{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,1,16]],"date-time":"2026-01-16T12:17:13Z","timestamp":1768565833737,"version":"3.49.0"},"posted":{"date-parts":[[2024,8,29]]},"group-title":"Infectious Diseases (except HIV\/AIDS)","reference-count":23,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2024,8,29]],"date-time":"2024-08-29T00:00:00Z","timestamp":1724889600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,4,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>BACKGROUND<\/jats:title>\n                  <jats:p>Lower respiratory tract infections (LRTIs) are a leading cause of mortality worldwide and can be difficult to diagnose in critically ill patients, as non-infectious causes of respiratory failure can present with similar clinical features.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>METHODS<\/jats:title>\n                  <jats:p>\n                    We developed a LRTI diagnostic method combining the pulmonary transcriptomic biomarker\n                    <jats:italic>FABP4<\/jats:italic>\n                    with electronic medical record (EMR) text assessment using the large language model Generative Pre-trained Transformer 4 (GPT-4). We evaluated this approach in a prospective cohort of critically ill adults with acute respiratory failure from whom tracheal aspirate\n                    <jats:italic>FABP4<\/jats:italic>\n                    expression was measured by RNA sequencing. Patients with LRTI or non-infectious conditions were identified using retrospective, multi-physician clinical adjudication. We then confirmed our findings by applying this method to an independent validation cohort of 115 adults with acute respiratory failure.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>RESULTS<\/jats:title>\n                  <jats:p>\n                    In the derivation cohort, a combined classifier incorporating\n                    <jats:italic>FABP4<\/jats:italic>\n                    expression and GPT-4\u2013 assisted EMR analysis achieved an AUC of 0.93 (\u00b10.08) and an accuracy of 84%, outperforming\n                    <jats:italic>FABP4<\/jats:italic>\n                    expression alone (AUC 0.84 \u00b1 0.11) and GPT-4\u2013based analysis alone (AUC 0.83 \u00b1 0.07). By comparison, the primary medical team\u2019s admission diagnosis had an accuracy of 72%. In the validation cohort, the combined classifier yielded an AUC of 0.98 (\u00b10.04) and an accuracy of 96%.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>CONCLUSIONS<\/jats:title>\n                  <jats:p>Integrating a host transcriptional biomarker with EMR text analysis using a large language model may offer a promising new approach to improving the diagnosis of LRTIs in critically ill adults.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Description<\/jats:title>\n                  <jats:p>We present the novel use of a host transcriptional biomarker combined with artificial intelligence analysis of electronic medical record data to diagnose lower respiratory tract infections in a derivation cohort of critically ill adults, then the validation of this approach in a second, fully independent, cohort. This approach demonstrated high diagnostic accuracy compared to a gold standard of post-hoc multi-physician adjudication.<\/jats:p>\n                <\/jats:sec>","DOI":"10.1101\/2024.08.28.24312732","type":"posted-content","created":{"date-parts":[[2024,8,29]],"date-time":"2024-08-29T11:28:36Z","timestamp":1724930916000},"source":"Crossref","is-referenced-by-count":0,"title":["Integrating a host transcriptomic biomarker with a large language model for diagnosis of lower respiratory tract infection"],"prefix":"10.64898","author":[{"given":"Hoang Van","family":"Phan","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Natasha","family":"Spottiswoode","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Emily C.","family":"Lydon","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Victoria T.","family":"Chu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Adolfo","family":"Cuesta","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Alexander D.","family":"Kazberouk","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Natalie L.","family":"Richmond","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Padmini","family":"Deosthale","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Carolyn S.","family":"Calfee","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Charles R.","family":"Langelier","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2025040501300949000_2024.08.28.24312732v2.1","unstructured":"The Top 10 Causes of Death. World Health Organization, 2021. (https:\/\/www.who.int\/news-room\/fact-sheets\/detail\/the-top-10-causes-of-death)."