{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T14:56:43Z","timestamp":1774882603258,"version":"3.50.1"},"posted":{"date-parts":[[2026,3,27]]},"group-title":"Health Informatics","reference-count":32,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,3,27]],"date-time":"2026-03-27T00:00:00Z","timestamp":1774569600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,3,27]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>Digital health technologies, including machine learning (ML), are transforming infectious disease management, however ML models for HIV care have been limited by data sharing restrictions that prevent multi-site collaboration. Federated Learning (FL) offers a privacy-preserving solution, enabling cross-site model training without sharing patient-level data. We evaluated FL for developing clinical prediction models using data from 22,234 people living with HIV (PLWH) across six sites in five countries within the Caribbean, Central, and South America network for HIV epidemiology (CCASAnet). Across four prediction tasks \u2014 1-year mortality, 3-year mortality, tuberculosis incidence, and AIDS-defining cancer incidence \u2014 FL algorithms achieved near-centralized performance while substantially outperforming site-specific models. Performance gains varied across sites, driven by both site size and between-site heterogeneity. Local fine-tuning often improved FL performance, though benefits were task dependent. These findings support FL as a scalable, privacy-preserving infrastructure for multi-site ML in international HIV research.<\/jats:p>","DOI":"10.64898\/2026.03.25.26349286","type":"posted-content","created":{"date-parts":[[2026,3,27]],"date-time":"2026-03-27T17:55:29Z","timestamp":1774634129000},"source":"Crossref","is-referenced-by-count":0,"title":["Federated Learning Performance Depends on Site Variation in Global HIV Data Consortia"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6763-599X","authenticated-orcid":false,"given":"Nicholas J.","family":"Jackson","sequence":"first","affiliation":[{"name":"Department of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6719-1388","authenticated-orcid":false,"given":"Chao","family":"Yan","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yanink","family":"Caro-Vega","sequence":"additional","affiliation":[{"name":"Departamento de Infectolog\u00eda, Instituto Nacional de Ciencias M\u00e9dicas y Nutrici\u00f3n Salvador Zubir\u00e1n, M\u00e9xico City, M\u00e9xico"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fabio","family":"Paredes","sequence":"additional","affiliation":[{"name":"Departamento de Estad\u00edstica, Pontificia Universidad Cat\u00f3lica de Chile, Santiago, Chile"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ronaldo","family":"Ismerio Moreira","sequence":"additional","affiliation":[{"name":"Instituto Nacional de Infectologia Evandro Chagas, Funda\u00e7\u00e3o Oswaldo Cruz, Rio de Janeiro, Brazil"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Stanley","family":"Cadet","sequence":"additional","affiliation":[{"name":"GHESKIO, Port-au-Prince, Haiti"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Diana","family":"Varela","sequence":"additional","affiliation":[{"name":"Instituto Hondure\u00f1o de Seguridad Social, Tegucigalpa, Honduras"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Carina","family":"Cesar","sequence":"additional","affiliation":[{"name":"Fundaci\u00f3n Hu\u00e9sped, Buenos Aires, Argentina"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Stephany N","family":"Duda","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Bryan E.","family":"Shepherd","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, USA"},{"name":"Department of Biostatistics, Vanderbilt University Medical Center, Nashville, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Bradley A.","family":"Malin","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, USA"},{"name":"Department of Biostatistics, Vanderbilt University Medical Center, Nashville, USA"},{"name":"Department of Computer Science, Vanderbilt University, Nashville, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2026033007101149000_2026.03.25.26349286v1.1","doi-asserted-by":"publisher","DOI":"10.1126\/science.adh1114"},{"key":"2026033007101149000_2026.03.25.26349286v1.2","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmi.2019.09.009"},{"key":"2026033007101149000_2026.03.25.26349286v1.3","doi-asserted-by":"crossref","first-page":"7666","DOI":"10.3934\/mbe.2021380","article-title":"Mechanical and Industrial Engineering Department, University of Massachusetts Amherst, Amherst, MA 01003, USA. A reinforcement learning model to inform optimal decision paths for HIV elimination","volume":"18","year":"2021","journal-title":"Math. Biosci. Eng"},{"key":"2026033007101149000_2026.03.25.26349286v1.4","doi-asserted-by":"crossref","first-page":"474","DOI":"10.1186\/s12879-024-09368-z","article-title":"The predictive accuracy of machine learning for the risk of death in HIV patients: a systematic review and meta-analysis","volume":"24","year":"2024","journal-title":"BMC Infect. Dis"},{"key":"2026033007101149000_2026.03.25.26349286v1.5","doi-asserted-by":"publisher","DOI":"10.1093\/cid\/ciaf149"},{"key":"2026033007101149000_2026.03.25.26349286v1.6","doi-asserted-by":"crossref","first-page":"393","DOI":"10.1186\/s12905-024-03232-7","article-title":"Prediction of precancerous cervical cancer lesions among women living with HIV on antiretroviral therapy in Uganda: a comparison of supervised machine learning algorithms","volume":"24","year":"2024","journal-title":"BMC Womens Health"},{"key":"2026033007101149000_2026.03.25.26349286v1.7","doi-asserted-by":"crossref","first-page":"ofab130","DOI":"10.1093\/ofid\/ofab130","article-title":"Development