{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:28:51Z","timestamp":1785536931530,"version":"3.56.0"},"reference-count":33,"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 Postoperative nausea and vomiting are common complications after anesthesia. However, vomiting represents a clinically distinct and objectively measurable endpoint.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study aimed to develop and internally validate predictive models for postoperative vomiting within 24 hours using structured perioperative data and unstructured clinical text, while introducing a structured framework that separates feature construction from interpretability using large language models (LLMs).<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n We analyzed 33,460 anesthesia records from a single center (2019\u20102022). Two temporally defined prediction tasks were constructed to reflect real-world clinical decision-making and prevent information leakage: a preoperative model using variables available before anesthesia induction, and a perioperative model using variables available up to the end of surgery. Structured data were modeled using machine learning algorithms (logistic regression, Extreme Gradient Boosting, Light Gradient Boosting Machine [LightGBM]). Unstructured clinical text was incorporated through a deterministic, concept-driven preprocessing pipeline, where LLMs were used solely for normalization (temperature=0) without feature generation, followed by rule-based concept mapping and feature encoding. Post hoc interpretability was further supported using an LLM-based Question Answering Chain module. Model performance was evaluated using receiver operating characteristic-area under the curve (AUC), precision-recall AUC, calibration metrics, and threshold-based operating characteristics. Classification thresholds were selected using the Youden J statistic, and all metrics were reported with 95% CIs derived from bootstrap resampling. Decision curve analysis was performed to assess clinical utility.<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n A total of 33,460 surgical procedures were included, of which 3607 (10.8%) experienced postoperative vomiting within 24 hours. In the preoperative task, LightGBM achieved an AUC of 0.729 (95% CI 0.706\u20100.749), compared with 0.610 (95% CI 0.588\u20100.632) for the Apfel score. In the end-of-surgery task, LightGBM achieved an AUC of 0.735 (95% CI 0.714\u20100.757). At the Youden-optimal threshold, the negative predictive value exceeded 0.95 across all models. Decision curve analysis demonstrated positive net benefit across clinically relevant threshold probabilities. Incorporating text-derived features provided modest improvements, while LLM-based explanation modules generated structured, natural-language explanations intended to enhance interpretability without substantially improving predictive performance.<\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n Machine learning models can effectively predict postoperative vomiting within 24 hours using perioperative data. The proposed framework demonstrates that LLMs can be integrated in a controlled and reproducible manner\u2014restricted to deterministic normalization and post hoc reasoning\u2014to generate natural-language explanations intended to enhance the interpretability of model predictions, without introducing information leakage or altering predictive modeling. As no formal clinician-based evaluation was conducted, this interpretability benefit cannot yet be objectively confirmed, and the generated explanations should be regarded as a useful interpretability aid to be validated in future clinician-centered studies. External, multicenter validation is required before broader clinical applicability can be assumed.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/84260","type":"journal-article","created":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T19:56:06Z","timestamp":1785527766000},"page":"e84260-e84260","source":"Crossref","is-referenced-by-count":0,"title":["Prediction of Postoperative Vomiting Within 24 Hours Using Machine Learning With Large Language Model\u2013Enhanced Interpretability: Development and Validation Study"],"prefix":"10.2196","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-0983-6205","authenticated-orcid":false,"given":"Huan-Jun","family":"Wang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0256-8068","authenticated-orcid":false,"given":"Wei-Po","family":"Lee","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9331-505X","authenticated-orcid":false,"given":"Tz-Ping","family":"Gau","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3766-6450","authenticated-orcid":false,"given":"Kuang-I","family":"Cheng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-7894-7902","authenticated-orcid":false,"given":"Cheng-Ru","family":"Wei","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,7,31]]},"reference":[{"issue":"2","key":"R1","doi-asserted-by":"publisher","first-page":"143","DOI":"10.1016\/j.clinthera.2024.11.018","article-title":"A