{"doi":"10.2196/57257","title":"Assessing Racial and Ethnic Bias in Text Generation by Large Language Models for Health Care–Related Tasks: Cross-Sectional Study","abstract":"BACKGROUND: Racial and ethnic bias in large language models (LLMs) used for health care tasks is a growing concern, as it may contribute to health disparities. In response, LLM operators implemented safeguards against prompts that are overtly seeking certain biases. OBJECTIVE: This study aims to investigate a potential racial and ethnic bias among 4 popular LLMs: GPT-3.5-turbo (OpenAI), GPT-4 (OpenAI), Gemini-1.0-pro (Google), and Llama3-70b (Meta) in generating health care consumer-directed text in the absence of overtly biased queries. METHODS: In this cross-sectional study, the 4 LLMs were prompted to generate discharge instructions for patients with HIV. Each patient's encounter deidentified metadata including race/ethnicity as a variable was passed over in a table format through a prompt 4 times, altering only the race/ethnicity information (African American, Asian, Hispanic White, and non-Hispanic White) each time, while keeping all other information constant. The prompt requested the model to write discharge instructions for each encounter without explicitly mentioning race or ethnicity. The LLM-generated instructions were analyzed for sentiment, subjectivity, reading ease, and word frequency by race/ethnicity. RESULTS: The only observed statistically significant difference between race/ethnicity groups was found in entity count (GPT-4, df=42, P=.047). However, post hoc chi-square analysis for GPT-4's entity counts showed no significant pairwise differences among race/ethnicity categories after Bonferroni correction. CONCLUSIONS: A total of 4 LLMs were relatively invariant to race/ethnicity in terms of linguistic and readability measures. While our study used proxy linguistic and readability measures to investigate racial and ethnic bias among 4 LLM responses in a health care-related task, there is an urgent need to establish universally accepted standards for measuring bias in LLM-generated responses. Further studies are needed to validate these results and assess their implications.","journal":"Journal of Medical Internet Research","year":2025,"id":510920,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":17,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9653,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2025-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1213566,"name":"Abdi D Wakene","orcid":"0009-0000-7270-0506","position":1,"is_corresponding":false},{"id":1368645,"name":"Andrew O. Johnson","orcid":null,"position":2,"is_corresponding":false},{"id":703181,"name":"Christoph U. Lehmann","orcid":"0000-0001-9559-4646","position":3,"is_corresponding":false},{"id":859006,"name":"Richard J Medford","orcid":"0000-0001-9814-8043","position":4,"is_corresponding":false},{"id":784968,"name":"John Hanna","orcid":"0000-0003-0909-9396","position":0,"is_corresponding":true}],"reference_count":10,"raw_metadata":null,"created_at":"2026-07-19T02:47:50.575618Z","pmid":"40080818","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}