Study Finds Biopsy Decisions, Diagnoses Differ by Skin Type

BOSTON ― Among dermatology residents and attending dermatologists, rates of diagnostic accuracy and appropriate biopsy recommendations were significantly lower for patients with skin of color compared with White patients, new research shows.

Dr Loren Krueger

"Our findings suggest diagnostic biases based on skin color exist in dermatology practice," lead author Loren Krueger, MD, assistant professor in the Department of Dermatology, Emory University School of Medicine, Atlanta, said at the Annual Meeting of the Skin of Color Society Scientific Symposium (SOCS) 2022. "A lower likelihood of biopsy of malignancy in darker skin types could contribute to disparities in cutaneous malignancies," she added.

Disparities in dermatologic care among Black patients in comparison with White patients have been well documented. Recent evidence includes a 2020 study that showed significant shortcomings among medical students in correctly diagnosing squamous cell carcinoma, urticaria, and atopic dermatitis for patients with skin of color.

"It's no secret that our images do not accurately or in the right quantity include skin of color," Krueger said. "Yet few papers talk about how these biases actually impact our care. Importantly, this study demonstrates that diagnostic bias develops as early as the medical student level."

To further investigate the role of skin color in the assessment of neoplastic and inflammatory skin conditions and decisions to perform biopsy, Krueger and her colleagues surveyed 144 dermatology residents and attending dermatologists to evaluate their clinical decision-making skills in assessing skin conditions for patients with lighter skin and those with darker skin. Almost 80% (113) provided complete responses and were included in the study.

For the survey, participants were shown photos of 10 neoplastic and 10 inflammatory skin conditions. Each image was matched in lighter (skin types I–II) and darker (skin types IV–VI) skinned patients in random order. Participants were asked to identify the suspected underlying etiology (neoplastic–benign, neoplastic–malignant, papulosquamous, lichenoid, infectious, bullous, or no suspected etiology) and whether they would choose to perform biopsy for the pictured condition.

Overall, their responses showed a slightly higher probability of recommending a biopsy for patients with skin types IV–V (odds ratio [OR], 1.18; P = .054).

However, respondents were more than twice as likely to recommend a biopsy for benign neoplasms for patients with skin of color in comparison with those with lighter skin types (OR, 2.57; P < .0001). They were significantly less likely to recommend a biopsy for a malignant neoplasm for patients with skin of color (OR, 0.42; P < .0001).

In addition, the correct etiology was much more commonly missed in diagnosing patients with skin of color, even after adjusting for years in dermatology practice (OR, 0.569; P < .0001).

Conversely, respondents were significantly less likely to recommend a biopsy for benign neoplasms and were more likely to recommend a biopsy for malignant neoplasms among White patients. Etiology was more commonly correct.

The findings underscore that "for skin of color patients, you're more likely to have a benign neoplasm biopsied, you're less likely to have a malignant neoplasm biopsied, and more often, your etiology may be missed," Krueger said at the meeting.

Of note, while 45% of respondents were dermatology residents or fellows, 20.4% had 1 to 5 years of experience, and about 28% had 10 to more than 25 years of experience.

And while 75% of the dermatology residents, fellows, and attendings were White, there was no difference in the probability of correctly identifying the underlying etiology in dark or light skin types based on the provider's self-identified race.

Importantly, the patterns in the study of diagnostic discrepancies are reflected in broader dermatologic outcomes. The 5-year melanoma survival rate is 74.1% among Black patients and 92.9% among White patients. Krueger referred to data showing that only 52.6% of Black patients have stage I melanoma at diagnosis, whereas among White patients, the rate is much higher, at 75.9%.

"We know skin malignancy can be more aggressive and late-stage in skin of color populations, leading to increased morbidity and later stage at initial diagnosis," Krueger told Medscape Medical News. "We routinely attribute this to limited access to care and lack of awareness on skin malignancy. However, we have no evidence on how we, as dermatologists, may be playing a role."

Furthermore, the decision to perform biopsy or not can affect the size and stage at diagnosis of a cutaneous malignancy, she noted.

Key changes needed to prevent the disparities ― and their implications ― should start at the training level, she emphasized. "I would love to see increased photo representation in training materials ― this is a great place to start," Krueger told Medscape.

In addition, "encouraging medical students, residents, and dermatologists to learn from skin of color experts is vital," she said. "We should also provide hands-on experience and training with diverse patient populations."

The first step to addressing biases "is to acknowledge they exist," Krueger added. "I am hopeful this inspires others to continue to investigate these biases, as well as how we can eliminate them."

The study was funded by the Rudin Resident Research Award. The authors have disclosed no relevant financial relationships.

Annual Meeting of the Skin of Color Society Scientific Symposium (SOCS) 2022: Oral abstract. Presented March 24, 2022.

For more news, follow Medscape on Facebook, Twitter, Instagram, and YouTube.

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.