biases and that these biases have the potential to affect one's actions toward others. In order to overcome such biases, some institutions and programs have implemented optional implicit bias education curricula. These programs often invite participants to complete a Harvard Implicit Association Test (https://implicit.harvard.edu/) and may include workshops on bias literacy and alternative strategies.^{16–19,22,41,67}  In a study by Carnes et al,¹⁶  all Bias Literacy Workshop attendees who completed a written evaluation found the workshop to be at least somewhat useful, and 74% (134 of 220 attendees) reportedly found it very useful. Other studies have shown that participation in similar interventions led to increases in concern about discrimination,¹⁷  personal awareness of bias,^17,19  and engagement in equity-promoting behaviors.⁴¹ 

Perceived fit or inclusion is often an important factor as applicants rank residency programs. A perceived lack of diversity in the residency population, and also among the faculty, has been described to have a negative impact on a program's ability to recruit UIM trainees.^14,68,69  With this in mind, many programs aiming to improve diversity seek to create an inclusive interview experience by increasing the visibility of UIM trainees and faculty (Table 1), explicitly stressing their commitment to diversity and inclusion, and describing diversity and inclusion efforts within the department and institution.^14,70 

As physicians, we have professed to assure and protect health and well-being for all of our patients. The events of 2020, including the COVID-19 pandemic and the Black Lives Matter movement, have hopefully sensitized us all to how pervasive health disparities remain and how our most vulnerable patients continue to disproportionately suffer in a time when health care is not always inclusive or equitable. The roots of inclusive and equitable health care extend far beyond the delivery of culturally and linguistically inappropriate care in the clinics, originating in the fundamental research and education driving medical decision-making. For pathologists, this is where we are at the greatest risk of perpetuating health disparities. As pathologists, there are innumerable opportunities for us to increase health equity: collaboration with clinical colleagues to ensure appropriate emphasis on health disparities in research and education, rigorously validating diagnostic algorithms adjusted by race, improving patient access to screening and diagnostic laboratory testing, and patient advocacy, just to name a few. Awareness of how we can use our daily pathology practice to mitigate health disparities is the most heightened when there is diversity within our workforce, diversity to bring forth new ideas and vet existing ones under a new lens. The best opportunity pathologists have to accomplish these goals is through assuring and protecting the diversity within our trainee pipeline into pathology.

The pipeline into pathology in 2021 looks a bit different than it did in decades past. In recent years, medical student interest in pathology among US medical graduates has declined, which makes recruitment to pathology in general difficult. There has been much debate on why medical student interest in pathology continues to decline. Possible reasons cited in the literature include limited patient interaction, decreasing exposure to the practice of pathology in medical school curricula, perceived challenges with the pathology job market, and the lack of medical students' identification with the practicing pathologists with whom they interact.^3,24  Each of these suggested reasons may be addressed, at least in part, by increasing exposure to pathology, ideally early in medical school tenure.

As emphasized in the literature, increasing clinical exposure to a field can increase student interest in and ultimately pursuit of careers in that field.^68,71,72  One-year pathology postsophomore fellowships and 1- to 2-month elective programs have the potential to increase student interest in pathology.^25,27,28  Therefore, increasing the availability of these experiences is critical. Although securing dedicated funding and faculty and administrative support to ensure program longevity are essential, direct outreach is equally as important. Direct outreach has the potential to increase the likelihood that early-undifferentiated students will consider these opportunities. Outreach can include creative and fun presentations to student groups and organizations, or even publicizing through social media platforms. Similar to pathology-specific medical student experiences, direct engagement with even younger students in undergraduate and high school programs such as mini–medical school programs can also generate early interest in pathology. These programs have been successful in increasing the number of students who eventually pursue careers in medicine,^73–77  including UIM students,^75,76  and also have an important role of supporting an antiracist culture through community engagement.

