{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,23]],"date-time":"2026-07-23T19:08:26Z","timestamp":1784833706690,"version":"3.55.0"},"reference-count":41,"publisher":"Public Library of Science (PLoS)","issue":"7","license":[{"start":{"date-parts":[[2026,7,23]],"date-time":"2026-07-23T00:00:00Z","timestamp":1784764800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.plosdigitalhealth.org"],"crossmark-restriction":false},"short-container-title":["PLOS Digit Health"],"abstract":"<jats:p>The heterogeneity of health data systems remains a major barrier to interoperability and data integration, and a challenge for clinical decision support and AI-driven research. The coexistence of multiple Common Data Models (CDMs) complicates integration, requiring complex manual or rule-based schema mapping, which is labor-intensive, error-prone, and difficult to scale. We evaluated three leading Large Language Models (LLMs): ChatGPT, Gemini, and DeepSeek, for their ability to match data elements from four widely used CDMs (PCORnet, OMOP, Sentinel, and i2b2) to FHIR R4 via web-based chat interfaces. Reference mappings from the HL7 Common Data Models Harmonization Implementation Guide were used as the benchmark. Standardized prompts were submitted across 36 total runs (3 LLMs\u2009\u00d7\u20094 schemas\u2009\u00d7\u20093 repetitions), and accuracy was measured using F1-scores with 95% confidence intervals, alongside strict (AND) and relaxed (OR) match criteria. DeepSeek obtained the highest numerical mean F1-scores (0.78-0.99) with strong consistency (AND Match\u2009=\u20090.87). Gemini showed good coverage (OR Match\u2009=\u20090.96) but greater variability (AND Match\u2009=\u20090.64), while ChatGPT achieved moderate accuracy with lower consistency (AND Match\u2009=\u20090.51). A two-way ANOVA indicated no statistically significant differences by LLM type (p\u2009=\u20090.138), CDM schema (p\u2009=\u20090.194), or their interaction (p\u2009=\u20090.451). Performance varied by domain, with Demographics and Condition mapped most accurately, while Observation showed the highest error rates (44% of total errors). Out of 2,583 total matchings, 314 errors were observed and categorized into four failure patterns: semantic confusion (42%), structural misalignment (28%), ambiguous terminology (18%), and terminology variance (12%). This study demonstrates the feasibility of using LLMs to automate health data integration as a complementary approach to traditional ontology-based methods, particularly for well-structured, context-rich CDMs. However, safe deployment requires human-in-the-loop validation, governance, prompt optimization, metadata integration, and validation to ensure clinical and research reliability.<\/jats:p>","DOI":"10.1371\/journal.pdig.0001511","type":"journal-article","created":{"date-parts":[[2026,7,23]],"date-time":"2026-07-23T18:53:28Z","timestamp":1784832808000},"page":"e0001511","update-policy":"https:\/\/doi.org\/10.1371\/journal.pdig.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Enabling interoperability across disparate health data sources using Large Language Models"],"prefix":"10.1371","volume":"5","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5385-7954","authenticated-orcid":true,"given":"Obinwa","family":"Ozonze","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-6163-9648","authenticated-orcid":true,"given":"Ogechukwu","family":"Okonor","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-4586-094X","authenticated-orcid":true,"given":"Uzondu","family":"Dike","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7201-7202","authenticated-orcid":true,"given":"Taiwo","family":"Adedeji","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,7,23]]},"reference":[{"issue":"6","key":"pdig.0001511.ref001","doi-asserted-by":"crossref","first-page":"395","DOI":"10.1038\/nrg3208","article-title":"Mining electronic health records: towards better research applications and clinical care","volume":"13","author":"PB Jensen","year":"2012","journal-title":"Nat Rev Genet"},{"issue":"1","key":"pdig.0001511.ref002","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1007\/s10840-016-0104-y","article-title":"The application of Big Data in medicine: current implications and future directions","volume":"47","author":"C Austin","year":"2016","journal-title":"J Interv Card Electrophysiol"},{"key":"pdig.0001511.ref003","unstructured":"Topol E. The Topol review. Preparing the healthcare workforce to deliver the digital future. 2019."