{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,30]],"date-time":"2025-12-30T13:14:57Z","timestamp":1767100497074,"version":"3.48.0"},"reference-count":54,"publisher":"BMJ","issue":"11","license":[{"start":{"date-parts":[[2025,11,29]],"date-time":"2025-11-29T00:00:00Z","timestamp":1764374400000},"content-version":"unspecified","delay-in-days":28,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003710","name":"Korea Health Industry Development Institute","doi-asserted-by":"publisher","award":["RS-2025-24535069 and HI23C0733"],"award-info":[{"award-number":["RS-2025-24535069 and HI23C0733"]}],"id":[{"id":"10.13039\/501100003710","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100007033","name":"Eisai Korea Inc","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100007033","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["BMJ Open"],"accepted":{"date-parts":[[2025,11,7]]},"published-print":{"date-parts":[[2025,11]]},"abstract":"<jats:sec>\n                    <jats:title>Objectives<\/jats:title>\n                    <jats:p>Methotrexate (MTX) effectively controls rheumatoid arthritis (RA) but often leads to side effects (SE) such as gastrointestinal (GI) issues, liver toxicity and bone marrow suppression. To develop clinically interpretable machine learning (ML) models that accurately predict MTX-related SE in patients with RA taking MTX. The aim was to enhance predictive accuracy and to identify patient-specific risk factors using explainable artificial intelligence (XAI), thereby enabling transparent clinical interpretation. We specifically sought to address the unmet need for individualised risk stratification using real-world, multicentre observational data.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Design<\/jats:title>\n                    <jats:p>Retrospective case-control study.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Setting<\/jats:title>\n                    <jats:p>Across 23 rheumatology clinics in South Korea, based on data from a nationwide multicentre cohort.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Participants<\/jats:title>\n                    <jats:p>A total of 5077 patients with RA were initially enrolled from the Korean Observational Study Network for Arthritis. After excluding those with missing clinical, demographic or prescription data and those not receiving MTX, 2375 patients remained eligible. Among these, 1654 and 1218 patients were included in the overall SE and GI SE analysis groups, respectively, after 1:1 propensity score matching. All patients were aged \u226518 years and met the 1987 American College of Rheumatology classification criteria.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Primary and secondary outcome measures<\/jats:title>\n                    <jats:p>The primary outcome was the presence of SE in patients with RA taking MTX, categorised into overall SE and GI SE, based on standardised patient questionnaires and clinical assessments. The secondary outcome was the identification of key predictors using SHapley Additive exPlanations (SHAP) to enhance the interpretability of ML predictions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Among six ML classifiers, extreme gradient boosting demonstrated the highest performance in predicting overall SE (area under the curve (AUC) 0.781, F1 score 0.672, area under the precision-recall curve (AUPRC) 0.757) and GI SE (AUC 0.701, F1 score 0.690, AUPRC 0.670). SHAP analysis identified key predictive features including age, physician visual analogue scale score, alanine aminotransferase, Health Assessment Questionnaire score, celecoxib use and drug adherence. Logistic regression confirmed statistical significance for multiple variables (eg, OR 4.63; 95% CI 1.41 to 20.90 for non-adherence &gt;30\u2009days; OR 1.45; 95% CI 1.14 to 1.85 for celecoxib use). DeLong\u2019s test indicated that boosting models significantly outperformed support vector machine (p&lt;0.001).