# Does clinical exposure to different skin tones during training improve diagnostic ability?

## TL;DR

Medical students and doctors struggle to diagnose skin conditions in people of color, even with clinical exposure, suggesting education materials need more diverse skin tones.

## Contribution

Shows that clinical exposure to different skin tones during training does not improve diagnostic accuracy for skin of color.

## Key findings

- Both groups performed better in diagnosing white skin than skin of color.
- Exposure to non-white patients did not significantly improve diagnostic ability in skin of color.
- Educational resources must include diverse skin tones to reduce diagnostic disparities.

## Abstract

Previous studies have shown that medical students demonstrate poorer performance when diagnosing pathology in skin of colour (SOC) compared to white skin (WS); it is important to understand the reasons underpinning this. If not addressed, poorer differential diagnostic ability in certain skin tones could entrench existing racial inequities in health care. We investigated whether exposure to a predominant patient skin colour during clinical practice affects diagnostic ability in WS and SOC.

Participants were international medical graduates (IMGs) and medical students from Imperial College London and the University of Dundee, recruited between January and May 2024. Participants were divided into two groups, based on whether they were predominantly exposed to white patients (WP) or non‐white patients (NWP) in their practice. Participants sat a dermatology quiz, in which they were asked to provide a diagnosis for 22 image‐based vignettes, covering 11 clinical presentations, each shown in WS and SOC. For each of the WP and NWP exposed groups, we compared their diagnostic ability in WS and SOC presentations.

A total of 411 participants were analysed; 187 predominantly exposed to WP and 224 predominantly exposed to NWP. Both groups demonstrated a statistically significantly better diagnostic ability in WS compared to SOC (p < 0.01). Overall, there was no significant difference in differential diagnostic ability in WS and SOC between the WP‐exposed and NWP‐exposed groups (p = 0.731).

Regardless of the predominant patient skin colour participants saw in their practice, participants were worse at diagnosing pathology in SOC. This highlights that clinical exposure to SOC is not sufficient to mitigate clinicians' inferior diagnostic ability in non‐white skin tones. Therefore, effort must be made to improve the diversity of skin colours represented in medical education resources, to improve clinicians' familiarity with pathology in different skin tones and minimize the risk of patients being misdiagnosed due to their skin colour.

Clinical exposure isn't enough. Medical students and IMGs diagnose skin of colour less accurately, highlighting the need for better representation in education.