{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T05:47:09Z","timestamp":1775022429971,"version":"3.50.1"},"reference-count":25,"publisher":"Public Library of Science (PLoS)","issue":"2","license":[{"start":{"date-parts":[[2022,2,15]],"date-time":"2022-02-15T00:00:00Z","timestamp":1644883200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000009","name":"Foundation for the National Institutes of Health","doi-asserted-by":"publisher","award":["K01 HL136687"],"award-info":[{"award-number":["K01 HL136687"]}],"id":[{"id":"10.13039\/100000009","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000009","name":"Foundation for the National Institutes of Health","doi-asserted-by":"publisher","award":["R01 LM013325"],"award-info":[{"award-number":["R01 LM013325"]}],"id":[{"id":"10.13039\/100000009","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100007270","name":"University of Michigan","doi-asserted-by":"crossref","award":["Precision Health Award"],"award-info":[{"award-number":["Precision Health Award"]}],"id":[{"id":"10.13039\/100007270","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Quanta Computer"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS ONE"],"abstract":"<jats:sec id=\"sec001\"><jats:title>Importance<\/jats:title><jats:p>When hospitals are at capacity, accurate deterioration indices could help identify low-risk patients as potential candidates for home care programs and alleviate hospital strain. To date, many existing deterioration indices are based entirely on structured data from the electronic health record (EHR) and ignore potentially useful information from other sources.<\/jats:p><\/jats:sec><jats:sec id=\"sec002\"><jats:title>Objective<\/jats:title><jats:p>To improve the accuracy of existing deterioration indices by incorporating unstructured imaging data from chest radiographs.<\/jats:p><\/jats:sec><jats:sec id=\"sec003\"><jats:title>Design, setting, and participants<\/jats:title><jats:p>Machine learning models were trained to predict deterioration of patients hospitalized with acute dyspnea using existing deterioration index scores and chest radiographs. Models were trained on hospitalized patients without coronavirus disease 2019 (COVID-19) and then subsequently tested on patients with COVID-19 between January 2020 and December 2020 at a single tertiary care center who had at least one radiograph taken within 48 hours of hospital admission.<\/jats:p><\/jats:sec><jats:sec id=\"sec004\"><jats:title>Main outcomes and measures<\/jats:title><jats:p>Patient deterioration was defined as the need for invasive or non-invasive mechanical ventilation, heated high flow nasal cannula, IV vasopressor administration or in-hospital mortality at any time following admission. The EPIC deterioration index was augmented with unstructured data from chest radiographs to predict risk of deterioration. We compared discriminative performance of the models with and without incorporating chest radiographs using area under the receiver operating curve (AUROC), focusing on comparing the fraction and total patients identified as low risk at different negative predictive values (NPV).<\/jats:p><\/jats:sec><jats:sec id=\"sec005\"><jats:title>Results<\/jats:title><jats:p>Data from 6278 hospitalizations were analyzed, including 5562 hospitalizations without COVID-19 (training cohort) and 716 with COVID-19 (216 in validation, 500 in held-out test cohort). At a NPV of 0.95, the best-performing image-augmented deterioration index identified 49 more (9.8%) individuals as low-risk compared to the deterioration index based on clinical data alone in the first 48 hours of admission. At a NPV of 0.9, the EPIC image-augmented deterioration index identified 26 more individuals (5.2%) as low-risk compared to the deterioration index based on clinical data alone in the first 48 hours of admission.<\/jats:p><\/jats:sec><jats:sec id=\"sec006\"><jats:title>Conclusion and relevance<\/jats:title><jats:p>Augmenting existing deterioration indices with chest radiographs results in better identification of low-risk patients. The model augmentation strategy could be used in the future to incorporate other forms of unstructured data into existing disease models.