However, these experiences address only the broader issue of student interest in pathology. The more significant challenge may be how to increase workforce diversity within an applicant and resident pool that is limited at baseline. Although data suggest that the number of Hispanic pathology residents has increased each year, the numbers of black and Native American/American Indian/Alaska native/native Hawaiian residents have remained stable during the past 25 years.³⁴  Furthermore, the challenges to diversity in pathology extend beyond residency and fellowship training programs and become more significant for the pathology workforce, most notably at higher faculty ranks in academia.^32,69  Limited numbers of UIM senior faculty within a training program may further hurt recruitment efforts, as UIM applicants may perceive limited diversity among faculty and staff as an indirect sign of a noninclusive training environment. Some programs have found success in increasing the visibility of diverse faculty during interviews and emphasizing ongoing diversity and inclusion efforts within their departments.^14,68–70  Additional offerings, such as funded Second Look Weekend activities with involvement by UIM department leadership and trainees, have also been shown to help increase recruitment of UIM residency applicants.⁷⁸ 

Successful implementation of a holistic review process, which promotes diversity as a prerequisite for program excellence, is one effective strategy to enhance diversity within a training program, and hopefully in the pathology workforce long term. As established in the literature, reliance on traditional academic and experiential metrics such as USMLE scores, AOA status, research experiences, and medical school clerkship grades in isolation may have a disproportionately negative impact on UIM applicants selected to receive invitations to interview.^5–13  In summary, successful implementation of a holistic review process should begin with ensuring selection committee awareness of program-specific mission and goals and how diversity serves as a cornerstone to their achievement. A program needs assessment should then follow, to include a review of longitudinal applicant and resident demographics and the diversity of the selection committee. After a program's baseline and needs are established, the selection committee should then perform a critical review of the existing applicant screening and selection process. This review should pay specific attention to how the committee considers applicant attributes (ie, race, ethnicity, ties to a marginalized community, socioeconomic status, disability) as well as traditional academic metrics (examination scores, class rank, AOA status, LORs) and how the committee can confirm it is equitably assessing each domain.

At a minimum, the selection process review should establish action items and identify the necessary internal institutional and/or external resources and persons responsible for follow-up. Internal institutional resources may include institution-wide graduate medical education or faculty diversity committees or DEI leadership. If the selection process review identifies a need for committee diversity that may not exist within the program, these resources could assist in identifying the appropriate faculty or trainees outside a program. External resources include the AAMC, which offers numerous free tools and resources to guide programs in holistic review implementation (https://www.aamc.org/services/member-capacity-building/holistic-review).⁴⁰ 

Action items may vary widely and will be program specific, but could include de-emphasis on minimum USMLE examination scores or even interviewer blinding to USMLE examination scores, revision of applicant and/or interview scoring rubrics, and/or unconscious bias training for committee members.^37,79  Some programs have found that lowering or eliminating the minimum USMLE step 1 examination score requirement for an interview invitation, as well as placing less emphasis on AOA status and research experiences, may increase the number of UIM applicants invited to interview with and ultimately match into their program.^10,37  This strategy may be useful in helping programs to at least consider more UIM applicants for interview. The slated transition of the USMLE step 1 examination to pass/fail grading may expedite the change in the weight of examination scores on the evaluation of residency applicants. However, it is unclear whether this transition will lead to a de-emphasis of standardized test scores, or whether it will lead programs to place more weight on USMLE step 2 clinical knowledge scores.

Another example of an action item for implementing a holistic review process is increasing committee awareness of potential bias in subjective academic metrics such as LORs and narrative student evaluations. Although some studies have shown gender-based differences in LORs, demonstrating that letters for male applicants often use a more authentic tone and use more achievement and power words than those written for female applicants,^56–58  few studies examine potential racial and ethnic biases in LORs. At least one training program has found that letter writers for their applicants describe UIM applicants as less agentic than white and Asian applicants.⁵⁵  Further studies are needed to determine whether racial and ethnic disparities exist in LORs for other specialties, including pathology.

Nonetheless, awareness of how some LOR writers avoid use of bias-driven language is important. Writers of LORs aiming to avoid bias-driven language may construct letters that use an authentic tone and/or agentic language (ie, describing applicants as self-motivated), avoid stereotypes, emphasize accomplishments, and avoid mentioning an applicant's personal life. There are several resources available to guide letter writers to avoid writing biased LORs,^80–82  including a gender bias calculator,⁸³  which may be applicable to racial bias as well. As an evaluator of LORs, it is important to keep in mind possible bias using the aforementioned guidelines.