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>Interpretable ML models using real-world clinical data can accurately predict SE in patients with RA taking MTX. These models may facilitate early identification of high-risk individuals and inform personalised treatment strategies. Integration into clinical decision support systems could improve MTX safety monitoring. Further prospective validation in external cohorts is warranted.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1136\/bmjopen-2025-108527","type":"journal-article","created":{"date-parts":[[2025,11,30]],"date-time":"2025-11-30T05:00:28Z","timestamp":1764478828000},"page":"e108527","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":0,"title":["Development of explainable machine learning models to predict side effects in patients with rheumatoid arthritis taking methotrexate treatment: a nationwide multicentre cohort study"],"prefix":"10.1136","volume":"15","author":[{"given":"Junbeom","family":"Jang","sequence":"first","affiliation":[{"name":"Department of Medical Sciences, Kangwon National University, Chuncheon-si, The Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2927-370X","authenticated-orcid":false,"given":"Woo Jin","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Internal Medicine, Kangwon National University, Chuncheon-si, Gangwon-do, The Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9796-2799","authenticated-orcid":false,"given":"Sang Won","family":"Park","sequence":"additional","affiliation":[{"name":"Next-Generation Information Center, Kangwon National University Hospital, Chuncheon-si, Gangwon, The Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ki Won","family":"Moon","sequence":"additional","affiliation":[{"name":"Division of Rheumatology, Department of Internal Medicine, Kangwon National University Hospital, Chuncheon-si, Kangwon-do, The Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"239","published-online":{"date-parts":[[2025,11,29]]},"reference":[{"key":"2025123005100394000_15.11.e108527.1","doi-asserted-by":"publisher","DOI":"10.1007\/s10067-023-06628-2"},{"key":"2025123005100394000_15.11.e108527.2","doi-asserted-by":"crossref","first-page":"526","DOI":"10.1016\/j.semarthrit.2019.11.007","article-title":"An increased disease burden of autoimmune inflammatory rheumatic diseases in Korea","volume":"50","author":"Kim","year":"2020","journal-title":"Semin Arthritis Rheum"},{"key":"2025123005100394000_15.11.e108527.3","doi-asserted-by":"crossref","first-page":"11","DOI":"10.4078\/jkra.2008.15.1.11","article-title":"The Prevalence and Trend of Arthritis in Korea: Results from Korea National Health and Nutrition Examination Surveys","volume":"15","author":"Hur","year":"2008","journal-title":"J Korean Rheum Assoc"},{"key":"2025123005100394000_15.11.e108527.4","doi-asserted-by":"publisher","DOI":"10.1136\/annrheumdis-2019-216655"},{"key":"2025123005100394000_15.11.e108527.5","doi-asserted-by":"crossref","first-page":"162","DOI":"10.1186\/s13075-022-02851-5","article-title":"Clinical predictors of response to methotrexate in patients with rheumatoid arthritis: a machine learning approach using clinical trial data","volume":"24","author":"Duong","year":"2022","journal-title":"Arthritis Res Ther"},{"key":"2025123005100394000_15.11.e108527.6","article-title":"Folic acid and folinic acid for reducing side effects in patients receiving methotrexate for rheumatoid arthritis","volume":"2014","author":"Shea","year":"2013","journal-title":"Cochrane Database Syst Rev"},{"key":"2025123005100394000_15.11.e108527.7","doi-asserted-by":"publisher","DOI":"10.1001\/archinte.1996.00440130066007"},{"key":"2025123005100394000_15.11.e108527.8","first-page":"8","article-title":"NSAID