<\/jats:p><\/jats:sec>","DOI":"10.1371\/journal.pone.0263922","type":"journal-article","created":{"date-parts":[[2022,2,15]],"date-time":"2022-02-15T18:40:19Z","timestamp":1644950419000},"page":"e0263922","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":7,"title":["Augmenting existing deterioration indices with chest radiographs to predict clinical deterioration"],"prefix":"10.1371","volume":"17","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7136-1012","authenticated-orcid":true,"given":"Emily","family":"Mu","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0808-4662","authenticated-orcid":true,"given":"Sarah","family":"Jabbour","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Adrian V.","family":"Dalca","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"John","family":"Guttag","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jenna","family":"Wiens","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Michael W.","family":"Sjoding","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2022,2,15]]},"reference":[{"issue":"5","key":"pone.0263922.ref001","doi-asserted-by":"crossref","first-page":"837","DOI":"10.1016\/j.jbi.2013.06.011","article-title":"Development and validation of a continuous measure of patient condition using the electronic medical record","volume":"46","author":"MJ Rothman","year":"2013","journal-title":"Journal of biomedical informatics"},{"issue":"6","key":"pone.0263922.ref002","doi-asserted-by":"crossref","first-page":"649","DOI":"10.1164\/rccm.201406-1022OC","article-title":"Multicenter development and validation of a risk stratification tool for ward patients","volume":"190","author":"MM Churpek","year":"2014","journal-title":"American journal of respiratory and critical care medicine"},{"key":"pone.0263922.ref003","article-title":"Evaluating a Widely Implemented Proprietary Deterioration Index Model Among Hospitalized COVID-19 Patients","author":"K Singh","year":"2020","journal-title":"Annals of the American Thoracic Society"},{"key":"pone.0263922.ref004","doi-asserted-by":"crossref","first-page":"e068576","DOI":"10.1136\/bmj-2021-068576","article-title":"\"Early Identification of patients admitted to hospital for covid-19 at risk of clinical deterioration: model development and multisite external validation study\"","volume":"376","author":"F Kamran","year":"2022","journal-title":"BMJ"},{"key":"pone.0263922.ref005","author":"DNP Cardenas","year":"2018","journal-title":"Differences in the Rothman Index Score in Evolving Emergent Events in Medical-Surgical Patients"},{"issue":"4","key":"pone.0263922.ref006","doi-asserted-by":"crossref","first-page":"425","DOI":"10.1017\/ice.2018.16","article-title":"A generalizable, data-driven approach to predict daily risk of Clostridium difficile infection at two large academic health centers","volume":"39","author":"J Oh","year":"2018","journal-title":"infection control & hospital epidemiology"},{"key":"pone.0263922.ref007","article-title":"Patient risk stratification for hospital-associated c","author":"JAM Wiens","year":"2012","journal-title":"diff as a time-series classification task"},{"issue":"3","key":"pone.0263922.ref008","doi-asserted-by":"crossref","first-page":"201","DOI":"10.7326\/0003-4819-144-3-200602070-00009","article-title":"Translating clinical research into clinical practice: impact of using prediction rules to make decisions","volume":"144","author":"BM Reilly","year":"2006","journal-title":"Annals of internal medicine"},{"issue":"4","key":"pone.0263922.ref009","doi-asserted-by":"crossref","first-page":"357","DOI":"10.1007\/s10278-008-9115-2","article-title":"Radiologist use of and perceived need for patient data access","volume":"22","author":"WW Boonn","year":"2009","journal-title":"Journal of digital imaging"},{"issue":"874","key":"pone.0263922.ref010","doi-asserted-by":"crossref","first-page":"1052","DOI":"10.1259\/bjr.73.874.11271897","article-title":"The influence of clinical information on the reporting of CT by radiologists","volume":"73","author":"A Leslie","year":"2000","journal-title":"The British journal of radiology"},{"issue":"1","key":"pone.0263922.ref011","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s12890-020-01286-5","article-title":"Chest x-ray findings and temporal lung changes in patients with COVID-19 pneumonia","volume":"20","author":"LA