{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,29]],"date-time":"2026-06-29T23:46:59Z","timestamp":1782776819064,"version":"3.54.5"},"reference-count":26,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T00:00:00Z","timestamp":1776988800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T00:00:00Z","timestamp":1782777600000},"content-version":"vor","delay-in-days":67,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/100007270","name":"University of Michigan","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100007270","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["npj Digit. Med."],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Challenges with model generalizability and data privacy have led to a shift in health artificial intelligence (AI) models being trained locally within individual health systems rather than relying on multicenter data. Localization carries the promise of capturing local practice patterns and patient demographics, presumably resulting in better models. Our study empirically tests this hypothesis in a national research network by comparing locally trained models predicting acute kidney injury (AKI) after cardiac surgery with two multicenter modeling approaches, pooling and a novel federated model stacking method. Trained on 43,926 cases across 23 hospitals, the study finds that multicenter models outperform single-center approaches, with higher area under the receiver operating characteristic curves (AUCs) for all AKI severity levels in both temporal and external validation sets. Hospitals with smaller case volumes benefit the most from multicenter approaches, showing the greatest AUC increase over locally trained models.<\/jats:p>","DOI":"10.1038\/s41746-026-02634-1","type":"journal-article","created":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T08:46:34Z","timestamp":1777020394000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Exploring the limits of localization: federated model stacking improves hospital-level prediction in a national research network"],"prefix":"10.1038","volume":"9","author":[{"given":"Jie","family":"Cao","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Emily","family":"Balczewski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rahul","family":"Ladhania","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Paramveer","family":"Dhillon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Donald S.","family":"Likosky","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Allison M.","family":"Janda","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"name":"MPOG Collaborators","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nathan L.","family":"Pace","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robert E.","family":"Freundlich","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Douglas A.","family":"Colquhoun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bradley A.","family":"Fritz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robert B.","family":"Schonberger","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sachin","family":"Kheterpal","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael R.","family":"Mathis","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Karandeep","family":"Singh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,4,24]]},"reference":[{"key":"2634_CR1","doi-asserted-by":"publisher","first-page":"283","DOI":"10.1056\/NEJMc2104626","volume":"385","author":"SG Finlayson","year":"2021","unstructured":"Finlayson, S. G. et al. The clinician and dataset shift in artificial intelligence. N. Engl. J. Med. 385, 283\u2013286 (2021).","journal-title":"N. Engl. J. Med."},{"key":"2634_CR2","doi-asserted-by":"publisher","first-page":"515","DOI":"10.7326\/0003-4819-130-6-199903160-00016","volume":"130","author":"AC Justice","year":"1999","unstructured":"Justice, A. C., Covinsky, K. E. & Berlin, J. A. Assessing the generalizability of prognostic information. Ann. Intern. Med. 130, 515\u2013524 (1999).","journal-title":"Ann. Intern. Med."},{"key":"2634_CR3","unstructured":"Ross, C. Epic overhauls popular sepsis algorithm criticized for faulty alarms. STAT https:\/\/www.statnews.com\/2022\/10\/03\/epic-sepsis-algorithm-revamp-training\/ (2022)."