Large Language Model Biases in Healthcare: A Scoping Review and Call for an Integrated Assessment Framework

Large Language Model Biases in Healthcare: A Scoping Review and Call for an Integrated Assessment Framework | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Systematic Review Large Language Model Biases in Healthcare: A Scoping Review and Call for an Integrated Assessment Framework Lu He, D. Phuong Do, Vishesh Girish Shet, Omar Farghaly, Priya Deshpande, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8604147/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Objective: To conduct a scoping review of bias assessment in studies applying Large Language Models (LLMs) to health data and to synthesize their prevailing conceptualization of bias. Material and methods: Following PRISMA guidelines, we queried PubMed and Scopus. Two annotators screened titles, abstracts, and full texts for eligibility, calibrating their assessments throughout the process. For included studies, we extracted and summarized data on LLMs (name and version, development domain, open- or closed- sourced status, and commercial or academic origin), NLP tasks (task formulation, gold standard dataset, evaluation metrics, prompting or fine-tuning strategies), and biases (type, assessment, and bias summary). Results: Of the 1,585 records retrieved, 76 papers met the eligibility criteria for full review. Among these, 59 reported identifying bias. Three major conceptualizations of bias emerged: behavioral output bias (non-stereotyping and stereotyping), predictive outcome bias, and representational bias. Studies generally adopted an observational approach (measuring bias using the existing dataset) or an experimental approach (altering prompts, e.g., with different demographic information, and comparing outputs). Discussion and Conclusion: Behavioral output bias and predictive outcome bias, both of which emphasize parity, dominate existing studies. Whether evaluated against external accuracy or internal equality benchmarks, these approaches often assume that equal performance across groups is inherently desirable. Treating all disparities as bias risks conflating poor model behavior with real-world disparities, and researchers should remain aware of potential trade-offs between parity and accuracy objectives. We introduce an integrated framework that combines parity and accuracy benchmarks and encourages transparent, context-aware interpretation of group differences. Medical Informatics Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8604147","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Systematic Review","associatedPublications":[],"authors":[{"id":574775050,"identity":"967a500b-93bb-4416-aeb0-53c1ee66998c","order_by":0,"name":"Lu He","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAnElEQVRIiWNgGAWjYBACAwb2gw8+wNhEauFJNpxBohYGM2kekrSYszckSNu2Hc5jYG/eJkGUFsuegweMc9sOFzPwHCsjTovBjYSE5Ny224kNEjlmRGq5/8DgsCVIi/wbYrXcYDBsZgTbwkOsljM5yYw95/4ntvGkFVsQp+X48eM/fpSlJfazH954gygtcMBGmvJRMApGwSgYBXgBAOzkL2ukiaswAAAAAElFTkSuQmCC","orcid":"","institution":"University of Wisconsin-Milwaukee","correspondingAuthor":true,"prefix":"","firstName":"Lu","middleName":"","lastName":"He","suffix":""},{"id":574775051,"identity":"dcc13925-cb2d-4f91-8cab-c59a3fb18c86","order_by":1,"name":"D. Phuong Do","email":"","orcid":"","institution":"University of Wisconsin-Milwaukee","correspondingAuthor":false,"prefix":"","firstName":"D.","middleName":"Phuong","lastName":"Do","suffix":""},{"id":574775052,"identity":"f5abe6b4-98fe-4f2e-8dd1-1f7c6c9ae2b8","order_by":2,"name":"Vishesh Girish Shet","email":"","orcid":"","institution":"University of Wisconsin-Milwaukee","correspondingAuthor":false,"prefix":"","firstName":"Vishesh","middleName":"Girish","lastName":"Shet","suffix":""},{"id":574775053,"identity":"47f0f021-1867-4514-a6c4-361d7b685913","order_by":3,"name":"Omar Farghaly","email":"","orcid":"","institution":"Marquette University","correspondingAuthor":false,"prefix":"","firstName":"Omar","middleName":"","lastName":"Farghaly","suffix":""},{"id":574775054,"identity":"70f9d016-10d4-4a1f-a1df-4f2b46635fec","order_by":4,"name":"Priya Deshpande","email":"","orcid":"","institution":"Marquette University","correspondingAuthor":false,"prefix":"","firstName":"Priya","middleName":"","lastName":"Deshpande","suffix":""},{"id":574775055,"identity":"e2c33000-0a8d-4b42-af8d-a9a9469b1e20","order_by":5,"name":"Praveen Madiraju","email":"","orcid":"","institution":"Marquette