{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T10:48:28Z","timestamp":1785754108210,"version":"3.56.0"},"reference-count":25,"publisher":"Oxford University Press (OUP)","license":[{"start":{"date-parts":[[2026,5,15]],"date-time":"2026-05-15T00:00:00Z","timestamp":1778803200000},"content-version":"am","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"name":"JSPS KAKENHI","award":["JP23K24465"],"award-info":[{"award-number":["JP23K24465"]}]},{"name":"Program for the Promotion of Precision Medicine in Rheumatoid Arthritis, Japan College of Rheumatology"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>To evaluate whether medical-domain pre-training and parameter-efficient fine-tuning improve the ability of locally deployable Large Language Models (LLMs) to predict long-term disease activity and disability in rheumatoid arthritis, and to benchmark their performance against established tabular machine-learning models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We trained on-premises Llama-2 (70B) models with and without medical pre-training (Meditron) and applied Quantized Low-Rank Adapter fine-tuning using structured data from 11\u2009865 patients in the Institute of Rheumatology, Rheumatoid Arthritis cohort. Models predicted eight binary outcomes of rheumatoid arthritis disease activity and disability at 2\u00a0years. Performance was compared with logistic regression, random forest, and XGBoost using receiver operating characteristic-area under the curve, Brier score, calibration plots, and decision-curve analysis.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Medical pre-training and Quantized Low-Rank Adapter fine-tuning both improved discrimination and calibration, with effects varying by endpoint. Fine-tuned LLMs achieved the highest or comparable receiver operating characteristic-area under the curves for most Disease Activity Score-based outcomes and matched conventional models for remission tasks. In low-prevalence outcomes, LLMs showed more stable calibration and small but consistent net-benefit advantages, while some tabular models exhibited clinically harmful decision curves.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>Endpoint-specific, privacy-preserving LLMs can complement or replace conventional tabular models for rheumatoid arthritis disease-activity prediction, providing reliable, locally deployable decision support without sacrificing data privacy.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/mr\/roag039","type":"journal-article","created":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T12:20:48Z","timestamp":1778674848000},"source":"Crossref","is-referenced-by-count":0,"title":["Medical pre-training and fine-tuning improve large-language-model prediction of rheumatoid-arthritis disease activity"],"prefix":"10.1093","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4846-3234","authenticated-orcid":false,"given":"Suguru","family":"Honda","sequence":"first","affiliation":[{"name":"Department of Rheumatology, Tokyo Women\u2019s Medical University School of Medicine , Tokyo ,","place":["Japan"]},{"name":"Institute of Rheumatology, Tokyo Women\u2019s Medical University Hospital , Tokyo ,","place":["Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katsunori","family":"Ikari","sequence":"additional","affiliation":[{"name":"Institute of Rheumatology, Tokyo Women\u2019s Medical University Hospital , Tokyo ,","place":["Japan"]},{"name":"Department of Orthopedic Surgery, Tokyo Women\u2019s Medical University School of Medicine , Tokyo ,","place":["Japan"]},{"name":"Section of Surgery for Inflammatory Arthritis, Department of Orthopedic Surgery, Tokyo Women's Medical University , Tokyo ,","place":["Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mayuko","family":"Fujisaki","sequence":"additional","affiliation":[{"name":"Department of Rheumatology, Tokyo Women\u2019s Medical University School of Medicine , Tokyo ,","place":["Japan"]},{"name":"Institute of Rheumatology, Tokyo Women\u2019s Medical University Hospital , Tokyo ,","place":["Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eiichi","family":"Tanaka","sequence":"additional","affiliation":[{"name":"Department of Rheumatology, Tokyo Women\u2019s Medical University School of Medicine , Tokyo ,","place":["Japan"]},{"name":"Institute of Rheumatology, Tokyo Women\u2019s Medical University Hospital , Tokyo ,","place":["Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6418-2603","authenticated-orcid":false,"given":"Masayoshi","family":"Harigai","sequence":"additional","affiliation":[{"name":"Department of Rheumatology, Tokyo Women\u2019s Medical University School of Medicine , Tokyo ,","place":["Japan"]},{"name":"Institute of Rheumatology, Tokyo Women\u2019s Medical University Hospital , Tokyo ,","place":["Japan"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2026,5,15]]},"reference":[{"key":"2026080306214024700_ref1","doi-asserted-by":"publisher","first-page":"1134","DOI":"10.1038\/s41591-024-02855-5","article-title":"Adapted large language models can outperform medical experts in clinical text summarization","volume":"30","author":"Van Veen","year":"2024","journal-title":"Nat Med"},{"key":"2026080306214024700_ref2","doi-asserted-by":"publisher","first-page":"357","DOI":"10.1038\/s41586-023-06160-y","article-title":"Health system-scale language models are all-purpose prediction engines","volume":"619","author":"Jiang","year":"2023","journal-title":"Nature."