},{"key":"2634_CR4","doi-asserted-by":"crossref","unstructured":"Dameff, C. et al. Ransomware attack associated with disruptions at adjacent emergency departments in the US. JAMA Network Open 6, e2312270 (2023).","DOI":"10.1001\/jamanetworkopen.2023.12270"},{"key":"2634_CR5","doi-asserted-by":"publisher","first-page":"1283","DOI":"10.1001\/jamainternmed.2024.3162","volume":"184","author":"HT Neprash","year":"2024","unstructured":"Neprash, H. T., Dameff, C. & Tully, J. Cybersecurity lessons from the change healthcare attack. JAMA Intern Med 184, 1283\u20131284 (2024).","journal-title":"JAMA Intern Med"},{"key":"2634_CR6","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/1471-2288-14-137","volume":"14","author":"T van der Ploeg","year":"2014","unstructured":"van der Ploeg, T., Austin, P. C. & Steyerberg, E. W. Modern modelling techniques are data hungry: a simulation study for predicting dichotomous endpoints. BMC Med. Res. Methodol. 14, 1\u201313 (2014).","journal-title":"BMC Med. Res. Methodol."},{"key":"2634_CR7","doi-asserted-by":"publisher","first-page":"1121","DOI":"10.1038\/s42256-022-00563-8","volume":"4","author":"J Cao","year":"2022","unstructured":"Cao, J. et al. Generalizability of an acute kidney injury prediction model across health systems. Nat. Mach. Intell. 4, 1121\u20131129 (2022).","journal-title":"Nat. Mach. Intell."},{"key":"2634_CR8","doi-asserted-by":"publisher","first-page":"119","DOI":"10.1038\/s41746-020-00323-1","volume":"3","author":"N Rieke","year":"2020","unstructured":"Rieke, N. et al. The future of digital health with federated learning. npj Digit. Med. 3, 119 (2020).","journal-title":"npj Digit. Med."},{"key":"2634_CR9","doi-asserted-by":"publisher","first-page":"59","DOI":"10.1016\/j.ijmedinf.2018.01.007","volume":"112","author":"TS Brisimi","year":"2018","unstructured":"Brisimi, T. S. et al. Federated learning of predictive models from federated Electronic Health Records. Int. J. Med. Inform. 112, 59\u201367 (2018).","journal-title":"Int. J. Med. Inform."},{"key":"2634_CR10","doi-asserted-by":"publisher","first-page":"103291","DOI":"10.1016\/j.jbi.2019.103291","volume":"99","author":"L Huang","year":"2019","unstructured":"Huang, L. et al. Patient clustering improves efficiency of federated machine learning to predict mortality and hospital stay time using distributed electronic medical records. J. Biomed. Inform. 99, 103291 (2019).","journal-title":"J. Biomed. Inform."},{"key":"2634_CR11","doi-asserted-by":"crossref","first-page":"205520762211344","DOI":"10.1177\/20552076221134455","volume":"8","author":"TJ Loftus","year":"2022","unstructured":"Loftus, T. J. et al. Federated learning for preserving data privacy in collaborative healthcare research. Digit. Health 8, 20552076221134456 (2022).","journal-title":"Digit. Health"},{"key":"2634_CR12","doi-asserted-by":"publisher","first-page":"1081","DOI":"10.1038\/s42256-021-00421-z","volume":"3","author":"X Bai","year":"2021","unstructured":"Bai, X. et al. Advancing COVID-19 diagnosis with privacy-preserving collaboration in artificial intelligence. Nat. Mach. Intell. 3, 1081\u20131089 (2021).","journal-title":"Nat. Mach. Intell."},{"key":"2634_CR13","doi-asserted-by":"publisher","first-page":"1735","DOI":"10.1038\/s41591-021-01506-3","volume":"27","author":"I Dayan","year":"2021","unstructured":"Dayan, I. et al. Federated learning for predicting clinical outcomes in patients with COVID-19. Nat. Med. 27, 1735\u20131743 (2021).","journal-title":"Nat. Med."},{"key":"2634_CR14","unstructured":"Roth, H. R. et al. Federated learning for breast density classification: a real-world implementation. in Domain Adaptation and Representation Transfer, and Distributed and Collaborative Learning 181\u2013191 (Springer International Publishing, 2020)."},{"key":"2634_CR15","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-022-33407-5","volume":"13","author":"S Pati","year":"2022","unstructured":"Pati, S. et al. Federated learning enables big data for rare cancer boundary detection. Nat. Commun. 13, 7346 (2022).","journal-title":"Nat. Commun."},{"key":"2634_CR16","unstructured":"Epic Cosmos. https:\/\/cosmos.epic.com\/."},{"key":"2634_CR17","unstructured":"Real-world data for the life sciences and healthcare. TriNetX. https:\/\/trinetx.com\/ (2021)."