{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,19]],"date-time":"2026-05-19T13:15:14Z","timestamp":1779196514862,"version":"3.51.4"},"reference-count":38,"publisher":"BMJ","issue":"5","license":[{"start":{"date-parts":[[2026,5,19]],"date-time":"2026-05-19T00:00:00Z","timestamp":1779148800000},"content-version":"unspecified","delay-in-days":18,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/100031019","name":"Digital Technology Supercluster","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100031019","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["BMJ Open"],"accepted":{"date-parts":[[2026,4,30]]},"published-print":{"date-parts":[[2026,5]]},"abstract":"<jats:sec>\n                    <jats:title>Objectives<\/jats:title>\n                    <jats:p>To develop, train and test a deep learning model Image-based PROgnostication applied to Chest X Rays (IPRO-X) tool that predicts the inpatient (IP) admission risk in patients with suspected pneumonia presenting to the emergency department (ED).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Design<\/jats:title>\n                    <jats:p>The study consists of a retrospective (training) and prospective non-interventional shadow deployment (validation) of a deep learning model.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Setting<\/jats:title>\n                    <jats:p>Three-hospital tertiary care system with two EDs.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Participants<\/jats:title>\n                    <jats:p>Consecutive adult patients (18 years old or older) who presented to the ED from December 2022 to February 2023 with a chief clinical complaint potentially related to pneumonia were reviewed for eligibility (n=5567), with a final number of 3677 included in the validation study.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Primary and secondary outcome measures<\/jats:title>\n                    <jats:p>IPRO-X was developed using standard two-dimensional convolutional neural network InceptionNet architecture and processes chest radiographs (CXRs) acquired at admission generating a continuous value from 0 (IP-negative) to 1 (IP-positive). We examined IPRO-X\u2019s ability to predict IP admission using accuracy, area under the curve (AUC), receiver operating characteristic, sensitivity, specificity, positive predictive value and negative predictive value. IPRO-X scores were compared against observed outcomes (admission and discharge) to determine clinical utility in an ED setting. Four thresholds were defined from the retrospective phase: Youden optimal, highest specificity and thresholds equivalent to 30-day mortality rates of Pneumonia Severity Index (PSI) Risk class IV and Confusion, Respiratory Rate, Blood Pressure and Age 65 or older Score (CRB-65) scores 1 and 2. Performance was also analysed per group of chief complaints.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>In the validation set, 3677 patients (1777 (48%) female, median age 56 years (min 18, max 99)) were included. IPRO-X predicted IP admissions with an AUC of 0.795, and in the pneumonia-specific chief complaints, AUC increased to 0.828, whereas in non-related pneumonia chief complaints, the AUC decreased to 0.755. IPRO-X score was significantly higher in admitted patients compared with discharged patients (p&lt;0.001). The accuracy of IPRO-X when using the PSI-anchored threshold was 0.751 and with the CRB-anchored threshold was 0.729, with a specificity of 0.803 and 0.968, respectively.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>IPRO-X applied to CXR of patients with signs and symptoms commonly related to pneumonia in the ED can accurately predict IP admission.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1136\/bmjopen-2025-108694","type":"journal-article","created":{"date-parts":[[2026,5,19]],"date-time":"2026-05-19T13:00:20Z","timestamp":1779195620000},"page":"e108694","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":0,"title":["Evaluation of a deep learning model applied to chest X-rays of patients with suspected pneumonia presenting to the emergency department designed to predict admission risk: a retrospective training and prospective non-intervention validation"],"prefix":"10.1136","volume":"16","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8196-866X","authenticated-orcid":false,"given":"Eduardo","family":"Pimenta Ribeiro Pontes Almeida","sequence":"first","affiliation":[{"name":"Joint Department of Medical Imaging, University Medical Imaging Toronto, University Health Network, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sameer","family":"Masood","sequence":"additional","affiliation":[{"name":"Division of Emergency Medicine, Temerty Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Felix","family":"Baldauf-Lenschen","sequence":"additional","affiliation":[{"name":"Altis Labs Inc, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Thomas Jay","family":"Hennessy","sequence":"additional","affiliation":[{"name":"Altis