Skin color representation in orthopedic literature

Abstract Introduction Underrepresentation of skin color in educational materials can undermine diagnostic accuracy and contribute to disparities in orthopedic care for minority populations. While underrepresentation of darker skin types has already been identified in orthopedic literature, this study assessed whether underrepresentation is apparent in open access journals, is uniform across all orthopedic subspecialties, differs between figure types, and has shifted over time as awareness of skin of color disparities has grown. Materials and methods 40,674 figures were accessed from 14 open-access orthopedic journals and assigned a subspecialty and a country of origin based on author affiliation. Automatic filters narrowed the dataset to 4,809 figures. Each figure was processed by a GPT-4o artificial intelligence (AI) pipeline to obtain skin presence (yes/no), figure type (clinical, intraoperative, illustration), Fitzpatrick skin type (I-VI), and a confidence score (1–5). Agreement was high with human review for skin presence and figure type and moderate with a board-certified dermatologist for skin tone, with disagreement biased toward darker ratings. 2,025 skin-containing figures were analyzed. Results Among 2,025 figures, most depicted light skin (65.6%, Fitzpatrick I-II), 29.6% medium (III-IV), and only 4.8% dark (V-VI). This demonstrated severe overrepresentation of light skin ( p < 0.001) and severe underrepresentation of dark skin ( p < 0.001) relative to both a diverse-population reference and US population estimates. Underrepresentation of dark skin was consistent across subspecialties and figure types. Although subgroup analysis revealed modest improvement in dark-skin representation from 2015 to 2024 in non-U.S. figures (r = + 0.72, p = 0.02), the year-by-region interaction was not significant ( p = 0.24). Conclusions Within open-access orthopedic literature, dark skin is consistently underrepresented and falls short of US and diverse-population benchmarks, regardless of figure type or subspecialty. This systemic pattern narrows the implicit norm of who an orthopedic patient appears to be and warrants editorial attention given documented disparities in orthopedic care.