{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T18:36:42Z","timestamp":1788547002538,"version":"build-2803163510"},"reference-count":20,"publisher":"American Medical Association (AMA)","issue":"7","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JAMA Netw Open"],"abstract":"<jats:sec id=\"ab-zoi250690-4\">\n                    <jats:title>Importance<\/jats:title>\n                    <jats:p>Injury is a leading cause of morbidity and mortality among children worldwide. Prevention strategies rely on timely and accurate injury surveillance. Many national programs, including the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP), depend on manual review of emergency department (ED) medical records to track injury trends. Rising ED volumes have strained manual processes, delaying data analysis and compromising early detection of public health risks.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"ab-zoi250690-5\">\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>To evaluate whether natural language processing (NLP) transformer models can automate injury case detection in ED medical records, improving CHIRPP workflow efficiency while maintaining high sensitivity.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"ab-zoi250690-6\">\n                    <jats:title>Design, Setting, and Participants<\/jats:title>\n                    <jats:p>Prognostic study of ED visits from January 1, 2017, to December 31, 2023, at The Hospital for Sick Children, a high-volume tertiary pediatric referral center in Toronto, Canada, and a core CHIRPP site. The dataset included pediatric ED visits across all age groups. All medical records were labeled as requiring or not requiring CHIRPP reporting, with no exclusions. Two transformer-based NLP models, DistilBERT-base-uncased (model 1) and BERT-large-uncased (model 2), were fine tuned using supervised learning to classify medical records as CHIRPP-reportable or not.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"ab-zoi250690-7\">\n                    <jats:title>Exposure<\/jats:title>\n                    <jats:p>Application of fine-tuned NLP transformer models to routine ED visit data to automate classification of injury-related cases for surveillance reporting.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"ab-zoi250690-8\">\n                    <jats:title>Main Outcomes and Measures<\/jats:title>\n                    <jats:p>Outcomes included true positive rate (TPR), true negative rate (TNR), false positive rate (FPR), false negative rate (FNR), area under the receiver operating characteristic curve (AUROC), and area under the precision-recall curve (AUPRC). Additional outcomes included the proportion of medical records requiring manual review to achieve 90% sensitivity.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"ab-zoi250690-9\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Among 217\u202f173 pediatric ED visits across all age groups, model 1 achieved an AUROC of 0.983, AUPRC of 0.932, TPR of 0.90, TNR of 0.99, FPR of 0.014, and FNR of 0.10. Model 2 showed similar performance with an AUROC of 0.983, AUPRC of 0.931, TPR of 0.90, TNR of 0.99, FPR of 0.012, and FNR of 0.09. Both models identified 90% of injury cases while reducing manual medical record review from 100% to 17% of ED visits.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"ab-zoi250690-10\">\n                    <jats:title>Conclusions and Relevance<\/jats:title>\n                    <jats:p>NLP transformer models accurately automated detecting injury cases in ED patient medical records, with the potential of enabling real-time injury surveillance monitoring.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1001\/jamanetworkopen.2025.24154","type":"journal-article","created":{"date-parts":[[2025,7,31]],"date-time":"2025-07-31T15:31:04Z","timestamp":1753975864000},"page":"e2524154","source":"Crossref","is-referenced-by-count":7,"title":["AI-Driven Injury Reporting in Pediatric Emergency Departments"],"prefix":"10.1001","volume":"8","author":[{"given":"Devin","family":"Singh","sequence":"first","affiliation":[{"name":"Division of Pediatric Emergency Medicine, Hospital for Sick Children (SickKids), Toronto, Ontario, Canada"},{"name":"Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada"},{"name":"Department of Computer Science, University of Toronto, Toronto, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alper","family":"Celik","sequence":"additional","affiliation":[{"name":"Center for Computational Medicine, SickKids, Toronto, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Evangeline W. J.","family":"Zhang","sequence":"additional","affiliation":[{"name":"Division of Pediatric Emergency Medicine, Hospital for Sick Children (SickKids), Toronto, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eric","family":"Liu","sequence":"additional","affiliation":[{"name":"Schulich School of Medicine and Dentistry, University of Western Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Daniel","family":"Rosenfield","sequence":"additional","affiliation":[{"name":"Division of Pediatric Emergency Medicine, Hospital for Sick Children (SickKids), Toronto, Ontario, Canada"},{"name":"Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"10","published-online":{"date-parts":[[2025,7,31]]},"reference":[{"issue":"4","key":"zoi250690r5","doi-asserted-by":"publisher","first-page":"227","DOI":"10.1177\/0885066606288944","article-title":"Injury