{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,23]],"date-time":"2026-06-23T11:01:58Z","timestamp":1782212518535,"version":"3.54.5"},"reference-count":54,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,1,8]],"date-time":"2024-01-08T00:00:00Z","timestamp":1704672000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>The move from a reactive model of care which treats conditions when they arise to a proactive model which intervenes early to prevent adverse healthcare events will benefit from advances in the predictive capabilities of Artificial Intelligence and Machine Learning. This paper investigates the ability of a Deep Learning (DL) approach to predict future disease diagnosis from Electronic Health Records (EHR) for the purposes of Population Health Management.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>In this study, embeddings were created using a Word2Vec algorithm from structured vocabulary commonly used in EHRs e.g., Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) codes. This study is based on longitudinal medical data from ~50 m patients in the USA. We introduced a novel method of including binned observation values into an embeddings model. We also included novel features associated with wider determinants of health. Patient records comprising these embeddings were then fed to a Bidirectional Gated Recurrent Unit (GRU) model to predict the likelihood of patients developing Type 2 Diabetes Mellitus, Chronic Obstructive Pulmonary Disorder (COPD), Hypertension or experiencing an Acute Myocardial Infarction (MI) in the next 3 years. SHapley Additive exPlanations (SHAP) values were calculated to achieve model explainability.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Increasing the data scope to include binned observations and wider determinants of health was found to improve predictive performance. We achieved an area under the Receiver Operating Characteristic curve value of 0.92 for Diabetes prediction, 0.94 for COPD, 0.92 for Hypertension and 0.94 for MI. The SHAP values showed that the models had learned features known to be associated with these outcomes.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>The DL approach outlined in this study can identify clinically-relevant features from large-scale EHR data and use these to predict future disease outcomes. This study highlights the promise of DL solutions for identifying patients at future risk of disease and providing clinicians with the means to understand and evaluate the drivers of those predictions.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frai.2023.1287541","type":"journal-article","created":{"date-parts":[[2024,1,8]],"date-time":"2024-01-08T05:04:10Z","timestamp":1704690250000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":24,"title":["Predicting disease onset from electronic health records for population health management: a scalable and explainable Deep Learning approach"],"prefix":"10.3389","volume":"6","author":[{"given":"Robert","family":"Grout","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rishab","family":"Gupta","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ruby","family":"Bryant","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mawada A.","family":"Elmahgoub","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yijie","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Khushbakht","family":"Irfanullah","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rahul F.","family":"Patel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jake","family":"Fawkes","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Catherine","family":"Inness","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,1,8]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1409.0473","article-title":"Neural machine translation by jointly learning to align and translate","author":"Bahdanau","year":"2014","journal-title":"arXiv"},{"key":"B2","doi-asserted-by":"crossref","DOI":"10.1142\/9789811215636_0027","article-title":"\u201cClinical concept embeddings learned from massive sources of multimodal medical data,\u201d","volume-title":"Pacific Symposium on Biocomputing 2020","author":"Beam","year":"2019"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1377\/hlthaff.27.3.759","article-title":"The triple aim: care, health, and cost","author":"Berwick","year":"2008","journal-title":"Health Aff"},{"key":"B4","doi-asserted-by":"publisher","first-page":"420","DOI":"10.1016\/j.ypmed.2015.09.016","article-title":"Lack of private health insurance is associated with higher mortality from cancer and other chronic diseases, poor diet quality, and inflammatory biomarkers in the united states","volume":"81","author":"Bittoni","year":"2015","journal-title":"Prev. Med."},{"key":"B5","article-title":"A Vision for Population Health: Towards a Healthier Future","volume-title":"Towards a Healthier Future","author":"Buck","year":"2018"},{"key":"B6","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2018\/554","article-title":"Medical concept embedding with time-aware attention","author":"Cai","year":"2018","journal-title":"arXiv [preprint]"},{"key":"B7","doi-asserted-by":"publisher","first-page":"357","DOI":"10.1016\/j.ejor.2023.06.036","article-title":"Interpretable machine learning for imbalanced credit scoring datasets","volume":"312","author":"Chen","year":"2023","journal-title":"Eur. J. Oper. Res."},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1602.03686","article-title":"Medical concept representation learning from electronic health records and its application on heart failure prediction","author":"Choi","year":"","journal-title":"arXiv [preprint]."