Behind the Lens: Research Gaps That Impact Skin of Color

From pavements-turned-murals to White Coats for Black Lives -- a bold declaration by physicians to dismantle racism in medicine -- the Black Lives Matter movement is one of many defining elements of 2020. As protests ignited across the nation, the American Academy of Dermatology (AAD) stood in solidarity with Black communities calling for change and equality.

For many years, dermatologists have worked toward spreading awareness and uncovering disparities in care, namely the inequities that impact their skin of color patients. According to the Skin of Color Society (SOCS), in roughly 30 years, half of the U.S. population will have skin of color, underscoring the need for dermatologists to have adequate training in treating common diseases that present differently on dark skin. Yet, dermatology training has traditionally focused on diagnosing and managing common skin conditions in lightly pigmented skin.

In 2020, the Academy interviewed AAD members with expertise and interest in treating patients with skin of color for their input on how best to create culturally responsive and inclusive public education content. Common themes emerged, specifically on how the lack of images of skin conditions on darker skin impacts proper treatment for skin of color patients, and the need for medical education for the future generation of dermatologists.

Here, experts explore the lack of equitable training and research in the specialty and its impact on quality of care for skin of color patients.

Images with darker skin

Skin conditions that present in a red or pink hue in light skin can be subtle and harder to see in dark skin. Physicians who do not work with patients with skin of color in their residency or haven't been trained with images that show diseases on darker pigmented skin are more likely to misdiagnose cutaneous disorders in people of color. Access to high-quality images of conditions in dark skin and a robust skin of color curriculum can help alleviate this issue.

But a recent analysis of commonly referenced dermatology textbooks revealed that only 4% to 18% of images showed skin conditions in darkly pigmented skin. The analysis, which explored textbooks used to educate trainees, dermatologists, and generalists, found they have limited representations of common dermatologic conditions in darker skin. For example, only 50% of dermatology textbooks in the analysis featured an image of dark-skinned patients with acne.

Jean Bolognia, MD, professor of dermatology at the Yale School of Medicine, has spent more than 20 years editing the widely used textbook, Dermatology, which was among the six textbooks analyzed in the study.

"Dermatology is so visual," she said. "Our textbook often features the same disease in two different skin tones, side-by-side."

Bolognia, who updated Dermatology in 2018, said that side-by-side images of skin conditions in light skin and dark skin help students recognize the spectrum -- appreciating a blending of colors when melanin is present -- and illustrate the changes in the disease from pink in light skin, to red-brown in pigmented skin.

"In general, there is an underrepresentation of skin of color in most areas," said Jules B. Lipoff, MD, assistant professor of clinical dermatology at the University of Pennsylvania. "There are certain diseases where there isn't that problem like in vitiligo. You tend to see vitiligo better when photographed on darker skin."

However, there are alarming cases where overrepresentation of skin conditions in people of color perpetuates stereotypes. Black skin, for example, is more likely to be used to display how sexually transmitted diseases present on skin.

"You can't justify overrepresentation in other areas like sexually transmitted infections," Lipoff said. "STI diseases don't photograph any better [on Black patients versus white patients]."

An analysis, led by Jenna Lester, MD, assistant professor of dermatology at the University of California San Francisco, found nearly 30% of images of infectious diseases used images of darker skin, but the number doubled for sexually transmitted diseases presented on dark skin.

Lester also published an analysis on the absence of images of skin of color in publications of COVID-19 skin manifestations. Lipoff argues that there are many ways to rationalize these findings, but some are a result of longstanding access to care issues prevalent in underserved communities.

"I don't think it's really anything specific to dermatology," Lester said. "It's another example of how Black patients are less likely to be tested, less likely to get care, and less likely to see a dermatologist who has the wherewithal to take pictures and submit them."

COVID-19 and race

For an encompassing approach to dermatologic conditions in skin of color, Lipoff encourages residents and dermatologists to reference specialized texts focusing on patients with skin of color, notably Dermatology for Skin of Color, co-authored by the late A. Paul Kelly, MD, and Susan Taylor, MD, AAD vice president and director of diversity, equity, and inclusion for the Department of Dermatology at the University of Pennsylvania in Philadelphia.

"In my early years of practice, there were many patients who came to me and relayed the fact that other dermatologists couldn't make the diagnosis," said Taylor, reflecting on one of the defining factors that inspired Dermatology for Skin of Color. "This could span from rashes to hair loss ... dermatologists weren't able to correctly diagnose the disease."

Taylor noted that awareness of the nuances of disorders in skin of color patients will improve patient care and patient satisfaction with their dermatologist, and will overall strengthen the specialty of dermatology.

"We are beginning to have the necessary resources to educate all dermatologists," she said.

Research into machine learning (ML) has raised awareness about the biases in health care that surface when artificial intelligence (AI) tools are not developed with inclusivity at the forefront. A JAMA Dermatology paper co-published by Adewole S. Adamson, MD, MPP, addresses machine learning algorithms, most of which have been developed in lighter skin types.

"Many machine learning algorithms are developed using thousands of images of skin diseases," said Adamson, assistant professor and dermatologist in the department of internal medicine at Dell Medical School at the University of Texas at Austin. "If these images do not include diverse skin types, machine learning algorithms are at risk of not being able to correctly identify skin disease across everyone."