University","correspondingAuthor":false,"prefix":"","firstName":"Praveen","middleName":"","lastName":"Madiraju","suffix":""},{"id":574775056,"identity":"85361771-d90d-45b5-ae39-1281ed6b3f62","order_by":6,"name":"Jiancheng Ye","email":"","orcid":"","institution":"Weill Cornell","correspondingAuthor":false,"prefix":"","firstName":"Jiancheng","middleName":"","lastName":"Ye","suffix":""},{"id":574775057,"identity":"e60ca2ce-45ee-4cfe-bd0d-9209b4662933","order_by":7,"name":"Molly Beestrum","email":"","orcid":"","institution":"Northwestern University","correspondingAuthor":false,"prefix":"","firstName":"Molly","middleName":"","lastName":"Beestrum","suffix":""}],"badges":[],"createdAt":"2026-01-14 17:28:34","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-8604147/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8604147/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100547082,"identity":"1c9f21c6-7c5b-48c7-a4a9-12ba81e77a19","added_by":"auto","created_at":"2026-01-19 08:14:21","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":625800,"visible":true,"origin":"","legend":"","description":"","filename":"LLMBiasReviewPreprint.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8604147/v1_covered_20146e26-9f7e-4d65-99a0-cb4eefcb1162.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eLarge Language Model Biases in Healthcare: A Scoping Review and Call for an Integrated Assessment Framework\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University of Wisconsin-Milwaukee","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8604147/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8604147/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjective:\u003c/strong\u003e To conduct a scoping review of bias assessment in studies applying Large Language Models (LLMs) to health data and to synthesize their prevailing conceptualization of bias.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMaterial and methods:\u003c/strong\u003e Following PRISMA guidelines, we queried PubMed and Scopus. Two annotators screened titles, abstracts, and full texts for eligibility, calibrating their assessments throughout the process. For included studies, we extracted and summarized data on LLMs (name and version, development domain, open- or closed- sourced status, and commercial or academic origin), NLP tasks (task formulation, gold standard dataset, evaluation metrics, prompting or fine-tuning strategies), and biases (type, assessment, and bias summary).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eOf the 1,585 records retrieved, 76 papers met the eligibility criteria for full review. Among these, 59 reported identifying bias. Three major conceptualizations of bias emerged: behavioral output bias (non-stereotyping and stereotyping), predictive outcome bias, and representational bias. Studies generally adopted an observational approach (measuring bias using the existing dataset) or an experimental approach (altering prompts, e.g., with different demographic information, and comparing outputs).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDiscussion and Conclusion: \u003c/strong\u003eBehavioral output bias and predictive outcome bias, both of which emphasize parity, dominate existing studies. Whether evaluated against external accuracy or internal equality benchmarks, these approaches often assume that equal performance across groups is inherently desirable. Treating all disparities as bias risks conflating poor model behavior with real-world disparities, and researchers should remain aware of potential trade-offs between parity and accuracy objectives. We introduce an integrated framework that combines parity and accuracy benchmarks and encourages transparent, context-aware interpretation of group differences.\u003c/p\u003e","manuscriptTitle":"Large Language Model Biases in Healthcare: A Scoping Review and Call for an Integrated Assessment Framework","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-16 15:01:11","doi":"10.21203/rs.3.rs-8604147/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d440ac77-f59a-417d-94b9-abd24d11ef63","owner":[],"postedDate":"January 16th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":61163693,"name":"Medical Informatics"}],"tags":[],"updatedAt":"2026-01-16T15:01:11+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-16 15:01:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8604147","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8604147","identity":"rs-8604147","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}