},{"key":"2026080306214024700_ref3","doi-asserted-by":"publisher","DOI":"10.1056\/AIoa2300138","article-title":"Towards generalist biomedical AI","volume":"1","author":"Tu","year":"2024","journal-title":"NEJM AI [Internet]"},{"key":"2026080306214024700_ref4","doi-asserted-by":"publisher","first-page":"685","DOI":"10.1136\/annrheumdis-2019-216655","article-title":"EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2019 update","volume":"79","author":"Smolen","year":"2020","journal-title":"Ann Rheum Dis"},{"key":"2026080306214024700_ref5","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1136\/annrheumdis-2015-207524","article-title":"Treating rheumatoid arthritis to target: 2014 update of the recommendations of an international task force","volume":"75","author":"Smolen","year":"2016","journal-title":"Ann Rheum Dis"},{"key":"2026080306214024700_ref6","doi-asserted-by":"publisher","author":"Nori","year":"2023","DOI":"10.48550\/ARXIV.2303.13375"},{"key":"2026080306214024700_ref7","doi-asserted-by":"publisher","first-page":"3256","DOI":"10.1093\/rheumatology\/kead291","article-title":"AI am a rheumatologist: a practical primer to large language models for rheumatologists","volume":"62","author":"Venerito","year":"2023","journal-title":"Rheumatology (Oxford)"},{"key":"2026080306214024700_ref8","doi-asserted-by":"publisher","DOI":"10.1093\/rap\/rkae119","article-title":"Large language models and rheumatology: are we there yet?","volume":"9","author":"Benavent","year":"2025","journal-title":"Rheumatol Adv Pract"},{"key":"2026080306214024700_ref9","doi-asserted-by":"publisher","author":"Touvron","year":"2023","DOI":"10.48550\/ARXIV.2307.09288"},{"key":"2026080306214024700_ref10","doi-asserted-by":"publisher","author":"Chen","year":"2023","DOI":"10.48550\/ARXIV.2311.16079"},{"key":"2026080306214024700_ref11","doi-asserted-by":"publisher","first-page":"479","DOI":"10.1002\/art.42737","article-title":"Doctor versus artificial intelligence: patient and physician evaluation of large language model responses to rheumatology patient questions in a cross-sectional study","volume":"76","author":"Ye","year":"2024","journal-title":"Arthritis Rheumatol"},{"key":"2026080306214024700_ref12","doi-asserted-by":"publisher","author":"Rusinovich Lovgach","year":"2024","DOI":"10.1136\/annrheumdis-2024-eular.1032"},{"key":"2026080306214024700_ref13","doi-asserted-by":"publisher","first-page":"509","DOI":"10.1007\/s00296-023-05473-5","article-title":"Assessing the accuracy and completeness of artificial intelligence language models in providing information on methotrexate use","volume":"44","author":"Coskun","year":"2024","journal-title":"Rheumatol Int"},{"key":"2026080306214024700_ref14","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1080\/14397595.2019.1660028","article-title":"A large observational cohort study of rheumatoid arthritis, IORRA: providing context for today\u2019s treatment options","volume":"30","author":"Yamanaka","year":"2020","journal-title":"Mod Rheumatol"},{"key":"2026080306214024700_ref15","doi-asserted-by":"publisher","author":"Fang","year":"2024","DOI":"10.48550\/arXiv.2402.17944"},{"key":"2026080306214024700_ref16","doi-asserted-by":"publisher","author":"Dettmers","year":"2023","DOI":"10.48550\/ARXIV.2305.14314"},{"key":"2026080306214024700_ref17","doi-asserted-by":"publisher","first-page":"565","DOI":"10.1177\/0272989X06295361","article-title":"Decision curve analysis: a novel method for evaluating prediction models","volume":"26","author":"Vickers","year":"2006","journal-title":"Med Decis Mak"},{"key":"2026080306214024700_ref18","doi-asserted-by":"publisher","author":"Lundberg","year":"2017","DOI":"10.48550\/arXiv.1705.07874"},{"key":"2026080306214024700_ref19","doi-asserted-by":"publisher","first-page":"112","DOI":"10.1093\/bioinformatics\/btr597","article-title":"MissForest\u2014non-parametric missing value imputation for mixed-type data","volume":"28","author":"Stekhoven","year":"2012","journal-title":"Bioinformatics."},{"key":"2026080306214024700_ref20","doi-asserted-by":"publisher","first-page":"938","DOI":"10.1016\/j.jclinepi.2009.11.009","article-title":"Use of Brier score to assess binary predictions","volume":"63","author":"Rufibach","year":"2010","journal-title":"J Clin Epidemiol"},{"key":"2026080306214024700_ref21","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1175\/1520-0493(1950)078&lt;0001:vofeit&gt;2.0.co;2","article-title":"Verification of forecasts expressed IN terms of probability","volume":"78","author":"Brier","year":"1950","journal-title":"Mon Wea Rev American Meteorological Society"},{"key":"2026080306214024700_ref22","author":"Efron","year":"1998"},{"key":"2026080306214024700_ref23","doi-asserted-by":"publisher","author":"Luo","year":"2023","DOI":"10.48550\/ARXIV.2308.08747"},{"key":"2026080306214024700_ref24","doi-asserted-by":"publisher","author":"Khattab","year":"2023","DOI":"10.48550\/arXiv.2310.03714"},{"key":"2026080306214024700_ref25","unstructured":"Colvin S, Victorien, Runkle S, Montague D, Badaracco AG, Hewitt D, et\u00a0al. pydantic\/pydantic: v1.10.26 2025-12-18 [Internet]. Zenodo; 2025 \u00a0[cited 2026 Feb 13]. \u00a010.5281\/ZENODO.17979236"}],"container-title":["Modern Rheumatology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/mr\/advance-article-pdf\/doi\/10.1093\/mr\/roag039\/68293456\/roag039.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/mr\/advance-article-pdf\/doi\/10.1093\/mr\/roag039\/68293456\/roag039.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T10:21:52Z","timestamp":1785752512000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/mr\/advance-article\/doi\/10.1093\/mr\/roag039\/8678969"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,15]]},"references-count":25,"URL":"https:\/\/doi.org\/10.1093\/mr\/roag039","relation":{},"ISSN":["1439-7595","1439-7609"],"issn-type":[{"value":"1439-7595","type":"print"},{"value":"1439-7609","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,15]]},"article-number":"roag039"}}