admissions to pediatric intensive care are predictable and preventable: a call to action.","volume":"21","author":"Joffe","year":"2006","journal-title":"J Intensive Care Med"},{"issue":"5","key":"zoi250690r7","doi-asserted-by":"publisher","first-page":"405","DOI":"10.2310\/8000.2013.131280","article-title":"Trends in use in a Canadian pediatric emergency department.","volume":"16","author":"Doan","year":"2014","journal-title":"CJEM"},{"issue":"3","key":"zoi250690r8","doi-asserted-by":"publisher","DOI":"10.1542\/peds.2022-060971","article-title":"Crowding in the emergency department: challenges and recommendations for the care of children.","volume":"151","author":"Gross","year":"2023","journal-title":"Pediatrics"},{"key":"zoi250690r9","doi-asserted-by":"publisher","first-page":"184","DOI":"10.1016\/j.ijmedinf.2019.06.008","article-title":"Prediction of emergency department patient disposition based on natural language processing of triage notes.","volume":"129","author":"Sterling","year":"2019","journal-title":"Int J Med Inform"},{"issue":"e2","key":"zoi250690r10","doi-asserted-by":"publisher","first-page":"e212","DOI":"10.1136\/amiajnl-2013-001962","article-title":"Developing and evaluating an automated appendicitis risk stratification algorithm for pediatric patients in the emergency department.","volume":"20","author":"Deleger","year":"2013","journal-title":"J Am Med Inform Assoc"},{"issue":"12","key":"zoi250690r11","doi-asserted-by":"publisher","first-page":"2326","DOI":"10.1016\/j.jpedsurg.2021.03.007","article-title":"Using deep learning and natural language processing models to detect child physical abuse.","volume":"56","author":"Shahi","year":"2021","journal-title":"J Pediatr Surg"},{"key":"zoi250690r12","doi-asserted-by":"publisher","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":"zoi250690r18","doi-asserted-by":"publisher","DOI":"10.2196\/44467","article-title":"Pediatric injury surveillance from uncoded emergency department admission records in Italy: machine learning\u2013based text-mining approach.","volume":"9","author":"Azzolina","year":"2023","journal-title":"JMIR Public Health Surveill"},{"issue":"3","key":"zoi250690r19","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2022.2599","article-title":"Assessment of machine learning\u2013based medical directives to expedite care in pediatric emergency medicine.","volume":"5","author":"Singh","year":"2022","journal-title":"JAMA Netw Open"},{"issue":"1","key":"zoi250690r20","doi-asserted-by":"publisher","first-page":"3574","DOI":"10.1038\/s41598-025-87161-x","article-title":"A pediatric emergency prediction model using natural language process in the pediatric emergency department.","volume":"15","author":"Choi","year":"2025","journal-title":"Sci Rep"},{"key":"zoi250690r2","author":"Pike","year":"2011"},{"key":"zoi250690r13","author":"Devlin","year":"2019"},{"key":"zoi250690r16","author":"Eyre","year":"2021"},{"key":"zoi250690r1","unstructured":"US Centers for Disease Control and Prevention. Child Injury. 2020. Accessed June 24, 2025. https:\/\/www.cdc.gov\/vitalsigns\/childinjury\/index.html"},{"key":"zoi250690r3","unstructured":"Commission? UCPS. National Electronic Injury Surveillance System. 2024. Accessed June 24, 2025. https:\/\/www.cpsc.gov\/Research--Statistics\/NEISS-Injury-Data"},{"key":"zoi250690r4","unstructured":"Swansea University. All Wales Injury Surveillance System. Accessed June 24, 2025. https:\/\/popdatasci.swan.ac.uk\/portfolio\/all-wales-injury-surveillance-system-awiss\/"},{"key":"zoi250690r6","unstructured":"Health Quality Ontario. Under Pressure: Emergency Department Performance in Ontario. 2016. Accessed June 24, 2025. https:\/\/www.hqontario.ca\/portals\/0\/Documents\/system-performance\/under-pressure-report-en.pdf"},{"key":"zoi250690r14","unstructured":"Sanh? V, Debut? L, Chaumond? J, Wolf? T. DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter.? arXiv. Preprint posted online October 2, 2019. doi:10.48550\/arXiv.1910.01108"},{"key":"zoi250690r15","unstructured":"Honnibal? M, Montani? I. spaCy 2: Natural language understanding with Bloom embeddings, convolutional neural networks and incremental parsing. 2017. https:\/\/sentometrics-research.com\/publication\/72\/"},{"key":"zoi250690r17","doi-asserted-by":"crossref","unstructured":"Wolf? T, Debut? L, Sanh? V, . Transformers: state-of-the-art natural language processing. Conference on Empirical Methods in Natural Language Processing, Proceedings of Systems Demonstrations. 2020. Accessed June 24, 2025. https:\/\/aclanthology.org\/2020.emnlp-demos.6\/","DOI":"10.18653\/v1\/2020.emnlp-demos.6"}],"container-title":["JAMA Network Open"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/jamanetwork.com\/journals\/jamanetworkopen\/articlepdf\/2837053\/singh_2025_oi_250690_1753367761.85452.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,7,31]],"date-time":"2025-07-31T15:31:05Z","timestamp":1753975865000},"score":1,"resource":{"primary":{"URL":"https:\/\/jamanetwork.com\/journals\/jamanetworkopen\/fullarticle\/2837053"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7,31]]},"references-count":20,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2025,7,1]]}},"URL":"https:\/\/doi.org\/10.1001\/jamanetworkopen.2025.24154","relation":{},"ISSN":["2574-3805"],"issn-type":[{"value":"2574-3805","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,7,31]]}}}