{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T16:06:36Z","timestamp":1789401996177,"version":"build-2803163510"},"reference-count":28,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Med Inform"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Neonatal sepsis remains a major cause of neonatal morbidity and mortality in low- and middle-income countries (LMICs). Early diagnosis is challenging because of nonspecific clinical manifestations and delays in laboratory confirmation. Machine learning (ML) approaches using structured electronic health record (EHR) data may improve early risk stratification in neonatal intensive care units (NICUs).<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study aimed to evaluate ML models for predicting culture-confirmed neonatal sepsis among neonates admitted to a tertiary NICU in Jordan, with the objective of addressing diagnostic gaps in resource-limited settings. Specifically, we aimed to identify key predictors through feature importance analysis, evaluate model performance with class-imbalanced data, and propose strategies to improve interpretability and generalizability in LMICs.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n A retrospective cohort study was conducted using structured EHRs of 3274 neonates admitted to a tertiary NICU in Jordan between 2018 and 2024. Neonates who underwent blood culture testing were included. The dataset was divided into training (n=2619, 80%) and testing (n=655, 20%) subsets using stratified sampling. Three ML models\u2014Extreme Gradient Boosting (XGBoost), decision trees, and neural networks\u2014were trained using clinical, laboratory, and demographic variables. Class imbalance was addressed using the synthetic minority oversampling technique (SMOTE) applied to the training dataset. Model performance was evaluated using accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC).<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n Among 3274 neonates included in the study, the XGBoost model demonstrated the best predictive performance on the independent test set (n=655, 20%), achieving an accuracy of 94% (616\/655 correct predictions, 95% CI 92% to 96%), sensitivity of 98% (95\/97 sepsis cases correctly identified, 95% CI 96% to 99%), and an AUC of 0.98 (95% CI 0.97 to 0.99). Decision trees provided interpretable classification rules with moderate performance, whereas neural networks showed lower discriminative ability, with an AUC of 0.81 (95% CI 0.78 to 0.84). Important predictive features included C-reactive protein, platelet count, and gestational age.<\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n XGBoost demonstrated strong predictive performance in this retrospective cohort, supporting its potential as a foundation for future prospective clinical decision support tools. External validation and prospective studies are required before clinical implementation.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/88732","type":"journal-article","created":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T15:41:16Z","timestamp":1789400476000},"page":"e88732-e88732","source":"Crossref","is-referenced-by-count":0,"title":["Machine Learning\u2013Based Prediction of Culture-Confirmed Neonatal Sepsis in a Tertiary Neonatal Intensive Care Unit: Retrospective Cohort Study"],"prefix":"10.2196","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1124-592X","authenticated-orcid":false,"given":"Eman","family":"Badran","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-6823-7392","authenticated-orcid":false,"given":"Oraib","family":"Al-Smadi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-6526-1757","authenticated-orcid":false,"given":"Loiy T","family":"Algazo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7029-4487","authenticated-orcid":false,"given":"Alaa T","family":"Al Ghazo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8460-0748","authenticated-orcid":false,"given":"Arwa","family":"Al Anber","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-9118-2342","authenticated-orcid":false,"given":"Areej","family":"Sharaqa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2386-6443","authenticated-orcid":false,"given":"Lena","family":"Abu-Argoub","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-0867-1316","authenticated-orcid":false,"given":"Shatha","family":"Al Jaberi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1860-4294","authenticated-orcid":false,"given":"Abdulrahman","family":"Alhanbali","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-0808-3086","authenticated-orcid":false,"given":"Taimein","family":"Yacoub","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1235-3228","authenticated-orcid":false,"given":"Shahd","family":"Rihan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3337-9657","authenticated-orcid":false,"given":"Hala","family":"Al-Jaberi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,9,14]]},"reference":[{"key":"R1","unstructured":"Newborns: improving survival and well-being. World Health Organization. 