},{"key":"B9","article-title":"\u201cMulti-layer representation learning for medical concepts,\u201d","author":"Choi","year":"","journal-title":"Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining"},{"key":"B10","article-title":"\u201cLearning low-dimensional representations of medical concepts,\u201d","author":"Choi","year":"","journal-title":"AMIA Summits on Translational Science Proceedings"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.1093\/jamiaopen\/ooac097","article-title":"Predicting hypertension onset from longitudinal electronic health records with deep learning","author":"Datta","year":"2022","journal-title":"JAMIA Open 5"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1810.04805","article-title":"Bert: pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018","journal-title":"arXiv [preprint]"},{"key":"B13","article-title":"Snomed-ct: the advanced terminology and coding system for ehealth","author":"Donnelly","year":"2006","journal-title":"Stud. Health Technol. Inform"},{"key":"B14","year":"2023","journal-title":"Towards Better Prevention and Management of Chronic Diseases"},{"key":"B15","unstructured":"Self-Perceived Health Statistics2023"},{"key":"B16","doi-asserted-by":"publisher","first-page":"1994","DOI":"10.2337\/dc22-0242","article-title":"Undiagnosed diabetes in us adults: prevalence and trends","volume":"45","author":"Fang","year":"2022","journal-title":"Diabetes Care"},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.1016\/j.pmedr.2018.10.008","article-title":"The global burden of multiple chronic conditions: a narrative review","author":"Hajat","year":"2018","journal-title":"Prev. Med. Rep"},{"key":"B18","doi-asserted-by":"publisher","first-page":"146","DOI":"10.1080\/00437956.1954.11659520","article-title":"Distributional structure","volume":"10","author":"Harris","year":"1954","journal-title":"Word"},{"key":"B19","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-018-0307-0","article-title":"The practical implementation of artificial intelligence technologies in medicine","author":"He","year":"2019","journal-title":"Nat. Med"},{"key":"B20","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory","author":"Hochreiter","year":"1997","journal-title":"Neural Comput"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.1002\/acr2.11114","article-title":"The relation of the chronic disease epidemic to the health care crisis","author":"Holman","year":"2020","journal-title":"ACR Open Rheumatol"},{"key":"B22","volume-title":"Health Insurance Coverage in the United States: 2019","author":"Keisler-Starkey","year":"2020"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2022.3224727","article-title":"Hi-behrt: hierarchical transformer-based model for accurate prediction of clinical events using multimodal longitudinal electronic health records","author":"Li","year":"2022","journal-title":"IEEE J. Biomed. Health Inf"},{"key":"B24","doi-asserted-by":"publisher","first-page":"7155","DOI":"10.1038\/s41598-020-62922-y","article-title":"Behrt: transformer for electronic health records","volume":"10","author":"Li","year":"2020","journal-title":"Sci. Rep."},{"key":"B25","doi-asserted-by":"publisher","first-page":"107161","DOI":"10.1016\/j.cmpb.2022.107161","article-title":"Application of explainable artificial intelligence for healthcare: a systematic review of the last decade (2011-2022)","volume":"226","author":"Loh","year":"2022","journal-title":"Comp. Methods Prog. Biomed"},{"key":"B26","article-title":"A unified approach to interpreting model predictions","author":"Lundberg","year":"2017","journal-title":"Adv. Neural Inf. Process. Syst"},{"key":"B27","author":"Lundberg","year":"2021","journal-title":"SHAP: SHapley Additive exPlanations"},{"key":"B28","unstructured":"LundbergS. M.\n            LeeS.-I.\n          Shap Documentation: Shap Force Plot2023"},{"key":"B29","volume-title":"The Promise of Population Health Management in England: From Theory to Implementation","author":"Main","year":"2022"},{"key":"B30","doi-asserted-by":"publisher","first-page":"103655","DOI":"10.1016\/j.jbi.2020.103655","article-title":"The role of explainability in creating trustworthy artificial intelligence for health care: a comprehensive survey of the terminology, design choices, and evaluation strategies","volume":"113","author":"Markus","year":"2021","journal-title":"J. Biomed. Inform."},{"key":"B31","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2021.3063721","article-title":"Bidirectional representation learning from transformers using multimodal electronic health record data to predict depression","author":"Meng","year":"2021","journal-title":"IEEE J. Biomed. Health Inf"},{"key":"B32","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1301.3781","article-title":"Efficient estimation of word representations in vector space","author":"Mikolov","year":"2013","journal-title":"arXiv [preprint]"},{"key":"B33","article-title":"\u201cCehr-bert: incorporating temporal information from structured ehr data to improve prediction tasks,\u201d","volume-title":"Machine Learning for Health","author":"Pang","year":"2021"},{"key":"B34","first-page":"1","article-title":"Hospitals' use of electronic health records data, 2015-2017","volume":"46","author":"Parasrampuria","year":"2019","journal-title":"ONC Data Brief"},{"key":"B35","doi-asserted-by":"crossref","DOI":"10.3115\/v1\/D14-1162","article-title":"\u201cGlove: global vectors for word representation,\u201d","author":"Pennington","year":"2014","journal-title":"Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing"},{"key":"B36","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-31750-2_3","article-title":"\u201cDeepcare: a deep dynamic memory model for predictive medicine,\u201d","volume-title":"Advances in Knowledge Discovery and Data Mining: 20th Pacific-Asia Conference, PAKDD 2016, Auckland, New Zealand, April 19-22, 2016","author":"Pham","year":"2016"},{"key":"B37","doi-asserted-by":"publisher","first-page":"86","DOI":"10.1038\/s41746-021-00455-y","article-title":"Med-bert: pretrained contextualized embeddings on large-scale structured electronic health records for disease prediction","volume":"4","author":"Rasmy","year":"2021","journal-title":"NPJ Dig. Med."