},{"key":"2025040501300949000_2024.08.28.24312732v2.2","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMoa1500245"},{"key":"2025040501300949000_2024.08.28.24312732v2.3","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmi.2022.12.006"},{"key":"2025040501300949000_2024.08.28.24312732v2.4","unstructured":"2022 Special Report: COVID-19 US Impact on Antimicrobial Resistance. Centers for Disease Control and Prevention, 2022. (https:\/\/stacks.cdc.gov\/view\/cdc\/1190252022)."},{"key":"2025040501300949000_2024.08.28.24312732v2.5","doi-asserted-by":"publisher","DOI":"10.1126\/scitranslmed.aad6873"},{"key":"2025040501300949000_2024.08.28.24312732v2.6","doi-asserted-by":"publisher","DOI":"10.1172\/JCI165904"},{"key":"2025040501300949000_2024.08.28.24312732v2.7","doi-asserted-by":"publisher","DOI":"10.1164\/rccm.202403-0516RL"},{"key":"2025040501300949000_2024.08.28.24312732v2.8","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.38483.478183.EB"},{"key":"2025040501300949000_2024.08.28.24312732v2.9","doi-asserted-by":"publisher","DOI":"10.1093\/cid\/cix1090"},{"key":"2025040501300949000_2024.08.28.24312732v2.10","unstructured":"OpenAI. Introducing ChatGPT. (https:\/\/openai.com\/index\/chatgpt\/)."},{"key":"2025040501300949000_2024.08.28.24312732v2.11","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.233270"},{"key":"2025040501300949000_2024.08.28.24312732v2.12","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-021-00426-3"},{"key":"2025040501300949000_2024.08.28.24312732v2.13","doi-asserted-by":"publisher","DOI":"10.1093\/cid\/ciad632"},{"key":"2025040501300949000_2024.08.28.24312732v2.14","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMc2305286"},{"key":"2025040501300949000_2024.08.28.24312732v2.15","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2024.40969"},{"key":"2025040501300949000_2024.08.28.24312732v2.16","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1809700115"},{"key":"2025040501300949000_2024.08.28.24312732v2.17","unstructured":"CDC\/NHSN Surveillance Definition for Specific Types of Infections. Centers for Disease Control and Prevention, 2021.(https:\/\/www.cdc.gov\/nhsn\/pdfs\/pscmanual\/17pscnosinfdef_current.pdf)"},{"key":"2025040501300949000_2024.08.28.24312732v2.18","doi-asserted-by":"publisher","DOI":"10.1186\/s13059-014-0550-8"},{"key":"2025040501300949000_2024.08.28.24312732v2.19","doi-asserted-by":"publisher","DOI":"10.1186\/1471-2105-12-77"},{"key":"2025040501300949000_2024.08.28.24312732v2.20","unstructured":"Wei J , Wang, X. , Schuurmans, D. , Bosma, M. , Ichter, B. , Xia, F. , Chi, E. , Le, Q.V. , Zhou, D. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. Advances in Neural Information Processing Systems 35 (NeurIPS 2022): NeurIPS Proceedings 2022."},{"key":"2025040501300949000_2024.08.28.24312732v2.21","doi-asserted-by":"publisher","DOI":"10.1007\/s11606-024-08758-y"},{"key":"2025040501300949000_2024.08.28.24312732v2.22","unstructured":"Antibiotic Resistance Threats In the United States. U.S. Department of Health and Human Services, 2019. (https:\/\/www.cdc.gov\/antimicrobial-resistance\/media\/pdfs\/2019-ar-threats-report-508.pdf)."},{"key":"2025040501300949000_2024.08.28.24312732v2.23","unstructured":"Antimicrobial stewardship interventions: a practical guide. World Health Organization, 2021. (https:\/\/www.who.int\/europe\/publications\/i\/item\/9789289056267)."}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2024.08.28.24312732","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T20:10:32Z","timestamp":1768507832000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.1101\/2024.08.28.24312732"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,8,29]]},"references-count":23,"URL":"https:\/\/doi.org\/10.1101\/2024.08.28.24312732","relation":{},"subject":[],"published":{"date-parts":[[2024,8,29]]},"subtype":"preprint"}}