induced gastrointestinal complications: the ARAMIS perspective--1997. Arthritis, Rheumatism, and Aging Medical Information System","volume":"51","author":"Singh","year":"1998","journal-title":"J Rheumatol Suppl"},{"key":"2025123005100394000_15.11.e108527.9","doi-asserted-by":"publisher","DOI":"10.1001\/archinte.1989.00390030137027"},{"key":"2025123005100394000_15.11.e108527.10","doi-asserted-by":"crossref","first-page":"549","DOI":"10.1002\/mrc.3985","article-title":"NMR\u2010based metabolomics coupled with pattern recognition methods in biomarker discovery and disease diagnosis","volume":"51","author":"Zhang","year":"2013","journal-title":"Magnetic Reson in Chemistry"},{"key":"2025123005100394000_15.11.e108527.11","doi-asserted-by":"crossref","first-page":"526","DOI":"10.1002\/ptr.5032","article-title":"Potentiating therapeutic effects by enhancing synergism based on active constituents from traditional medicine","volume":"28","author":"Zhang","year":"2014","journal-title":"Phytother Res"},{"key":"2025123005100394000_15.11.e108527.12","doi-asserted-by":"crossref","DOI":"10.1186\/ar4413","article-title":"Prevalence of methotrexate intolerance in rheumatoid arthritis and psoriatic arthritis","volume":"15","author":"\u0106alasan","year":"2013","journal-title":"Arthritis Res Ther"},{"key":"2025123005100394000_15.11.e108527.13","doi-asserted-by":"crossref","first-page":"65022","DOI":"10.18632\/oncotarget.17531","article-title":"Metabolic characterization and pathway analysis of berberine protects against prostate cancer","volume":"8","author":"Li","year":"2017","journal-title":"Oncotarget"},{"key":"2025123005100394000_15.11.e108527.14","doi-asserted-by":"publisher","DOI":"10.1136\/ard.2008.093690"},{"key":"2025123005100394000_15.11.e108527.15","first-page":"S95","article-title":"Side effects and management of side effects of methotrexate in rheumatoid arthritis","volume":"28","author":"Albrecht","year":"2010","journal-title":"Clin Exp Rheumatol"},{"key":"2025123005100394000_15.11.e108527.16","doi-asserted-by":"publisher","DOI":"10.1002\/art.10135"},{"key":"2025123005100394000_15.11.e108527.17","doi-asserted-by":"crossref","DOI":"10.1186\/s12911-019-0801-4","article-title":"Predicting factors for survival of breast cancer patients using machine learning techniques","volume":"19","author":"Ganggayah","year":"2019","journal-title":"BMC Med Inform Decis Mak"},{"key":"2025123005100394000_15.11.e108527.18","doi-asserted-by":"crossref","first-page":"1453","DOI":"10.1093\/rheumatology\/kez048","article-title":"Methotrexate persistence and adverse drug reactions in patients with juvenile idiopathic arthritis","volume":"58","author":"Kearsley-Fleet","year":"2019","journal-title":"Rheumatology (Sunnyvale)"},{"key":"2025123005100394000_15.11.e108527.19","doi-asserted-by":"publisher","DOI":"10.1002\/art.27758"},{"key":"2025123005100394000_15.11.e108527.20","doi-asserted-by":"publisher","DOI":"10.3899\/jrheum.090838"},{"key":"2025123005100394000_15.11.e108527.21","doi-asserted-by":"publisher","DOI":"10.1136\/annrheumdis-2018-213035"},{"key":"2025123005100394000_15.11.e108527.22","doi-asserted-by":"crossref","first-page":"2402","DOI":"10.1093\/rheumatology\/keac645","article-title":"Machine learning identifies a profile of inadequate responder to methotrexate in rheumatoid arthritis","volume":"62","author":"Duquesne","year":"2023","journal-title":"Rheumatology (Oxford)"},{"key":"2025123005100394000_15.11.e108527.23","doi-asserted-by":"crossref","DOI":"10.1186\/s13075-021-02635-3","article-title":"Machine learning-based prediction model for responses of bDMARDs in patients with rheumatoid arthritis and ankylosing spondylitis","volume":"23","author":"Lee","year":"2021","journal-title":"Arthritis Res