},{"key":"B38","doi-asserted-by":"publisher","first-page":"e2111315","DOI":"10.1001\/jamanetworkopen.2021.11315","article-title":"Development and validation of a machine learning model using administrative health data to predict onset of type 2 diabetes","volume":"4","author":"Ravaut","year":"2021","journal-title":"JAMA Netw. Open"},{"key":"B39","doi-asserted-by":"publisher","DOI":"10.1111\/1468-0009.00171","author":"Ross","year":"2000","journal-title":"Does medical insurance contribute to socioeconomic differentials in health, Milbank Q"},{"key":"B40","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1907.09600","article-title":"Evaluation of embeddings of laboratory test codes for patients at a cancer center","author":"Rossi","year":"2019","journal-title":"arXiv [preprint]"},{"key":"B41","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-39539-0_7","article-title":"Exbehrt: Extended transformer for electronic health records to predict disease subtypes & progressions","author":"Rupp","year":"2023","journal-title":"arXiv [preprint]"},{"key":"B42","doi-asserted-by":"publisher","first-page":"866","DOI":"10.1001\/jama.2023.14217","article-title":"Creation and adoption of large language models in medicine","volume":"330","author":"Shah","year":"2023","journal-title":"JAMA"},{"key":"B43","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2019\/825","article-title":"Pre-training of graph augmented transformers for medication recommendation","author":"Shang","year":"2019","journal-title":"arXiv [preprint]."},{"key":"B44","doi-asserted-by":"publisher","DOI":"10.1016\/j.jbi.2020.103671","article-title":"Deep representation learning of patient data from electronic health records (ehr): a systematic review. J. Biomed","author":"Si","year":"2021","journal-title":"Inform"},{"key":"B45","article-title":"Population health perspective on high users of health care: Role of family physicians","author":"Stone","year":"2014","journal-title":"Can. Fam. Phys"},{"key":"B46","doi-asserted-by":"publisher","DOI":"10.1016\/j.amepre.2018.12.010","article-title":"Health insurance coverage among us workers: differences by work arrangements in 2010 and 2015","author":"Su","year":"2019","journal-title":"Am. J. Prev. Med"},{"key":"B47","doi-asserted-by":"publisher","DOI":"10.1002\/edm2.306","article-title":"Clinical outcomes in high-hypoglycaemia-risk patients with type 2 diabetes switching to insulin glargine 300u\/ml versus a first-generation basal insulin analogue in the united states: results from the deliver high risk real-world study","author":"Sullivan","year":"2021","journal-title":"Endocrinol. Diabetes Metab"},{"key":"B48","volume-title":"How Data-Driven Population Health Management Will Shape Systems, Response to Health Inequalities and Secondary Prevention","author":"Tang","year":"2023"},{"key":"B49","author":"Tonekaboni","year":"2019","journal-title":"What Clinicians Want: Contextualizing Explainable Machine Learning for Clinical End Use"},{"key":"B50","doi-asserted-by":"publisher","DOI":"10.1016\/j.jaad.2019.02.064","article-title":"Prevalence estimates for chronic urticaria in the united states: a sex-and age-adjusted population analysis","author":"Wertenteil","year":"2019","journal-title":"J. Am. Acad. Dermatol"},{"key":"B51","year":"2023","journal-title":"Population Health Management in Primary Health Care: a Proactive Approach to Improve Health and Well-Being: Primary Health Care Policy Paper Series"},{"key":"B52","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-023-00879-8","article-title":"The shaky foundations of large language models and foundation models for electronic health records","author":"Wornow","year":"2023","journal-title":"npj Dig. Med"},{"key":"B53","doi-asserted-by":"publisher","DOI":"10.1186\/s12911-019-0766-3","article-title":"Time-sensitive clinical concept embeddings learned from large electronic health records","author":"Xiang","year":"2019","journal-title":"BMC Med. Inform. Decis. Mak"},{"key":"B54","doi-asserted-by":"publisher","DOI":"10.1186\/1472-6947-15-S4-S1","article-title":"Predictive modeling of structured electronic health records for adverse drug event detection","author":"Zhao","year":"2015","journal-title":"BMC Med. Inform. Decis. Mak"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2023.1287541\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,1,8]],"date-time":"2024-01-08T05:04:31Z","timestamp":1704690271000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2023.1287541\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,1,8]]},"references-count":54,"alternative-id":["10.3389\/frai.2023.1287541"],"URL":"https:\/\/doi.org\/10.3389\/frai.2023.1287541","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,1,8]]},"article-number":"1287541"}}