Labs Inc, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shazia","family":"Akbar","sequence":"additional","affiliation":[{"name":"Altis Labs Inc, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yangqing","family":"Deng","sequence":"additional","affiliation":[{"name":"Department of Biostatistics, University Health Network, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Felipe","family":"Soares Torres","sequence":"additional","affiliation":[{"name":"Joint Department of Medical Imaging, University Medical Imaging Toronto, University Health Network, Toronto, Ontario, Canada"},{"name":"Department of Medical Imaging, Temerty Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"239","published-online":{"date-parts":[[2026,5,19]]},"reference":[{"key":"2026051906001130000_16.5.e108694.1","unstructured":"World Health Organization . The top 10 causes of death, 2020. Available: https:\/\/www.who.int\/news-room\/fact-sheets\/detail\/the-top-10-causes-of-death"},{"key":"2026051906001130000_16.5.e108694.2","unstructured":"Canadian Institute for Health Information . NACRS Emergency Department Visits: Volumes and Median Lengths of Stay, 2003\u20132004 to 2021\u20132022 \u2014 Supplementary Statistics. Ottawa. 2022."},{"key":"2026051906001130000_16.5.e108694.3","doi-asserted-by":"publisher","DOI":"10.1164\/rccm.201002-0272CI"},{"key":"2026051906001130000_16.5.e108694.4","doi-asserted-by":"crossref","first-page":"665","DOI":"10.1016\/j.emc.2018.07.001","article-title":"Community-Acquired Pneumonia","volume":"36","author":"Rider","year":"2018","journal-title":"Emerg Med Clin North Am"},{"key":"2026051906001130000_16.5.e108694.5","doi-asserted-by":"crossref","DOI":"10.1136\/bmj-2021-065871","article-title":"Management of pneumonia in critically ill patients","volume":"375","author":"Cill\u00f3niz","year":"2021","journal-title":"BMJ"},{"key":"2026051906001130000_16.5.e108694.6","doi-asserted-by":"crossref","first-page":"11","DOI":"10.12927\/hcq.2022.26717","article-title":"How Were Wait Times for Priority Procedures in Canada Impacted during the First Six Months of the COVID-19 Pandemic?","volume":"24","author":"Reason","year":"2022","journal-title":"Healthc Q"},{"key":"2026051906001130000_16.5.e108694.7","doi-asserted-by":"publisher","DOI":"10.1056\/NEJM199701233360402"},{"key":"2026051906001130000_16.5.e108694.8","doi-asserted-by":"publisher","DOI":"10.3399\/bjgp10X532422"},{"key":"2026051906001130000_16.5.e108694.9","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijid.2020.06.038"},{"key":"2026051906001130000_16.5.e108694.10","doi-asserted-by":"publisher","DOI":"10.1164\/ajrccm.163.7.at1010"},{"key":"2026051906001130000_16.5.e108694.11","doi-asserted-by":"publisher","DOI":"10.1136\/thoraxjnl-2014-206744"},{"key":"2026051906001130000_16.5.e108694.12","doi-asserted-by":"crossref","first-page":"174","DOI":"10.36497\/respirsci.v1i3.25","article-title":"Accuracy Between CURB-65 Score and PSI in Determining The Prognosis of Community-Acquired Pneumonia Patients at H. Adam Malik General Hospital, Medan","volume":"1","author":"Pakpahan","year":"2021","journal-title":"Respir Sci"},{"key":"2026051906001130000_16.5.e108694.13","doi-asserted-by":"publisher","DOI":"10.1136\/thx.2008.099028"},{"key":"2026051906001130000_16.5.e108694.14","doi-asserted-by":"publisher","DOI":"10.1136\/thorax.58.5.377"},{"key":"2026051906001130000_16.5.e108694.15","doi-asserted-by":"publisher","DOI":"10.3949\/ccjm.87a.19067"},{"key":"2026051906001130000_16.5.e108694.16","doi-asserted-by":"publisher","DOI":"10.1164\/rccm.201908-1581ST"},{"key":"2026051906001130000_16.5.e108694.17","doi-asserted-by":"crossref","first-page":"1","DOI":"10.15172\/pneu.2014.5\/464","article-title":"The diagnosis of pneumonia requires a chest radiograph (x-ray)\u2013yes, no or sometimes?","volume":"5","author":"Wootton","year":"2014","journal-title":"Pneumonia (Nathan)"},{"key":"2026051906001130000_16.5.e108694.18","first-page":"105","article-title":"STEPS: lean thinking, theory of constraints and identifying bottlenecks in an emergency department","volume":"106","author":"Ryan","year":"2013","journal-title":"Ir Med J"},{"key":"2026051906001130000_16.5.e108694.19","article-title":"Automated Assessment and Tracking of COVID-19 Pulmonary Disease Severity on Chest Radiographs Using Convolutional Siamese Neural Networks","volume":"2","author":"Li","year":"2020","journal-title":"Radiology: Artificial Intelligence"},{"key":"2026051906001130000_16.5.e108694.20","doi-asserted-by":"publisher","DOI":"10.1136\/neurintsurg-2019-015135"},{"key":"2026051906001130000_16.5.e108694.21","doi-asserted-by":"crossref","DOI":"10.1136\/bmjresp-2021-001045","article-title":"Chest radiograph-based artificial intelligence predictive model for mortality in community-acquired pneumonia","volume":"8","author":"Quah","year":"2021","journal-title":"BMJ Open Respir Res"},{"key":"2026051906001130000_16.5.e108694.22","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2020.103898"},{"key":"2026051906001130000_16.5.e108694.23","doi-asserted-by":"crossref","DOI":"10.1136\/bmj-2023-078378","article-title":"TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods","volume":"385","author":"Collins","year":"2024","journal-title":"BMJ"},{"key":"2026051906001130000_16.5.e108694.24","unstructured":"MIRC CTP - MircWiki, Available: https:\/\/mircwiki.rsna.org\/index.php?title=MIRC_CTP"},{"key":"2026051906001130000_16.5.e108694.25","doi-asserted-by":"crossref","unstructured":"Szegedy C , Liu W , Jia Y , et al . Going deeper with convolutions. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); Boston, MA, USA, 2015:1\u20139. doi:10.1109\/CVPR.2015.7298594","DOI":"10.1109\/CVPR.2015.7298594"},{"key":"2026051906001130000_16.5.e108694.26","doi-asserted-by":"crossref","unstructured":"Chen Y-X , Wang J-Y , Guo S-B . Use of CRB-65 and quick Sepsis-related Organ Failure Assessment to predict site of care and mortality in pneumonia patients in the emergency department: a retrospective study. Crit Care 2016;20. doi:10.1186\/s13054-016-1351-0","DOI":"10.1186\/s13054-016-1351-0"},{"key":"2026051906001130000_16.5.e108694.27","doi-asserted-by":"crossref","first-page":"927","DOI":"10.1016\/j.chest.2021.10.031","article-title":"Pneumonia Severity Index and CURB-65 Score Are Good Predictors of Mortality in Hospitalized Patients With SARS-CoV-2 Community-Acquired Pneumonia","volume":"161","author":"Bradley","year":"2022","journal-title":"Chest"},{"key":"2026051906001130000_16.5.e108694.28","doi-asserted-by":"crossref","first-page":"7294","DOI":"10.1007\/s00330-023-09664-3","article-title":"The yield of chest X-ray or ultra-low-dose chest-CT in emergency department patients suspected of pulmonary infection without respiratory symptoms or signs","volume":"33","author":"van den Berk","year":"2023","journal-title":"Eur Radiol"},{"key":"2026051906001130000_16.5.e108694.29","doi-asserted-by":"crossref","first-page":"271","DOI":"10.1016\/j.patrec.2019.11.013","article-title":"Deep-learning framework to detect lung abnormality \u2013 A study with chest X-Ray and lung CT scan images","volume":"129","author":"Bhandary","year":"2020","journal-title":"Pattern Recognit Lett"},{"key":"2026051906001130000_16.5.e108694.30","doi-asserted-by":"crossref","first-page":"1770","DOI":"10.1007\/s00330-020-07269-8","article-title":"Initial chest radiographs and artificial intelligence (AI) predict clinical outcomes in COVID-19 patients: analysis of 697 Italian patients","volume":"31","author":"Mushtaq","year":"2021","journal-title":"Eur Radiol"},{"key":"2026051906001130000_16.5.e108694.31","doi-asserted-by":"publisher","DOI":"10.1186\/s12879-022-07289-3"},{"key":"2026051906001130000_16.5.e108694.32","doi-asserted-by":"crossref","DOI":"10.1117\/1.JMI.10.4.044504","article-title":"Predicting intensive care need for COVID-19 patients using deep learning on chest radiography","volume":"10","author":"Li","year":"2023","journal-title":"J Med Imag"},{"key":"2026051906001130000_16.5.e108694.33","doi-asserted-by":"crossref","first-page":"e286","DOI":"10.1016\/S2589-7500(21)00039-X","article-title":"Prognostication of patients with COVID-19 using artificial intelligence based on chest x-rays and clinical data: a retrospective study","volume":"3","author":"Jiao","year":"2021","journal-title":"Lancet Digit Health"},{"key":"2026051906001130000_16.5.e108694.34","doi-asserted-by":"crossref","first-page":"249","DOI":"10.1177\/08465371211023861","article-title":"Impact of 24\/7 Onsite Emergency Radiology Staff Coverage on Emergency Department Workflow","volume":"73","author":"Macri","year":"2022","journal-title":"Canad Associat Radiol J"},{"key":"2026051906001130000_16.5.e108694.35","doi-asserted-by":"publisher","DOI":"10.1016\/j.annemergmed.2024.01.031"},{"key":"2026051906001130000_16.5.e108694.36","unstructured":"Toronto\u2019s Demographics, Economy & Labour Force \u2013 City of Toronto, Available: https:\/\/www.toronto.ca\/city-government\/data-research-maps\/toronto-economy-labour-force-demographics\/"},{"key":"2026051906001130000_16.5.e108694.37","doi-asserted-by":"publisher","DOI":"10.2307\/3001968"},{"key":"2026051906001130000_16.5.e108694.38","doi-asserted-by":"publisher","DOI":"10.1080\/14786440009463897"}],"container-title":["BMJ Open"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bmjopen-2025-108694","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,19]],"date-time":"2026-05-19T13:00:24Z","timestamp":1779195624000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmjopen.bmj.com\/lookup\/doi\/10.1136\/bmjopen-2025-108694"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5]]},"references-count":38,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2026,5,19]]},"published-print":{"date-parts":[[2026,5]]}},"alternative-id":["10.1136\/bmjopen-2025-108694"],"URL":"https:\/\/doi.org\/10.1136\/bmjopen-2025-108694","relation":{},"ISSN":["2044-6055","2044-6055"],"issn-type":[{"value":"2044-6055","type":"print"},{"value":"2044-6055","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5]]}}}