Ther"},{"key":"2025123005100394000_15.11.e108527.24","doi-asserted-by":"crossref","unstructured":"Ahmad MA , Eckert C , Teredesai A . Interpretable machine learning in healthcare. BCB \u201918; Washington DC USA, August 15, 2018:559\u201360. Available: https:\/\/dl.acm.org\/doi\/proceedings\/10.1145\/3233547","DOI":"10.1145\/3233547.3233667"},{"key":"2025123005100394000_15.11.e108527.25","doi-asserted-by":"publisher","DOI":"10.1016\/j.cgh.2022.07.006"},{"key":"2025123005100394000_15.11.e108527.26","doi-asserted-by":"publisher","DOI":"10.1016\/j.semarthrit.2011.09.007"},{"key":"2025123005100394000_15.11.e108527.27","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.39335.541782.AD"},{"key":"2025123005100394000_15.11.e108527.28","doi-asserted-by":"publisher","DOI":"10.1002\/art.1780380107"},{"key":"2025123005100394000_15.11.e108527.29","first-page":"1975","article-title":"Psychometric evaluation of a Korean Health Assessment Questionnaire for clinical research","volume":"25","author":"Bae","year":"1998","journal-title":"J Rheumatol"},{"key":"2025123005100394000_15.11.e108527.30","doi-asserted-by":"crossref","unstructured":"Boser BE , Guyon IM , Vapnik VN . A training algorithm for optimal margin classifiers. COLT92; Pittsburgh Pennsylvania USA, July 1992:144\u201352. doi:10.1145\/130385.130401","DOI":"10.1145\/130385.130401"},{"key":"2025123005100394000_15.11.e108527.31","doi-asserted-by":"publisher","DOI":"10.1023\/A:1010933404324"},{"key":"2025123005100394000_15.11.e108527.32","doi-asserted-by":"publisher","DOI":"10.1214\/aos\/1013203450"},{"key":"2025123005100394000_15.11.e108527.33","doi-asserted-by":"crossref","unstructured":"Chen T , Guestrin C . XGBoost: a scalable tree boosting system. 2016:785\u201394. doi:10.1145\/2939672.2939785","DOI":"10.1145\/2939672.2939785"},{"key":"2025123005100394000_15.11.e108527.34","unstructured":"Ke G , Meng Q , Finley T , et al . LightGBM: a highly efficient gradient boosting decision tree. 2017.3149\u201357."},{"key":"2025123005100394000_15.11.e108527.35","unstructured":"Prokhorenkova L , Gusev G , Vorobev A , et al . CatBoost: unbiased boosting with categorical features. 2019. Available: http:\/\/arxiv.org\/abs\/1706.09516"},{"key":"2025123005100394000_15.11.e108527.36","doi-asserted-by":"crossref","unstructured":"Patro SGK , sahu KK . Normalization: A Preprocessing Stage. International Advanced Research Journal in Science, Engineering and Technology 2015;20\u20132. doi:10.17148\/IARJSET.2015.2305","DOI":"10.17148\/IARJSET.2015.2305"},{"key":"2025123005100394000_15.11.e108527.37","doi-asserted-by":"crossref","unstructured":"Akiba T , Sano S , Yanase T , et al . Optuna: a next-generation hyperparameter optimization framework. October 28, 2019:2623\u201331. doi:10.1145\/3292500.3330701","DOI":"10.1145\/3292500.3330701"},{"key":"2025123005100394000_15.11.e108527.38","unstructured":"Lundberg SM , Lee SI . A unified approach to interpreting model predictions."},{"key":"2025123005100394000_15.11.e108527.39","doi-asserted-by":"crossref","first-page":"4499","DOI":"10.1038\/s41467-024-48825-w","article-title":"Acute and post-acute respiratory complications of SARS-CoV-2 infection: population-based cohort study in South Korea and Japan","volume":"15","author":"Choi","year":"2024","journal-title":"Nat Commun"},{"key":"2025123005100394000_15.11.e108527.40","doi-asserted-by":"crossref","first-page":"668","DOI":"10.3904\/kjim.2023.276","article-title":"Impact of socioeconomic status on biologics utilization in rheumatoid arthritis: revealing inequalities and healthcare efficiency","volume":"39","author":"Kim","year":"2024","journal-title":"Korean J Intern Med"},{"key":"2025123005100394000_15.11.e108527.41","doi-asserted-by":"crossref","DOI":"10.1136\/rmdopen-2022-002722","article-title":"Impact of early age at menopause on disease outcomes in postmenopausal