2020. URL: https:\/\/www.who.int\/westernpacific\/newsroom\/fact-sheets\/detail\/newborns-reducing-mortality [Accessed 04-07-2025]"},{"issue":"8","key":"R2","doi-asserted-by":"publisher","first-page":"745","DOI":"10.1136\/archdischild-2020-320217","article-title":"Global incidence and mortality of neonatal sepsis: a systematic review and meta-analysis","volume":"106","author":"Fleischmann","journal-title":"Arch Dis Child"},{"issue":"6","key":"R3","doi-asserted-by":"publisher","first-page":"1371","DOI":"10.12669\/pjms.36.6.2281","article-title":"Admission patterns and risk factors linked with neonatal mortality: a hospital-based retrospective study","volume":"36","author":"Al-Momani","journal-title":"Pak J Med Sci"},{"issue":"5","key":"R4","doi-asserted-by":"publisher","first-page":"1141","DOI":"10.1038\/s41390-021-01749-3","article-title":"Neonatal sepsis definitions from randomised clinical trials","volume":"93","author":"Hayes","journal-title":"Pediatr Res"},{"issue":"1","key":"R5","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1093\/tropej\/fmu079","article-title":"Challenges in the diagnosis and management of neonatal sepsis","volume":"61","author":"Zea-Vera","journal-title":"J Trop Pediatr"},{"issue":"6","key":"R6","doi-asserted-by":"publisher","first-page":"1120","DOI":"10.1016\/j.jpeds.2012.11.089","article-title":"Late-onset sepsis in very low birth weight infants from singleton and multiple-gestation births","volume":"162","author":"Boghossian","journal-title":"J Pediatr"},{"issue":"11","key":"R7","doi-asserted-by":"publisher","first-page":"3482","DOI":"10.1128\/JCM.02107-08","article-title":"Effects of volume and site of blood draw on blood culture results","volume":"47","author":"Gonsalves","journal-title":"J Clin Microbiol"},{"issue":"12","key":"R8","doi-asserted-by":"publisher","first-page":"1677","DOI":"10.1016\/S1473-3099(21)00050-5","article-title":"Effects of antibiotic resistance, drug target attainment, bacterial pathogenicity and virulence, and antibiotic access and affordability on outcomes in neonatal sepsis: an international microbiology and drug evaluation prospective substudy (BARNARDS)","volume":"21","author":"Thomson","journal-title":"Lancet Infect Dis"},{"issue":"2","key":"R9","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0212665","article-title":"Machine learning models for early sepsis recognition in the neonatal intensive care unit using readily available electronic health record data","volume":"14","author":"Masino","journal-title":"PLoS One"},{"issue":"6","key":"R10","doi-asserted-by":"publisher","first-page":"739","DOI":"10.1016\/j.jelectrocard.2017.08.013","article-title":"Early sepsis detection in critical care patients using multiscale blood pressure and heart rate dynamics","volume":"50","author":"Shashikumar","journal-title":"J Electrocardiol"},{"issue":"12","key":"R11","doi-asserted-by":"publisher","first-page":"3201","DOI":"10.1111\/apa.16083","article-title":"Neonatal sepsis prediction through clinical decision support algorithms: a systematic review","volume":"110","author":"Persad","journal-title":"Acta Paediatr"},{"issue":"6","key":"R12","doi-asserted-by":"publisher","first-page":"e481","DOI":"10.1097\/CCM.0000000000003011","article-title":"Development and evaluation of an automated machine learning algorithm for in-hospital mortality risk adjustment among critical care patients","volume":"46","author":"Delahanty","journal-title":"Crit Care Med"},{"issue":"4","key":"R13","doi-asserted-by":"publisher","first-page":"547","DOI":"10.1097\/CCM.0000000000002936","article-title":"An interpretable machine learning model for accurate prediction of sepsis in the ICU","volume":"46","author":"Nemati","journal-title":"Crit Care Med"},{"issue":"17","key":"R14","doi-asserted-by":"publisher","DOI":"10.3390\/jcm12175658","article-title":"Machine learning-based early prediction of sepsis using electronic health records: a systematic review","volume":"12","author":"Islam","journal-title":"J Clin