Acknowledgement of potential implicit biases and training in ways to circumvent the effects of such biases are excellent tools to encourage a more holistic application review process and create a more inclusive training environment. The Harvard Implicit Association Tests, which cover many different topics including racial/ethnic and gender biases, have been used by many programs to initiate a conversation surrounding implicit biases. Programs that have hosted workshops on bias literacy and strategies to combat implicit biases have reported successful outcomes.^{16–19,22,41,67}  Such programs are useful, particularly for faculty and trainees who will be reviewing residency applications and working directly with trainees, to encourage an inclusive climate. The Implicit Association Tests seem to be most effective when testing is coupled with reflective discussion about individual's results and ways to increase awareness of potential biases.⁸⁴  Ideally, all individuals involved in the interview process, including faculty, staff, and residents, should participate in implicit bias training, and it is important to emphasize the importance of the testing and its role in the applicant review process. At the very least, the program directors and any other persons directly involved in the applicant review or final ranking process should participate. Programs may offer testing during selection committee meetings or via email to be taken on an individual's own time.

However inclusive a program culture may be, if diversity is a long-term goal, then programs are encouraged to consider ways to increase equity. Although the literature frequently discusses inequities in education and access for UIM students, socioeconomic inequity has the potential to limit a student's future success as severely. On average, medical student indebtedness often exceeds $150 000,^85,86  with studies demonstrating that UIM students are more likely to graduate with a higher amount of student indebtedness compared with non-UIM students.^50,85–87  When education indebtedness is combined with additional fees and costs of visiting rotations and residency applications, visiting rotations become less accessible to UIM students, despite being critical to the medical student residency specialty and/or program selection process. Financially inequitable access to away rotations may put UIM students at a further disadvantage, as pathology residency program directors often cite audition elective/rotations within departments when selecting applicants to interview (78%; 31 of 40 programs; citing factor, 4.0 average rating) and ranking applicants (59%; 23 of 40 programs; citing factor, 4.3 average rating).⁵²  This is an important consideration as UIM students are less likely to attend medical schools associated with large academic and research centers,⁵  and smaller medical schools may not have an associated pathology training program.

The overall unfortunate consequences are that UIM students may not be able to obtain adequate pathology exposure prior to applying to residency programs, and may be at risk of having limited access to the networking opportunities gained through away rotations. Furthermore, as students who carry more student loan debt are less likely to participate in research electives or training years,⁵  students from UIM groups may be regarded as less competitive residency applicants in a traditional selection process. Although addressing the issue of student indebtedness is beyond the scope of most pathology residency programs, consideration of the financial challenges faced by many students from UIM groups may further support an equitable and holistic application review process and ultimately aid in increasing the representation of UIM trainees.²⁹ 

With the incredible tragedy the world has suffered during the COVID-19 pandemic, one positive has been how we have learned that when available, technology has the capacity to increase equity and access. Novel virtual away electives, such as the rotation offered by the Johns Hopkins University Department of Pathology, allow students to preview and sign out with a pathologist in real time. Such rotations deliver a similar learning experience to in-person rotations at a fraction of the cost, and may serve to help bridge the financial gap when students are considering the financial burden of away rotations. Another solution is having programs provide funded rotations, including coverage of travel, registration, and lodging expenses, for UIM students.²⁹ 

The field of pathology faces shared challenges with ensuring and protecting diversity in graduate medical education pathology training, and there are many strategies that may be used to increase diversity and encourage an inclusive and equitable training environment (Table 6). Increasing graduate medical education trainee diversity will ultimately lead to a long-term increase in faculty diversity, which will in turn serve as a positive force in continuing to recruit a diverse trainee cohort and reducing health disparities. Efforts such as increasing exposure to the field of pathology early in medical education, offering elective and immersive opportunities to increase the pipeline to pathology, implementing holistic pathology trainee application review, and encouraging implicit bias training and bias literacy have all proven effective in promoting DEI. A genuine and concerted effort by pathology training programs to implement the aforementioned efforts has the potential to make even more significant improvements in diversifying the pathology workforce.