},{"issue":"1","key":"pdig.0001511.ref004","article-title":"Use of EHRs data for clinical research: historical progress and current applications","volume":"3","author":"AH Nordo","year":"2019","journal-title":"Learn Health Syst"},{"issue":"1","key":"pdig.0001511.ref005","doi-asserted-by":"crossref","first-page":"6","DOI":"10.1080\/24709360.2019.1572344","article-title":"Clinical data quality: a data life cycle perspective","volume":"4","author":"C Weng","year":"2020","journal-title":"Biostat Epidemiol"},{"key":"pdig.0001511.ref006","first-page":"150","article-title":"Beyond volume: the impact of complex healthcare data on the machine learning pipeline.","author":"K Feldman","year":"2017","journal-title":"Lecture Notes in Computer Science. Springer International Publishing"},{"key":"pdig.0001511.ref007","doi-asserted-by":"crossref","first-page":"4","DOI":"10.1186\/2047-2501-2-4","article-title":"Traits and types of health data repositories","volume":"2","author":"TD Wade","year":"2014","journal-title":"Health Inf Sci Syst"},{"key":"pdig.0001511.ref008","doi-asserted-by":"crossref","first-page":"333","DOI":"10.1016\/j.jbi.2016.10.016","article-title":"Evaluating common data models for use with a longitudinal community registry","volume":"64","author":"M Garza","year":"2016","journal-title":"J Biomed Inform"},{"key":"pdig.0001511.ref009","doi-asserted-by":"crossref","DOI":"10.1093\/database\/bav121","article-title":"Portal of medical data models: information infrastructure for medical research and healthcare","volume":"2016","author":"M Dugas","year":"2016","journal-title":"Database (Oxford)"},{"key":"pdig.0001511.ref010","first-page":"247","article-title":"Moving towards an EHR data quality framework: the MIRACUM approach","volume":"267","author":"LA Kapsner","year":"2019","journal-title":"Stud Health Technol Inform"},{"issue":"3","key":"pdig.0001511.ref011","doi-asserted-by":"crossref","first-page":"553","DOI":"10.1093\/jamia\/ocu023","article-title":"Feasibility and utility of applications of the common data model to multiple, disparate observational health databases","volume":"22","author":"EA Voss","year":"2015","journal-title":"J Am Med Inform Assoc"},{"issue":"7","key":"pdig.0001511.ref012","doi-asserted-by":"crossref","DOI":"10.2196\/21929","article-title":"The Fast Health Interoperability Resources (FHIR) standard: systematic literature review of implementations, applications, challenges and opportunities","volume":"9","author":"M Ayaz","year":"2021","journal-title":"JMIR Med Inform"},{"issue":"2","key":"pdig.0001511.ref013","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0212463","article-title":"Data model harmonization for the all of us research program: transforming i2b2 data into the OMOP common data model","volume":"14","author":"JG Klann","year":"2019","journal-title":"PLoS One"},{"issue":"3","key":"pdig.0001511.ref014","doi-asserted-by":"crossref","first-page":"275","DOI":"10.1007\/s40273-020-00981-9","article-title":"Common problems, common data model solutions: evidence generation for health technology assessment","volume":"39","author":"S Kent","year":"2021","journal-title":"Pharmacoeconomics"},{"issue":"5","key":"pdig.0001511.ref015","doi-asserted-by":"crossref","first-page":"1634","DOI":"10.3390\/s24051634","article-title":"Building flexible, scalable, and machine learning-ready multimodal oncology datasets","volume":"24","author":"A Tripathi","year":"2024","journal-title":"Sensors"},{"key":"pdig.0001511.ref016","doi-asserted-by":"crossref","first-page":"104002","DOI":"10.1016\/j.ijmedinf.2019.104002","article-title":"Aggregating the syntactic and semantic similarity of healthcare data towards their transformation to HL7 FHIR through ontology matching","volume":"132","author":"A Kiourtis","year":"2019","journal-title":"Int J Med Inform"},{"key":"pdig.0001511.ref017","doi-asserted-by":"crossref","first-page":"106232","DOI":"10.1016\/j.cmpb.2021.106232","article-title":"Development of an application concerning fast healthcare interoperability resources based on standardized structured medical information exchange version 2 data","volume":"208","author":"D Xiao","year":"2021","journal-title":"Comput Methods Programs Biomed"},{"key":"pdig.0001511.ref018","article-title":"Managing data quality for a drug safety surveillance system","author":"AG Hartzema","year":"2013","journal-title":"Drug Saf"},{"issue":"4","key":"pdig.0001511.ref019","doi-asserted-by":"crossref","first-page":"334","DOI":"10.1007\/s007780100057","article-title":"A survey of approaches to automatic schema matching","volume":"10","author":"E Rahm","year":"2001","journal-title":"The VLDB Journal"},{"key":"pdig.0001511.ref020","doi-asserted-by":"crossref","unstructured":"Banek M, Vrdoljak B, Tjoa AM, Sko\u010dir Z. Automating the schema matching process for heterogeneous data warehouses. In: Data warehousing and knowledge discovery: 9th international conference, DaWaK 2007, Berlin, Heidelberg: Springer; 2007. pp. 45\u201354.","DOI":"10.1007\/978-3-540-74553-2_5"},{"issue":"11","key":"pdig.0001511.ref021","doi-asserted-by":"crossref","first-page":"695","DOI":"10.14778\/3402707.3402710","article-title":"Generic