Gender-Dependent Diagnostic Substitution in LLM Medical Triage: Same Symptoms, Unequal Urgency
A recent study investigates gender-dependent disparities in medical triage recommendations made by large language models. The research shows that young women receive significantly lower emergency room referral rates compared to age-matched men for identical neurological symptoms. This disparity is attributed to diagnostic substitution, where models favor gender-associated diagnoses, leading to lower urgency care for female patients.
- ▪The study analyzed triage recommendations from three large language models: Gemini, Claude, and GPT.
- ▪Young women were found to have lower emergency room referral rates than young men, with disparities reaching statistical significance.
- ▪The disparity in triage urgency disappears for patients aged 65 and older.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2606.03641 (cs) [Submitted on 2 Jun 2026] Title:Gender-Dependent Diagnostic Substitution in LLM Medical Triage: Same Symptoms, Unequal Urgency Authors:Qi Han Wong View a PDF of the paper titled Gender-Dependent Diagnostic Substitution in LLM Medical Triage: Same Symptoms, Unequal Urgency, by Qi Han Wong View PDF HTML (experimental) Abstract:We investigate whether large language models produce different medical triage recommendations for identical neurological symptoms when only the patient's stated gender and age vary.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
