FairMedQA: Benchmarking Bias in Large Language Models for Medical Question Answering

TL;DR

This paper introduces FairMedQA, a new benchmark for measuring biases in large language models used for medical question answering, revealing significant disparities and exposing limitations of existing benchmarks.

Contribution

The paper presents FairMedQA, a comprehensive benchmark with 4,806 question pairs, to evaluate bias in 12 LLMs, and demonstrates its superior sensitivity over previous benchmarks.

Findings

Substantial accuracy disparities across demographic groups (3-19%)

FairMedQA detects biases at least 12% larger than previous benchmarks

Highlights urgent need for debiasing and validation in clinical LLM applications

Abstract

Large language models (LLMs) are approaching expert-level performance in medical question answering (QA), demonstrating strong potential to improve public healthcare. However, underlying biases related to sensitive attributes such as sex and race pose life-critical risks. The extent to which such sensitive attributes affect diagnosis remains an open question and requires comprehensive empirical investigation. Additionally, even the latest Counterfactual Patient Variations (CPV) benchmark can hardly distinguish the bias levels of different LLMs. To further explore these dynamics, we propose a new benchmark, FairMedQA, and benchmark 12 representative LLMs. FairMedQA contains 4,806 counterfactual question pairs constructed from 801 clinical vignettes. Our results reveal substantial accuracy disparity ranging from 3 to 19 percentage points across sensitive demographic groups. Notably, FairMedQA exposes biases that are at least 12 percentage points larger than those identified by the latest CPV benchmark, presenting superior benchmarking sensitivity. Our results underscore an urgent need for targeted debiasing techniques and more rigorous, identity-aware validation protocols before LLMs can be safely integrated into practical clinical decision-support systems.