Med"},{"issue":"19","key":"R15","doi-asserted-by":"publisher","DOI":"10.3390\/jcm13195959","article-title":"AI algorithms for modeling the risk, progression, and treatment of sepsis, including early-onset sepsis-a systematic review","volume":"13","author":"T\u0105del","journal-title":"J Clin Med"},{"issue":"8","key":"R16","doi-asserted-by":"publisher","first-page":"695","DOI":"10.3390\/jpm11080695","article-title":"Machine learning approaches to predict in-hospital mortality among neonates with clinically suspected sepsis in the neonatal intensive care unit","volume":"11","author":"Hsu","journal-title":"J Pers Med"},{"issue":"2","key":"R17","doi-asserted-by":"publisher","first-page":"326","DOI":"10.1136\/amiajnl-2013-001854","article-title":"Medical decision support using machine learning for early detection of late-onset neonatal sepsis","volume":"21","author":"Mani","journal-title":"J Am Med Inform Assoc"},{"key":"R18","doi-asserted-by":"publisher","DOI":"10.1016\/j.clinsp.2022.100148","article-title":"Complete blood count and C-reactive protein to predict positive blood culture among neonates using machine learning algorithms","volume":"78","author":"Matsushita","journal-title":"Clinics (Sao Paulo)"},{"key":"R19","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2023.107156","article-title":"Development and clinical impact assessment of a machine-learning model for early prediction of late-onset sepsis","volume":"163","author":"van den Berg","journal-title":"Comput Biol Med"},{"issue":"2","key":"R20","doi-asserted-by":"publisher","first-page":"350","DOI":"10.1038\/s41390-022-02274-7","article-title":"Artificial and human intelligence for early identification of neonatal sepsis","volume":"93","author":"Sullivan","journal-title":"Pediatr Res"},{"key":"R21","doi-asserted-by":"publisher","DOI":"10.1155\/2023\/3567194","article-title":"Neonatal disease prediction using machine learning techniques","volume":"2023","author":"Robi","journal-title":"J Healthc Eng"},{"key":"R22","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpeds.2023.113869","article-title":"Clinical decision support for improved neonatal care: the development of a machine learning model for the prediction of late-onset sepsis and necrotizing enterocolitis","volume":"266","author":"Meeus","journal-title":"J Pediatr"},{"issue":"1","key":"R23","doi-asserted-by":"publisher","DOI":"10.1136\/bmjpo-2025-003561","article-title":"A machine learning model for prediction of early-onset neonatal sepsis in low-income and middle-income countries: development and validation study","volume":"10","author":"Kainth","journal-title":"BMJ Paediatr Open"},{"key":"R24","doi-asserted-by":"publisher","DOI":"10.1016\/j.ibmed.2025.100198","article-title":"A neonatal sepsis prediction algorithm using electronic medical record data from Mbarara Regional Referral Hospital","volume":"11","author":"Ezeobi Dennis","journal-title":"Intell Based Med"},{"issue":"1","key":"R25","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-025-30572-7","article-title":"Risk prediction model for neonatal mortality among neonates hospitalized with sepsis, Bahir Dar, Ethiopia","volume":"16","author":"Geremew","journal-title":"Sci Rep"},{"issue":"3","key":"R26","doi-asserted-by":"publisher","first-page":"160","DOI":"10.1007\/s12519-021-00505-1","article-title":"Prediction modelling in the early detection of neonatal sepsis","volume":"18","author":"Sahu","journal-title":"World J Pediatr"},{"issue":"1","key":"R27","doi-asserted-by":"publisher","first-page":"101687","DOI":"10.1016\/j.siny.2025.101687","article-title":"Artificial intelligence in neonatal sepsis: scope, challenges, and potential solutions!","volume":"31","author":"Kainth","journal-title":"Semin Fetal Neonatal Med"},{"issue":"1","key":"R28","doi-asserted-by":"publisher","DOI":"10.1186\/s12879-023-08409-3","article-title":"Machine learning applications on neonatal sepsis treatment: a scoping review","volume":"23","author":"O\u2019Sullivan","journal-title":"BMC Infect Dis"}],"container-title":["JMIR Medical Informatics"],"original-title":[],"language":"en","deposited":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T15:41:19Z","timestamp":1789400479000},"score":1,"resource":{"primary":{"URL":"https:\/\/medinform.jmir.org\/2026\/1\/e88732"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,14]]},"references-count":28,"URL":"https:\/\/doi.org\/10.2196\/88732","relation":{},"ISSN":["2291-9694"],"issn-type":[{"value":"2291-9694","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,14]]},"article-number":"v14i8e88732"}}