prediction model for postoperative nausea and vomiting after laparoscopic surgery for gynecologic cancers","volume":"47","author":"Zhu","journal-title":"Clin Ther"},{"issue":"1","key":"R2","doi-asserted-by":"publisher","first-page":"48","DOI":"10.4103\/sja.SJA_401_19","article-title":"Prevalence of postoperative nausea and vomiting: a systematic review and meta-analysis","volume":"14","author":"Amirshahi","journal-title":"Saudi J Anaesth"},{"issue":"7 Suppl 1","key":"R3","doi-asserted-by":"publisher","first-page":"24S","DOI":"10.1093\/bja\/69.supplement_1.24s","article-title":"Surgical and patient factors involved in postoperative nausea and vomiting","volume":"69","author":"Lerman","journal-title":"Br J Anaesth"},{"issue":"2","key":"R4","doi-asserted-by":"publisher","first-page":"411","DOI":"10.1213\/ANE.0000000000004833","article-title":"Fourth consensus guidelines for the management of postoperative nausea and vomiting","volume":"131","author":"Gan","journal-title":"Anesth Analg"},{"issue":"6","key":"R5","doi-asserted-by":"publisher","first-page":"1884","DOI":"10.1213\/01.ANE.0000219597.16143.4D","article-title":"Risk factors for postoperative nausea and vomiting","volume":"102","author":"Gan","journal-title":"Anesth Analg"},{"issue":"2","key":"R6","doi-asserted-by":"publisher","first-page":"135","DOI":"10.1093\/bja\/70.2.135","article-title":"Logistic regression analysis of fixed patient factors for postoperative sickness: a model for risk assessment","volume":"70","author":"Palazzo","journal-title":"Br J Anaesth"},{"issue":"1","key":"R7","doi-asserted-by":"publisher","first-page":"135","DOI":"10.1186\/s12871-025-02987-2","article-title":"Predicting postoperative nausea and vomiting using machine learning: a model development and validation study","volume":"25","author":"Glebov","journal-title":"BMC Anesthesiol"},{"issue":"1","key":"R8","doi-asserted-by":"publisher","first-page":"37161041","DOI":"10.1038\/s41598-023-34505-0","article-title":"Machine learning-based causal models for predicting the response of individual patients to dexamethasone treatment as prophylactic antiemetic","volume":"13","author":"Mizuguchi","journal-title":"Sci Rep"},{"issue":"5","key":"R9","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0155705","article-title":"Application of machine learning techniques to high-dimensional clinical data to forecast postoperative complications","volume":"11","author":"Thottakkara","journal-title":"PLoS One"},{"issue":"1","key":"R10","doi-asserted-by":"publisher","first-page":"85","DOI":"10.1213\/ANE.0000000000000002","article-title":"Consensus guidelines for the management of postoperative nausea and vomiting","volume":"118","author":"Gan","journal-title":"Anesth Analg"},{"issue":"8","key":"R11","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0308755","article-title":"Machine learning-based identification of the risk factors for postoperative nausea and vomiting in adults","volume":"19","author":"Hoshijima","journal-title":"PLoS One"},{"issue":"10","key":"R12","doi-asserted-by":"publisher","DOI":"10.3390\/bioengineering10101152","article-title":"Postoperative nausea and vomiting prediction: machine learning insights from a comprehensive analysis of perioperative data","volume":"10","author":"Kim","journal-title":"Bioengineering (Basel)"},{"issue":"1","key":"R13","doi-asserted-by":"publisher","first-page":"133","DOI":"10.1186\/s12874-023-01955-z","article-title":"Predicting early postoperative PONV using multiple machine-learning- and deep-learning-algorithms","volume":"23","author":"Zhou","journal-title":"BMC Med Res Methodol"},{"issue":"8","key":"R14","doi-asserted-by":"publisher","first-page":"6679","DOI":"10.1609\/aaai.v35i8.16826","article-title":"TabNet: Attentive Interpretable Tabular Learning","volume":"35","author":"Arik","journal-title":"AAAI"},{"issue":"1","key":"R15","doi-asserted-by":"publisher","DOI":"10.1002\/aisy.202400304","article-title":"A perspective on explainable artificial intelligence methods: SHAP and LIME","volume":"7","author":"Salih","journal-title":"Adv Intell Syst"},{"key":"R16","doi-asserted-by":"publisher","DOI":"10.1016\/j.measen.2024.101305","article-title":"Enhancing healthcare with AI: sustainable AI and IoT-powered ecosystem for patient aid and interpretability analysis using SHAP","volume":"36","author":"Paneru","journal-title":"Meas: Sens"},{"issue":"1","key":"R17","doi-asserted-by":"publisher","DOI":"10.1155\/2023\/9919269","article-title":"Interpretability of clinical decision support systems based on artificial intelligence from technological and medical perspective: a systematic review","volume":"2023","author":"Xu","journal-title":"J Healthc Eng"},{"issue":"4","key":"R18","doi-asserted-by":"publisher","DOI":"10.1002\/dmrr.3801","article-title":"Multi-feature, Chinese-Western