},{"key":"2634_CR18","doi-asserted-by":"crossref","unstructured":"Steyerberg, E. W. Updating for a new setting. Clin. Predict. Models 399\u2013429 https:\/\/link.springer.com\/book\/10.1007\/978-3-030-16399-0 (2019).","DOI":"10.1007\/978-3-030-16399-0_20"},{"key":"2634_CR19","doi-asserted-by":"publisher","unstructured":"Cantu-Cervini, E. Personalized federated learning via stacking. arXiv https:\/\/doi.org\/10.48550\/arXiv.2404.10957 (2024).","DOI":"10.48550\/arXiv.2404.10957"},{"key":"2634_CR20","doi-asserted-by":"publisher","first-page":"109929","DOI":"10.1016\/j.knosys.2022.109929","volume":"257","author":"T Shaik","year":"2022","unstructured":"Shaik, T. et al. FedStack: personalized activity monitoring using stacked federated learning. Knowl. Based Syst. 257, 109929 (2022).","journal-title":"Knowl. Based Syst."},{"key":"2634_CR21","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-19551-w","volume":"11","author":"X Song","year":"2020","unstructured":"Song, X. et al. Cross-site transportability of an explainable artificial intelligence model for acute kidney injury prediction. Nat. Commun. 11, 5668 (2020).","journal-title":"Nat. Commun."},{"key":"2634_CR22","doi-asserted-by":"publisher","DOI":"10.1186\/s12916-019-1466-7","volume":"17","author":"B Van Calster","year":"2019","unstructured":"Van Calster, B. et al. Calibration: the Achilles heel of predictive analytics. BMC Med. 17, 230 (2019).","journal-title":"BMC Med."},{"key":"2634_CR23","doi-asserted-by":"publisher","first-page":"1133","DOI":"10.1213\/ANE.0000000000004489","volume":"130","author":"DA Colquhoun","year":"2020","unstructured":"Colquhoun, D. A. et al. Considerations for integration of perioperative electronic health records across institutions for research and quality improvement: the approach taken by the multicenter perioperative outcomes group. Anesth. Analg. 130, 1133\u20131146 (2020).","journal-title":"Anesth. Analg."},{"key":"2634_CR24","doi-asserted-by":"publisher","first-page":"116","DOI":"10.1038\/s41586-019-1390-1","volume":"572","author":"N Toma\u0161ev","year":"2019","unstructured":"Toma\u0161ev, N. et al. A clinically applicable approach to continuous prediction of future acute kidney injury. Nature 572, 116\u2013119 (2019).","journal-title":"Nature"},{"key":"2634_CR25","doi-asserted-by":"publisher","unstructured":"GitHub - ML4LHS\/gpmodels: a grammar of prediction models. GitHub https:\/\/doi.org\/10.5281\/zenodo.7158500.","DOI":"10.5281\/zenodo.7158500"},{"key":"2634_CR26","doi-asserted-by":"publisher","first-page":"774846","DOI":"10.3389\/fgene.2021.774846","volume":"12","author":"S Xu","year":"2021","unstructured":"Xu, S. et al. Use to effectively utilize plotting space to deal with large datasets and outliers. Front. Genet. 12, 774846 (2021).","journal-title":"Front. Genet."}],"container-title":["npj Digital Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02634-1","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02634-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02634-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,29]],"date-time":"2026-06-29T23:09:46Z","timestamp":1782774586000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02634-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,24]]},"references-count":26,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["2634"],"URL":"https:\/\/doi.org\/10.1038\/s41746-026-02634-1","relation":{},"ISSN":["2398-6352"],"issn-type":[{"value":"2398-6352","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,24]]},"assertion":[{"value":"5 May 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 April 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"24 April 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"All authors declare: M.R.M. has received a research grant from the US National Institutes of Health. K.S. receives research grant funding from the US National Institutes of Health and serves on Google\u2019s consumer health advisory panel as a paid consultant for unrelated work. A.M.J. has received research grants from the US National Institutes of Health. Outside of this work, D.S.L. receives funding from the Agency for Healthcare Research and Quality and National Institutes of Health. Non-author collaborator Dr. Robert Schonberger reports owning stock in Johnson & Johnson unrelated to the present work. Dr. Schonberger reports that his institution received support from Merck, Inc. for a study in which he was involved unrelated to the present work. No other relationships or activities that could appear to have influenced the submitted work are reported.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"492"}}