Assessing implicit ethnic biases and feedback preferences among trainee GPs
V Lorant, C DuveauAbstract
Background
Healthcare providers’ decisions can be biased, sometimes resulting in less favourable care for patients with a migration background (PMB). Implicit and explicit biases contribute to ethnic disparities in healthcare, yet their determinants and potential levers for change remain poorly understood. This study examines the determinants of implicit and explicit biases toward PMB and assesses medical trainees’ willingness to engage with bias through personalised feedback.
Methods
Between 2021 and 2025, five cohorts of general practice trainees at a Belgian university were invited to complete an Implicit Association Test (IAT) and the Hudelson scale measuring explicit bias toward PMB. The IAT contrasted patients with Belgian versus North African names and faces. At the end of the survey, participants were asked whether they wished to receive individualised feedback on their results.
Findings
A total of 680 trainees participated. Overall, 68% showed moderate to high implicit bias unfavourable to North African patients, with no variation over time. Across cohorts, 60% of participants requested individual feedback. Willingness to receive feedback was not associated with implicit bias (D-score), but was lower among trainees with stronger explicit biases.
Conclusions
Implicit and explicit biases toward PMB persist among future primary care providers. Engagement with personalised feedback appears to be shaped more by explicit attitudes than by implicit bias levels. These findings highlight the importance of addressing both bias awareness and underlying attitudes in medical education.
Key messages
• Most GP trainees show implicit bias against North-African patients. Only 60% seek feedback, and engagement is driven more by explicit attitudes than by implicit bias levels.
• Implicit and explicit biases persist among future GPs. Willingness to receive bias feedback is limited and reflects explicit beliefs, highlighting barriers to engagement in bias-reduction efforts.