Rousan","year":"2020","journal-title":"BMC Pulmonary Medicine"},{"issue":"1","key":"pone.0263922.ref012","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s43055-020-00296-x","article-title":"Chest X-ray findings monitoring COVID-19 disease course and severity","volume":"51","author":"R Yasin","year":"2020","journal-title":"Egyptian Journal of Radiology and Nuclear Medicine"},{"issue":"1","key":"pone.0263922.ref013","doi-asserted-by":"crossref","first-page":"21","DOI":"10.7326\/M20-1260","article-title":"Locally informed simulation to predict hospital capacity needs during the COVID-19 pandemic","volume":"173","author":"GE Weissman","year":"2020","journal-title":"Annals of internal medicine"},{"key":"pone.0263922.ref014","article-title":"Densenet: Implementing efficient convnet descriptor pyramids","author":"F Iandola","year":"2014","journal-title":"arXiv preprint arXiv14041869"},{"key":"pone.0263922.ref015","first-page":"590","author":"J Irvin","journal-title":"Chexpert: A large chest radiograph dataset with uncertainty labels and expert comparison"},{"issue":"1","key":"pone.0263922.ref016","first-page":"1","article-title":"MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports","volume":"6","author":"AEW Johnson","year":"2019","journal-title":"Scientific data"},{"key":"pone.0263922.ref017","author":"A Sriram","year":"2021","journal-title":"COVID-19 Prognosis via Self-Supervised Representation Learning and Multi-Image Prediction"},{"key":"pone.0263922.ref018","article-title":"Validation and comparison of PICTURE analytic and Epic Deterioration Index for COVID-19","author":"BC Cummings","year":"2020","journal-title":"medRxiv"},{"key":"pone.0263922.ref019","article-title":"Deep learning models for COVID-19 chest x-ray classification: Preventing shortcut learning using feature disentanglement","author":"C Robinson","year":"2021","journal-title":"medRxiv"},{"issue":"3","key":"pone.0263922.ref020","doi-asserted-by":"crossref","first-page":"915","DOI":"10.1007\/s13246-020-00888-x","article-title":"Truncated inception net: COVID-19 outbreak screening using chest X-rays","volume":"43","author":"D Das","year":"2020","journal-title":"Physical and engineering sciences in medicine"},{"issue":"2","key":"pone.0263922.ref021","doi-asserted-by":"crossref","first-page":"537","DOI":"10.1148\/radiol.2018181422","article-title":"Assessment of convolutional neural networks for automated classification of chest radiographs","volume":"290","author":"JA Dunnmon","year":"2019","journal-title":"Radiology"},{"key":"pone.0263922.ref022","doi-asserted-by":"crossref","DOI":"10.1155\/2020\/9756518","article-title":"Review on diagnosis of COVID-19 from chest CT images using artificial intelligence","volume":"2020","author":"I Ozsahin","year":"2020","journal-title":"Computational and Mathematical Methods in Medicine"},{"key":"pone.0263922.ref023","article-title":"Scibert: A pretrained language model for scientific text","author":"I Beltagy","year":"2019","journal-title":"arXiv preprint arXiv:190310676"},{"key":"pone.0263922.ref024","doi-asserted-by":"crossref","first-page":"107826","DOI":"10.1016\/j.patcog.2021.107826","article-title":"Momentum contrastive learning for few-shot COVID-19 diagnosis from chest CT images","volume":"113","author":"X Chen","year":"2021","journal-title":"Pattern Recognition"},{"issue":"10","key":"pone.0263922.ref025","doi-asserted-by":"crossref","first-page":"910","DOI":"10.1056\/NEJMoa1214726","article-title":"Probability of cancer in pulmonary nodules detected on first screening CT","volume":"369","author":"A McWilliams","year":"2013","journal-title":"New England Journal of Medicine"}],"container-title":["PLOS ONE"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0263922","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,1,27]],"date-time":"2023-01-27T00:46:42Z","timestamp":1674780402000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0263922"}},"subtitle":[],"editor":[{"given":"Gopal Krishna","family":"Dhali","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2022,2,15]]},"references-count":25,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2022,2,15]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0263922","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,2,15]]}}}