women with rheumatoid arthritis: a large observational cohort study of Korean patients with rheumatoid arthritis","volume":"9","author":"Park","year":"2023","journal-title":"RMD Open"},{"key":"2025123005100394000_15.11.e108527.42","doi-asserted-by":"crossref","first-page":"21800","DOI":"10.1038\/s41598-023-48842-7","article-title":"Prediction of osteoporosis in patients with rheumatoid arthritis using machine learning","volume":"13","author":"Lee","year":"2023","journal-title":"Sci Rep"},{"key":"2025123005100394000_15.11.e108527.43","doi-asserted-by":"crossref","first-page":"1599","DOI":"10.3390\/jcm12041599","article-title":"Predicting Hepatotoxicity Associated with Low-Dose Methotrexate Using Machine Learning","volume":"12","author":"Hu","year":"2023","journal-title":"JCM"},{"key":"2025123005100394000_15.11.e108527.44","doi-asserted-by":"crossref","first-page":"4050","DOI":"10.3390\/jcm13144050","article-title":"Development of a Risk Prediction Model for Adverse Skin Events Associated with TNF-\u03b1 Inhibitors in Rheumatoid Arthritis Patients","volume":"13","author":"Kim","year":"2024","journal-title":"JCM"},{"key":"2025123005100394000_15.11.e108527.45","doi-asserted-by":"crossref","first-page":"67","DOI":"10.1186\/s13075-021-02439-5","article-title":"Advanced machine learning for predicting individual risk of flares in rheumatoid arthritis patients tapering biologic drugs","volume":"23","author":"Vodencarevic","year":"2021","journal-title":"Arthritis Res Ther"},{"key":"2025123005100394000_15.11.e108527.46","doi-asserted-by":"publisher","DOI":"10.1136\/ard.2008.101378"},{"key":"2025123005100394000_15.11.e108527.47","doi-asserted-by":"publisher","DOI":"10.3899\/jrheum.131170"},{"key":"2025123005100394000_15.11.e108527.48","doi-asserted-by":"publisher","DOI":"10.1016\/j.semarthrit.2008.08.001"},{"key":"2025123005100394000_15.11.e108527.49","doi-asserted-by":"publisher","DOI":"10.1007\/BF02230334"},{"key":"2025123005100394000_15.11.e108527.50","doi-asserted-by":"crossref","first-page":"4001","DOI":"10.1093\/rheumatology\/keab304","article-title":"Prevalence and predictors of adverse events with methotrexate mono- and combination-therapy for rheumatoid arthritis: a systematic review","volume":"60","author":"Sherbini","year":"2021","journal-title":"Rheumatology (Oxford)"},{"key":"2025123005100394000_15.11.e108527.51","doi-asserted-by":"crossref","DOI":"10.3346\/jkms.2024.39.e53","article-title":"Early Prediction of Mortality for Septic Patients Visiting Emergency Room Based on Explainable Machine Learning: A Real-World Multicenter Study","volume":"39","author":"Park","year":"2024","journal-title":"J Korean Med Sci"},{"key":"2025123005100394000_15.11.e108527.52","doi-asserted-by":"publisher","DOI":"10.1093\/aje\/153.11.1089"},{"key":"2025123005100394000_15.11.e108527.53","doi-asserted-by":"publisher","DOI":"10.2165\/00002018-200730100-00005"},{"key":"2025123005100394000_15.11.e108527.54","doi-asserted-by":"publisher","DOI":"10.7326\/0003-4819-114-9-735"}],"container-title":["BMJ Open"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bmjopen-2025-108527","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,30]],"date-time":"2025-12-30T13:10:18Z","timestamp":1767100218000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmjopen.bmj.com\/lookup\/doi\/10.1136\/bmjopen-2025-108527"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11]]},"references-count":54,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2025,11,29]]},"published-print":{"date-parts":[[2025,11]]}},"alternative-id":["10.1136\/bmjopen-2025-108527"],"URL":"https:\/\/doi.org\/10.1136\/bmjopen-2025-108527","relation":{},"ISSN":["2044-6055","2044-6055"],"issn-type":[{"type":"print","value":"2044-6055"},{"type":"electronic","value":"2044-6055"}],"subject":[],"published":{"date-parts":[[2025,11]]}}}