schema matching, ten years later","volume":"4","author":"PA Bernstein","year":"2011","journal-title":"Proc VLDB Endow"},{"issue":"2","key":"pdig.0001511.ref022","doi-asserted-by":"crossref","first-page":"192","DOI":"10.1016\/j.is.2010.09.004","article-title":"Automatic generation of probabilistic relationships for improving schema matching","volume":"36","author":"L Po","year":"2011","journal-title":"Inf Syst"},{"key":"pdig.0001511.ref023","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1007\/978-3-642-16518-4_1","article-title":"Towards large-scale schema and ontology matching.","volume-title":"Schema matching and mapping","author":"E Rahm","year":"2011"},{"issue":"5","key":"pdig.0001511.ref024","doi-asserted-by":"crossref","first-page":"1293","DOI":"10.1007\/s00778-007-0067-9","article-title":"Applications of corpus-based semantic similarity and word segmentation to database schema matching","volume":"17","author":"A Islam","year":"2007","journal-title":"The VLDB Journal"},{"issue":"3","key":"pdig.0001511.ref025","article-title":"Survey: models and prototypes of schema matching","volume":"6","author":"E Sutanta","year":"2016","journal-title":"Int J Electr Comput Eng"},{"key":"pdig.0001511.ref026","doi-asserted-by":"crossref","DOI":"10.6028\/NBS.SP.500-149","author":"JJ Newton","year":"1987","journal-title":"Guide on data entity (Naming Conventions)"},{"key":"pdig.0001511.ref027","unstructured":"ISO, Information technology \u2014 Metadata registries (MDR) \u2014 Part 5: Naming principles; 2015."},{"key":"pdig.0001511.ref028","doi-asserted-by":"crossref","unstructured":"Rachman MAF, Saptawati GAP. Database integration based on combination schema matching approach (case study: Multi-database of district health information system). In: 2017 2nd International conferences on Information Technology, Information Systems and Electrical Engineering (ICITISEE); 2017. pp. 430\u20135. https:\/\/doi.org\/10.1109\/icitisee.2017.8285544","DOI":"10.1109\/ICITISEE.2017.8285544"},{"issue":"10","key":"pdig.0001511.ref029","article-title":"A survey of schema matching research using database schemas and instances","volume":"8","author":"AA Alwan","year":"2017","journal-title":"Int J Adv Comput Sci Appl"},{"issue":"24","key":"pdig.0001511.ref030","doi-asserted-by":"crossref","first-page":"4975","DOI":"10.1016\/j.ins.2010.08.022","article-title":"Element similarity measures in XML schema matching","volume":"180","author":"A Algergawy","year":"2010","journal-title":"Inf Sci"},{"key":"pdig.0001511.ref031","unstructured":"OpenAi. OpenAi, ChatGPT; 2025."},{"key":"pdig.0001511.ref032","unstructured":"Google. An overview of the Gemini app. Gemini; 2025."},{"key":"pdig.0001511.ref033","unstructured":"DeepSeek. DeepSeek: into the unknown; 2025."},{"key":"pdig.0001511.ref034","article-title":"Gpt-4 technical report","author":"J Achiam","year":"2023","journal-title":"arXiv preprint"},{"key":"pdig.0001511.ref035","doi-asserted-by":"crossref","unstructured":"Rahman A, Mahir SH, Tashrif MT, Karim MA, Aishi AA, Kundu D, et al. Comparative analysis based on DeepSeek, ChatGPT, and Google Gemini: features, techniques, performance, future prospects; 2025.","DOI":"10.1016\/j.sasc.2025.200396"},{"key":"pdig.0001511.ref036","unstructured":"DeepSeek AI. DeepSeek-V3 Technical Report; 2025."},{"key":"pdig.0001511.ref037","unstructured":"International HL. Common data models harmonization; 2021."},{"key":"pdig.0001511.ref038","article-title":"Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation","author":"DM Powers","year":"2020","journal-title":"arXiv preprint"},{"issue":"2","key":"pdig.0001511.ref039","article-title":"Agreementmakerlight","volume":"16","author":"D Faria","year":"2025","journal-title":"Semant Web"},{"key":"pdig.0001511.ref040","volume-title":"SMAT: an attention-based deep learning solution to the automation of schema matching","author":"J Zhang","year":"2021"},{"key":"pdig.0001511.ref041","article-title":"LLMs4SchemaDiscovery: a Human-in-the-Loop workflow for scientific schema mining with large language models","author":"S Sadruddin","year":"2025","journal-title":"arXiv preprint"}],"container-title":["PLOS Digital Health"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pdig.0001511","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,23]],"date-time":"2026-07-23T18:53:40Z","timestamp":1784832820000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pdig.0001511"}},"subtitle":[],"editor":[{"given":"Hadi","family":"Amiri","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,7,23]]},"references-count":41,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2026,7,23]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pdig.0001511","relation":{},"ISSN":["2767-3170"],"issn-type":[{"value":"2767-3170","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,23]]}}}