When Faces Don't Match: Cross‑Race Misidentification in an Emergency Room
A real-world example of Cross-race effect in action
Context
A busy urban emergency department serves a diverse patient population and often relies on quick visual recognition and verbal confirmation from accompanying family members for rapid triage. Staff turnover is high, and many clinicians are from different racial and ethnic backgrounds than the patients they see most often.
Situation
A middle-aged man of East Asian descent arrived by ambulance after a fainting episode. He was semi-conscious and unable to speak clearly; a person claiming to be his brother provided a name and story, and a triage nurse—who was White and unfamiliar with East Asian facial features—relied on that information without obtaining the patient's ID band or checking the electronic photo on file.
The bias in action
The triage nurse's perceptual familiarity favored faces of her own race, making it harder for her to detect that the unconscious patient did not match the photo associated with the name given. Because the nurse relied on the accompanying relative's statement and her own quick facial recognition, she skipped the usual two-source verification (photo + wristband). Later, when clinicians prepared a high-risk medication, they used the recorded patient identity; the mismatch was only noticed after medication had been pulled but before administration. The error chain started with an unchallenged visual recognition and a missed verification step influenced by the cross‑race effect.
Outcome
The patient did not receive the intended medication but experienced a 45-minute delay in receiving the correct treatment while the team re-verified identity and clarified family relationships. The hospital opened an incident review, the accompanying relative admitted he was a friend who had used the patient's name, and the ED updated its immediate verification procedures. The patient recovered without lasting harm, but the event prompted a formal safety audit.