and Validation of a Multivariable Prediction Model for Missed HIV Health Care Provider Visits in a Large US Clinical Cohort","volume":"8","year":"2021","journal-title":"Open Forum Infect. Dis"},{"key":"2026033007101149000_2026.03.25.26349286v1.8","doi-asserted-by":"crossref","first-page":"e54","DOI":"10.1016\/S2352-3018(21)00247-2","article-title":"Application of artificial intelligence and machine learning for HIV prevention interventions","volume":"9","year":"2022","journal-title":"Lancet HIV"},{"key":"2026033007101149000_2026.03.25.26349286v1.9","doi-asserted-by":"crossref","first-page":"e70045","DOI":"10.1002\/jia2.70045","article-title":"Artificial intelligence for HIV care: a global systematic review of current studies and emerging trends","volume":"28","year":"2025","journal-title":"J. Int. AIDS Soc"},{"key":"2026033007101149000_2026.03.25.26349286v1.10","doi-asserted-by":"publisher","DOI":"10.1093\/ije\/dym073"},{"key":"2026033007101149000_2026.03.25.26349286v1.11","doi-asserted-by":"publisher","DOI":"10.1093\/ije\/dyr080"},{"key":"2026033007101149000_2026.03.25.26349286v1.12","doi-asserted-by":"publisher","DOI":"10.1093\/infdis\/jiaa236"},{"key":"2026033007101149000_2026.03.25.26349286v1.13","doi-asserted-by":"publisher","DOI":"10.1186\/1471-2458-14-1144"},{"key":"2026033007101149000_2026.03.25.26349286v1.14","doi-asserted-by":"crossref","first-page":"215","DOI":"10.1038\/s41746-024-01205-6","article-title":"EU-US data transfers: an enduring challenge for health research collaborations","volume":"7","year":"2024","journal-title":"NPJ Digit. Med"},{"key":"2026033007101149000_2026.03.25.26349286v1.15","doi-asserted-by":"crossref","first-page":"119","DOI":"10.1038\/s41746-020-00323-1","article-title":"The future of digital health with federated learning","volume":"3","year":"2020","journal-title":"Npj Digit. Med"},{"key":"2026033007101149000_2026.03.25.26349286v1.16","doi-asserted-by":"crossref","first-page":"60","DOI":"10.1038\/s41746-021-00431-6","article-title":"Federated deep learning for detecting COVID-19 lung abnormalities in CT: a privacy-preserving multinational validation study","volume":"4","year":"2021","journal-title":"Npj Digit. Med"},{"key":"2026033007101149000_2026.03.25.26349286v1.17","doi-asserted-by":"crossref","first-page":"59","DOI":"10.1016\/j.ijmedinf.2018.01.007","article-title":"Federated learning of predictive models from federated Electronic Health Records","volume":"112","year":"2018","journal-title":"Int. J. Med. Inf"},{"key":"2026033007101149000_2026.03.25.26349286v1.18","doi-asserted-by":"crossref","first-page":"1377","DOI":"10.3390\/s22041377","article-title":"FL-PMI: Federated Learning-Based Person Movement Identification through Wearable Devices in Smart Healthcare Systems","volume":"22","year":"2022","journal-title":"Sensors"},{"key":"2026033007101149000_2026.03.25.26349286v1.19","doi-asserted-by":"crossref","first-page":"83","DOI":"10.5334\/aogh.4277","article-title":"Advancing the Global Fight Against HIV\/Aids: Strategies, Barriers, and the Road to Eradication","volume":"89","year":"2023","journal-title":"Ann. Glob. Health"},{"key":"2026033007101149000_2026.03.25.26349286v1.20","doi-asserted-by":"publisher","DOI":"10.1016\/S1473-3099(24)00176-2"},{"key":"2026033007101149000_2026.03.25.26349286v1.21","doi-asserted-by":"crossref","first-page":"17334","DOI":"10.7448\/IAS.15.2.17334","article-title":"HIV treatment and care in resource-constrained environments: challenges for the next decade","volume":"15","year":"2012","journal-title":"J. Int. AIDS Soc"},{"key":"2026033007101149000_2026.03.25.26349286v1.22","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-27173-1_9"},{"key":"2026033007101149000_2026.03.25.26349286v1.23","doi-asserted-by":"publisher","DOI":"10.2217\/hiv.10.6"},{"key":"2026033007101149000_2026.03.25.26349286v1.24","doi-asserted-by":"publisher","DOI":"10.1007\/s11904-014-0204-1"},{"key":"2026033007101149000_2026.03.25.26349286v1.25","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2408.04154"},{"key":"2026033007101149000_2026.03.25.26349286v1.26","unstructured":"Flower: The Friendly Federated AI Framework."},{"key":"2026033007101149000_2026.03.25.26349286v1.27","doi-asserted-by":"publisher","DOI":"10.1093\/jamia\/ocaf198"},{"key":"2026033007101149000_2026.03.25.26349286v1.28","unstructured":"Vapnik, V . Principles of risk minimization for learning theory. in Proceedings of the 5th International Conference on Neural Information Processing Systems 831\u2013838 (Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, 1991)."},{"key":"2026033007101149000_2026.03.25.26349286v1.29","unstructured":"McMahan, H. B. , Moore, E. , Ramage, D. , Hampson, S. & Arcas, B. A. y. Communication-Efficient Learning of Deep Networks from Decentralized Data. Preprint at http:\/\/arxiv.org\/abs\/1602.05629 (2023)."},{"key":"2026033007101149000_2026.03.25.26349286v1.30","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1812.06127"},{"key":"2026033007101149000_2026.03.25.26349286v1.31","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1907.10902"},{"key":"2026033007101149000_2026.03.25.26349286v1.32","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1905.12022"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.03.25.26349286","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T14:10:22Z","timestamp":1774879822000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.03.25.26349286"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,27]]},"references-count":32,"URL":"https:\/\/doi.org\/10.64898\/2026.03.25.26349286","relation":{},"subject":[],"published":{"date-parts":[[2026,3,27]]},"subtype":"preprint"}}