AI or augmented intelligence (AuI) tools may enhance physician care and innovative ML software can potentially distinguish between images of benign and malignant melanomas with accuracy comparable to a board-certified dermatologist. However, most ML software are largely learning to detect the lesions on light skin. According to Adamson, this technology gap correlates with the lack of imagery of skin conditions on dark skin in textbooks and databases.

"There does not exist a large repository of images of skin conditions in skin of color that can readily be used to train machine learning algorithms," he said. "We need to make sure that our dermatology textbooks, board questions, and other educational material are inclusive of all skin types."

Training and expertise

In 2004, alongside her mentor Vincent DeLeo, MD, Taylor co-founded the SOCS, bringing together like-minded people from a variety of ethnicities and races who were interested in developing the area of expertise in skin of color dermatology. The society is committed to the education of physicians and the public on dermatologic health issues related to skin of color, supporting research, and mentorship of young, diverse medical students.

Conditions like alopecia, post-inflammatory hyperpigmentation, and keloids are more common in skin of color patients and present differently on darker skin types. Misdiagnosis occurs when dermatologists are not familiar with how these common conditions, among others, look on darker skin because their curriculum lacked training in this area. In some cases, residents train in predominantly white populations and aren't exposed to diverse patients and their consequent skin health needs.

"We need to examine even the most ingrained things within our specialty," said Lipoff. "A lot of things are held over based on tradition and not necessarily by what would be the best for the whole population. I was lucky that I studied in a predominantly darker skinned population in the Bronx."

Location-specific effects persist, so dermatology faculty members are incorporating related training to help develop culturally competent dermatologists. "I'm in mid-west Iowa, where our population is not as diverse," said Nkanyezi Ngwenyama Ferguson, MD, chair of the Academy's Diversity Committee and clinical associate professor at the University of Iowa Hospitals and Clinics.

Ferguson oversees the University of Iowa's Ethnic Skin Care Clinic, the first clinic dedicated to the conditions impacting skin of color patients in Iowa. "From a training perspective, it was very important that we train our future dermatologists for practice all over the country in settings where there are diverse patient populations."

Ferguson is event co-chair of the AAD's Diversity Champion Workshop, an annual event that explores best practices in recruitment to ultimately increase the number of practicing underrepresented minority dermatologists.

Diversifying medicine to improve access to care

"Understanding why it's important to diversify the dermatology workforce is important," said Ferguson, acknowledging data that suggest workforce diversity within the specialty, and medicine at-large, reduces health care disparities.

The residue of structural racism has a lot to do with why individuals from underserved populations face barriers to health care. Lack of insurance coverage, and limited specialty care under Medicaid, an insurance option that mostly covers people from low-income communities, many of whom are minorities, are obvious barriers.

"The fact that the specialty does not accept Medicaid as well [as other specialties] is indirectly discriminatory," said Lipoff. "I understand that there are financial pressures, and it's not easy, but we have a moral and ethical obligation to make sure that we're providing access to those who need it."

Diversifying the specialty workforce can help to remedy the access issue. Research shows that underrepresented minority physicians are more likely to practice in underserved populations, creating a culturally competent experience for marginalized patients and, thus, improving outcomes. In fact, it is documented that patient-physician race-concordant visits, in which the patient and doctor are of the same race, result in better outcomes and higher patient compliance. This is especially evident when there is a language gap that can be closed when the doctor can communicate in the patient's native language. Other studies find that even during race-discordant visits, cultural competency can help fill the gap in race-discordant visits.

"As a white doctor taking care of a Black woman who has an issue with her hair, I have to be aware of how that lands," said Lipoff. "For instance, telling a Black woman that she should be washing her hair every day shows a lack of knowledge and an assumption that people, regardless of culture, do the same things."

Increased cultural competence is just one element of building trust among skin of color patients, particularly when it comes to increasing minority participation for research studies and clinical trials. Lipoff advocates for calling out systemic racism in medicine as an essential part of improving medicine's trustworthiness in order to earn and be deserving of the Black community's support.

"The Tuskegee experiment was not an isolated issue but one that was representative of a pattern of abuses on Black bodies throughout history," he said, referencing the United States Public Health Service's Tuskegee Study of Untreated Syphilis in the Negro Male that left the disease diagnosis concealed and intentionally untreated for decades in a group of poor, Black men.

In addition to mistrust, Ferguson noted lack of access plays a role. She warned that if researchers are not intentional about recruiting a diverse patient population in clinical trials, skin of color patients will continue to be marginalized.

"If you're not including a diverse patient population in your clinical trials, that's a lost opportunity," she said. "When we're talking about melanoma, for example, and how these are diagnosed [in African Americans] later and have increased morbidity and mortality [in African Americans], we can go beyond knowing there's a disparity. By including a broad range of diverse research participation, we can dig into how to actually reduce the disparity."

Tackling the disparities looming in the specialty is far from a one-off project that is easily solved. For decades, pioneering dermatologists like Taylor, and many others, have taken careful consideration into the nuances that, when intentionally addressed, help to narrow gaps in care for skin of color patients. Yet, most physicians are likely to agree that there is no one answer to get the "work" done to ensure all people have equitable access to care. But research, awareness, education, and mentorship -- of young people motivated to take on a career in medicine when physicians who look like them cheer them on -- have proven to make a difference.