{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T20:53:36Z","timestamp":1792788816566,"version":"4.5.0"},"reference-count":55,"publisher":"Public Library of Science (PLoS)","issue":"4","license":[{"start":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T00:00:00Z","timestamp":1775692800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100018336","name":"National Institute for Health Research Health Protection Research Unit","doi-asserted-by":"publisher","award":["R56LM013784"],"award-info":[{"award-number":["R56LM013784"]}],"id":[{"id":"10.13039\/100018336","id-type":"DOI","asserted-by":"publisher"}]},{"name":"NIH\/NLM","award":["R01LM014407"],"award-info":[{"award-number":["R01LM014407"]}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>A significant proportion of intensive care unit (ICU) patients undergo surgical procedures, and some may develop postoperative infections. Accurately predicting postoperative infection risk and identifying key contributing factors is crucial for improving postoperative management and understanding infection mechanisms. However, this task is challenging due to the complex interplay of multiple risk factors. While machine learning models can model these intricate associations to predict postoperative infection risk, their lack of interpretability \u2013 failing to uncover each factor\u2019s impact\u2014hinders their adoption in clinical settings. To address this difficulty, we introduced an interpretable deep neural network (DNN) model that integrates a permutation feature importance test (PermFIT). PermFIT rigorously evaluates the impact of each feature on postoperative infection risk through a rigorous statistical inference. By using only the identified important features as inputs, the DNN\u2019s predictive performance can be further enhanced. We conducted an extensive study using electronic health records (EHRs) from the Medical Information Mart for Intensive Care (MIMIC-III), a large-scale ICU EHR database. Under the PermFIT framework, our DNN model effectively identifies significant factors associated with postoperative infections while delivering the most accurate postoperative infection risk predictions. These findings highlight the clinical utility of our proposed DNN framework in managing postoperative care for ICU surgical patients, ultimately improving their health outcomes.<\/jats:p>","DOI":"10.1371\/journal.pone.0346896","type":"journal-article","created":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T17:40:14Z","timestamp":1775756414000},"page":"e0346896","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":1,"title":["An interpretable deep learning framework for predictive modeling of postoperative infections in ICU patients"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0006-2186-3683","authenticated-orcid":true,"given":"Xiaoyu","family":"Wu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Isaac","family":"Luria","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Meisheng","family":"Xiao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Patrick","family":"Tighe","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fei","family":"Zou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7879-9460","authenticated-orcid":true,"given":"Baiming","family":"Zou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,4,9]]},"reference":[{"issue":"5","key":"pone.0346896.ref001","doi-asserted-by":"crossref","first-page":"427","DOI":"10.1007\/s00595-019-01827-4","article-title":"Clinical- and surgery-specific risk factors for post-operative sepsis: a systematic review and meta-analysis of over 30 million patients","volume":"50","author":"P Plaeke","year":"2020","journal-title":"Surg Today"},{"issue":"10","key":"pone.0346896.ref002","article-title":"Postoperative Staphylococcus aureus Infections in Patients With and Without Preoperative Colonization","volume":"6","author":"DPR Troeman","year":"2023","journal-title":"JAMA Netw Open"},{"issue":"6","key":"pone.0346896.ref003","doi-asserted-by":"crossref","first-page":"1122","DOI":"10.1097\/ALN.0000000000004017","article-title":"Prevention of Healthcare-associated Infections in Intensive Care Unit Patients","volume":"135","author":"M Mazzeffi","year":"2021","journal-title":"Anesthesiology"},{"issue":"22","key":"pone.0346896.ref004","doi-asserted-by":"crossref","first-page":"2039","DOI":"10.1001\/jamainternmed.2013.9763","article-title":"Health care-associated infections: a meta-analysis of costs and financial impact on the US health care system","volume":"173","author":"E Zimlichman","year":"2013","journal-title":"JAMA Intern Med"},{"issue":"5","key":"pone.0346896.ref005","doi-asserted-by":"crossref","first-page":"483","DOI":"10.1001\/jama.2014.18614","article-title":"Underlying reasons associated with hospital readmission following surgery in the United States","volume":"313","author":"RP Merkow","year":"2015","journal-title":"JAMA"},{"issue":"9","key":"pone.0346896.ref006","doi-asserted-by":"crossref","first-page":"858","DOI":"10.1001\/archsurg.2010.179","article-title":"Postoperative hyperglycemia and surgical site infection