mlsanityData to 14 Oct 2026

EQUITRIAGE: A Fairness Audit of Gender Bias in LLM-Based Emergency Department Triage

arXiv 2605.03998 ↗First posted 5 May 2026cs.CL cs.CYcited by 3

Not on the map. Why?

Abstract (arXiv)

Emergency department triage assigns patients an acuity score that determines treatment priority, and clinical evidence documents persistent gender disparities in human acuity assessment. As hospitals pilot large language models (LLMs) as triage decision support, a critical question is whether these models reproduce or mitigate known biases. We present EQUITRIAGE, a fairness audit of LLM-based ESI assignment evaluating five models (Gemini-3-Flash, Nemotron-3-Super, DeepSeek-V3.1, Mistral-Small-3.2, GPT-4.1-Nano) across 374,275 evaluations on 18,714 MIMIC-IV-ED vignettes under four prompt strategies. Of 9,368 originals, 9,346 are paired with a gender-swapped counterfactual. All five models produced flip rates above a pre-registered 5% threshold (9.9% to 43.8%). Two showed directional female undertriage (DeepSeek F/M 2.15:1, Gemini 1.34:1); two were near-parity; one had high sensitivity with weak male-direction asymmetry. DeepSeek's directional bias coexisted with a low outcome-linked calibration gap (0.013 against MIMIC-IV admission), a Chouldechova-style dissociation between within-group calibration and between-pair counterfactual invariance. Demographic blinding reduced Gemini's flip rate to 0.5%; an age-preserving blind variant left DeepSeek with residual F/M 1.25, implicating age as a residual channel. Chain-of-thought prompting degraded accuracy for all five models. A two-model ablation reveals opposite underlying mechanisms for the same directional phenotype: in Gemini the signal is emergent in the combined name+gender swap, while in DeepSeek the gender token alone carries it. EQUITRIAGE shows that group parity, counterfactual invariance, and gender calibration are distinct fairness properties, that intervention effectiveness is model-dependent, and that per-model counterfactual auditing should precede clinical deployment.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 25 of its 62 references that are in the map. First posted on arXiv 5 May 2026.

Builds on

25 of 62 references.

References · 25

  1. 2016Equality of Opportunity in Supervised Learningcited by 2,007
  2. 2017Counterfactual Fairnesscited by 804
  3. 2016Fair prediction with disparate impact: A study of bias in recidivism prediction instrumentscited by 672
  4. 2017Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraintscited by 420
  5. 2024Non-Determinism of "Deterministic" LLM Settingscited by 92
  6. 2020Hurtful Words: Quantifying Biases in Clinical Contextual Word Embeddingscited by 79
  7. 2024What's in a Name? Auditing Large Language Models for Race and Gender Biascited by 51
  8. 2024Evaluating Gender Bias in Large Language Models via Chain-of-Thought Promptingcited by 50
  9. 2025ER-Reason: A Benchmark Dataset for LLM Clinical Reasoning in the Emergency Roomcited by 22
  10. 2024Locating and Mitigating Gender Bias in Large Language Modelscited by 12
  11. 2024Development of a Large Language Model-based Multi-Agent Clinical Decision Support System for Korean Triage and Acuity Scale (KTAS)-Based Triage and Treatment Planning in Emergency Departmentscited by 10
  12. 2025Bias in Large Language Models Across Clinical Applications: A Systematic Reviewcited by 6
13 more
  1. 2025AI Agents for Conversational Patient Triage: Preliminary Simulation-Based Evaluation with Real-World EHR Datacited by 6
  2. 2025From Promising Capability to Pervasive Bias: Assessing Large Language Models for Emergency Department Triagecited by 5
  3. 2024How Can We Diagnose and Treat Bias in Large Language Models for Clinical Decision-Making?cited by 4
  4. 2025MEDEQUALQA: Evaluating Biases in LLMs with Counterfactual Reasoningcited by 3
  5. 2024Fair Machine Learning for Healthcare Requires Recognizing the Intersectionality of Sociodemographic Factors, a Case Studycited by 2
  6. 2024Fairness at Every Intersection: Uncovering and Mitigating Intersectional Biases in Multimodal Clinical Predictionscited by 2
  7. 2025mFARM: Towards Multi-Faceted Fairness Assessment based on HARMs in Clinical Decision Supportcited by 2
  8. 2026Uncovering Latent Bias in LLM-Based Emergency Department Triage Through Proxy Variablescited by 2
  9. 2023Multi-Modal Perceiver Language Model for Outcome Prediction in Emergency Departmentcited by 1
  10. 2025Bias Detection in Emergency Psychiatry: Linking Negative Language to Diagnostic Disparitiescited by 1
  11. 2025Gender Bias in Large Language Models for Healthcare: Assignment Consistency and Clinical Implicationscited by 1
  12. 2025Dr. Bias: Social Disparities in AI-Powered Medical Guidancecited by 1
  13. 2025Auditing Sex/Gender Disparities in Emergency Triage with LLM-based Paired Comparisonscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026AgentFairBench: Do LLM Agents Discriminate When They Act?cited by 3
  2. 2026Proxy reliance in large language model decisions is uncalibrated to predictive evidencecited by 0
  3. 2026Instability Floors: Separating Bias from Noise in Fairness Audits of Clinical LLM Agents with FairMedAgentcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Bias patterns in the application of LLMs for clinical decision support: A comprehensive study
  2. 2026Evaluating the Presence of Sex Bias in Clinical Reasoning by Large Language Models
  3. 2025HALF: Harm-Aware LLM Fairness Evaluation Aligned with Deployment
  4. 2026AcuityBench: Evaluating Clinical Acuity Identification and Uncertainty Alignment
  5. 2025Small Language Models for Emergency Departments Decision Support: A Benchmark Study
  6. 2026Where Do Apparent LLM Clinical Triage Failures Arise? Localizing the Multiple-Choice Format Effect
  7. 2026First, Do No Harm (With LLMs): Mitigating Racial Bias via Agentic Workflows
  8. 2024FMBench: Benchmarking Fairness in Multimodal Large Language Models on Medical Tasks