medicine-integrated prediction model for diabetic peripheral neuropathy based on machine learning and SHAP","volume":"40","author":"Jiang","journal-title":"Diabetes Metab Res Rev"},{"key":"R19","doi-asserted-by":"publisher","DOI":"10.1016\/j.amjsurg.2024.115912","article-title":"A machine learning-based prediction model for delayed clinically important postoperative nausea and vomiting in high-risk patients undergoing laparoscopic gastrointestinal surgery","volume":"237","author":"Zheng","journal-title":"Am J Surg"},{"issue":"4","key":"R20","doi-asserted-by":"publisher","first-page":"811","DOI":"10.1213\/ANE.0000000000004571","article-title":"Roadmap for transforming preoperative assessment to preoperative optimization","volume":"130","author":"Aronson","journal-title":"Anesth Analg"},{"key":"R21","doi-asserted-by":"publisher","DOI":"10.1016\/j.jclinane.2021.110540","article-title":"Virtual preoperative assessment in surgical patients: a systematic review and meta-analysis","volume":"75","author":"Zhang","journal-title":"J Clin Anesth"},{"issue":"5","key":"R22","doi-asserted-by":"publisher","first-page":"1333","DOI":"10.1016\/j.bja.2024.08.038","article-title":"Generation of preoperative anaesthetic plans by ChatGPT-4.0: a mixed-method study","volume":"134","author":"Abdel Malek","journal-title":"Br J Anaesth"},{"issue":"1","key":"R23","doi-asserted-by":"publisher","first-page":"187","DOI":"10.1038\/s41746-025-01519-z","article-title":"Retrieval augmented generation for 10 large language models and its generalizability in assessing medical fitness","volume":"8","author":"Ke","journal-title":"NPJ Digit Med"},{"issue":"3","key":"R24","doi-asserted-by":"publisher","first-page":"278","DOI":"10.7861\/clinmed.2023-0078","article-title":"Early applications of ChatGPT in medical practice, education and research","volume":"23","author":"Sedaghat","journal-title":"Clin Med (Lond)"},{"key":"R25","doi-asserted-by":"publisher","DOI":"10.3389\/frai.2023.1169595","article-title":"ChatGPT in medicine: an overview of its applications, advantages, limitations, future prospects, and ethical considerations","volume":"6","author":"Dave","journal-title":"Front Artif Intell"},{"issue":"7","key":"R26","doi-asserted-by":"publisher","first-page":"1237","DOI":"10.1093\/jamia\/ocad072","article-title":"Using AI-generated suggestions from ChatGPT to optimize clinical decision support","volume":"30","author":"Liu","journal-title":"J Am Med Inform Assoc"},{"issue":"6","key":"R27","doi-asserted-by":"publisher","first-page":"1095","DOI":"10.1016\/j.jopan.2024.01.027","article-title":"PONV Management in adult patients: evidence-based summary","volume":"39","author":"Wang","journal-title":"J Perianesth Nurs"},{"issue":"12","key":"R28","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0277957","article-title":"Machine learning for prediction of postoperative nausea and vomiting in patients with intravenous patient-controlled analgesia","volume":"17","author":"Shim","journal-title":"PLoS One"},{"key":"R29","doi-asserted-by":"publisher","DOI":"10.3389\/fdgth.2025.1403047","article-title":"Artificial intelligence in clinical decision support and the prediction of adverse events","volume":"7","author":"Oei","journal-title":"Front Digit Health"},{"issue":"1","key":"R30","doi-asserted-by":"publisher","first-page":"36","DOI":"10.1186\/s12911-025-03333-9","article-title":"Explainable extratreeclassifier model for early detection of type 2 diabetes: evidence from the PERSIAN Dena Cohort","volume":"26","author":"Ghaderzadeh","journal-title":"BMC Med Inform Decis Mak"},{"issue":"10","key":"R31","doi-asserted-by":"publisher","DOI":"10.7759\/cureus.95337","article-title":"Accuracy and reliability of artificial intelligence in surgical decision-making: a literature review","volume":"17","author":"Ruiz","journal-title":"Cureus"},{"issue":"4","key":"R32","doi-asserted-by":"publisher","first-page":"260","DOI":"10.1080\/12460125.2020.1819094","article-title":"Transparency and trust in artificial intelligence systems","volume":"29","author":"Schmidt","journal-title":"J Decis Syst"},{"issue":"2","key":"R33","doi-asserted-by":"publisher","first-page":"585","DOI":"10.1007\/s00146-020-01066-z","article-title":"Artificial intelligence and the value of transparency","volume":"36","author":"Walmsley","journal-title":"AI & Soc"}],"container-title":["JMIR Medical Informatics"],"original-title":[],"language":"en","deposited":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T19:56:07Z","timestamp":1785527767000},"score":1,"resource":{"primary":{"URL":"https:\/\/medinform.jmir.org\/2026\/1\/e84260"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,31]]},"references-count":33,"URL":"https:\/\/doi.org\/10.2196\/84260","relation":{},"ISSN":["2291-9694"],"issn-type":[{"value":"2291-9694","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,31]]},"article-number":"v14i2e84260"}}