All dermatology nonskin of color (SOC) atlases have a greater percentage of SOC images for stigmatizing disorders compared with the percentage of all images, researchers reported in study findings published in the Journal of the American Academy of Dermatology.
The qualitative study evaluated SOC images in 10 dermatology atlases.
Each image was classified by 1 reviewer as light skin (1-3), medium skin (4-5), or dark skin (6-10) according to the 10-point Massey-Martin Skin Color Scale, with the medium and dark groups counted as SOC images. Stigmatizing disorders were analyzed separately and included sexually transmitted infections, drug abuse, physical abuse, and psychiatric illness.
A total of 11,328 clinical images were analyzed. About half of the images were in the Andrews (2943) and du Vivier (2119) atlases, and fewer than 12% of all images were in SOC atlases. Among the images, 63.87% were of light skin, 21.12% were of medium skin, and 15.02% were of dark skin.
In the 6 non-SOC atlases, the representation of light skin tone images included Fitzpatrick (87.45%), Callen 1st edition (80%), Goodheart (76.47%), Callen 2nd edition (75.74%), du Vivier (70.22%), and Andrews (66.20%). The Andrews atlas had the highest overall SOC representation of the non-SOC atlases (33.7%) and the largest number of SOC images (983), accounting for about one-fourth of all SOC images (4093) from all of the atlases.
Overall images for stigmatizing disorders (498) included 48.3% of light skin, 25.5% of medium skin, and 26.2% of dark skin. All non-SOC atlases had a greater percentage of SOC images in these disorders vs their corresponding percentage of all images.
Study limitations include the subjectiveness of assessing skin tone, differing image quality, and environmental lighting, noted the study authors.
“It is necessary to recognize the presence of invisibility bias, which is the lasting effect of under-representation in education,” stated the investigators. “This can create and perpetuate stereotypes as is concerning with the higher percentage of stigmatized conditions being SOC.”
These stereotypes can result in delayed diagnosis and poorer patient outcomes, and educational materials need to be holistically expansive, noted the researchers.
“We recognize authors may be limited with image selection due to their patient population and space constraints; however, substantial, collaborative efforts should be made by authors and publishers to create all-inclusive resources,” the investigators commented.
Disclosure: A study author is a coauthor of 1 of the atlases. Please see the original reference for a full list of authors’ disclosures.
Hereford B, James WD. Skin of color representation in dermatology atlases. J Am Acad Dermatol. Published online August 8, 2022. doi:10.1016/j.jaad.2022.08.003