in general surgery patients","volume":"145","author":"A Ata","year":"2010","journal-title":"Arch Surg"},{"key":"pone.0346896.ref007","first-page":"87","article-title":"Reducing Surgical Site Infections","author":"KMF Itani","year":"2024"},{"key":"pone.0346896.ref008","doi-asserted-by":"crossref","first-page":"2365","DOI":"10.1007\/s00268-022-06638-2","article-title":"Associations Between Preoperative Risk, Postoperative Complications, and 30-Day Mortality","volume":"46","author":"HJ Madsen","year":"2022","journal-title":"World J Surg"},{"issue":"1","key":"pone.0346896.ref009","doi-asserted-by":"crossref","first-page":"59","DOI":"10.1016\/j.jamcollsurg.2016.10.029","article-title":"American College of Surgeons and Surgical Infection Society: Surgical Site Infection Guidelines, 2016 Update","volume":"224","author":"KA Ban","year":"2017","journal-title":"J Am Coll Surg"},{"issue":"12","key":"pone.0346896.ref010","article-title":"New WHO recommendations on preoperative measures for surgical site infection prevention: an evidence-based global perspective","volume":"16","author":"B Allegranzi","year":"2016","journal-title":"Lancet Infect Dis"},{"issue":"1","key":"pone.0346896.ref011","doi-asserted-by":"crossref","first-page":"320","DOI":"10.1186\/s12893-023-02231-z","article-title":"Risk factors analysis of surgical site infections in postoperative colorectal cancer: a nine-year retrospective study","volume":"23","author":"C Han","year":"2023","journal-title":"BMC Surg"},{"key":"pone.0346896.ref012","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1016\/S0195-6701(08)60017-1","article-title":"Surgical site infections: epidemiology, microbiology and prevention","volume":"70","author":"C Owens","year":"2008","journal-title":"J Hosp Infect"},{"issue":"3","key":"pone.0346896.ref013","doi-asserted-by":"crossref","DOI":"10.1016\/0002-9343(91)90361-Z","article-title":"Surgical wound infection rates by wound class, operative procedure, and patient risk index","volume":"91","author":"DH Culver","year":"1991","journal-title":"Am J Med"},{"issue":"1","key":"pone.0346896.ref014","first-page":"9","article-title":"Deep Learning-based Risk Prediction Model for Postoperative Healthcare-associated Infections","volume":"44","author":"C Sun","year":"2022","journal-title":"Zhongguo Yi Xue Ke Xue Yuan Xue Bao"},{"key":"pone.0346896.ref015","doi-asserted-by":"crossref","first-page":"1336588","DOI":"10.3389\/fmed.2024.1336588","article-title":"Unlocking the potential of big data and AI in medicine: insights from biobanking","volume":"11","author":"K Aky\u00fcz","year":"2024","journal-title":"Front Med (Lausanne)"},{"issue":"7","key":"pone.0346896.ref016","article-title":"A Practical Approach to Predicting Surgical Site Infection Risk Among Patients Before Leaving the Operating Room","volume":"15","author":"MS Woods","year":"2023","journal-title":"Cureus"},{"issue":"1","key":"pone.0346896.ref017","doi-asserted-by":"crossref","first-page":"123","DOI":"10.1186\/s12876-023-02713-7","article-title":"Influencing factors and predictive model of postoperative infection in patients with primary hepatic carcinoma","volume":"23","author":"Y Ma","year":"2023","journal-title":"BMC Gastroenterol"},{"issue":"15","key":"pone.0346896.ref018","first-page":"11189","article-title":"Precision Opioid Prescription in ICU Surgery: Insights from an Interpretable Deep Learning Framework","volume":"9","author":"X Zhu","year":"2024","journal-title":"J Surg (Lisle)"},{"issue":"4","key":"pone.0346896.ref019","article-title":"Machine learning application for prediction of surgical site infection after posterior cervical surgery","volume":"21","author":"K Lu","year":"2024","journal-title":"Int Wound J"},{"issue":"5","key":"pone.0346896.ref020","doi-asserted-by":"crossref","first-page":"544","DOI":"10.1016\/j.ajic.2022.08.013","article-title":"Machine-learning models for predicting surgical site infections using patient pre-operative risk and surgical procedure factors","volume":"51","author":"REA Mamlook","year":"2023","journal-title":"Am J Infect Control"},{"issue":"5","key":"pone.0346896.ref021","doi-asserted-by":"crossref","DOI":"10.1097\/01.XCS.0000896036.40703.8a","article-title":"Development and Validation of Models for Preoperative Risk of Postoperative Infections Using Electronic Health Records Data","volume":"235","author":"K Colborn","year":"2022","journal-title":"J Am Coll Surg"},{"issue":"3","key":"pone.0346896.ref022","doi-asserted-by":"crossref","first-page":"273","DOI":"10.1023\/A:1022627411411","article-title":"Support-vector networks","volume":"20","author":"C Cortes","year":"1995","journal-title":"Mach Learn"},{"key":"pone.0346896.ref023","volume-title":"Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond","author":"B Sch\u00f6lkopf","year":"2018"},{"issue":"5","key":"pone.0346896.ref024","doi-asserted-by":"crossref","first-page":"1189","DOI":"10.1214\/aos\/1013203451","article-title":"Greedy function approximation: A gradient boosting machine","volume":"29","author":"JH Friedman","year":"2001","journal-title":"Ann Statist"},{"key":"pone.0346896.ref025","doi-asserted-by":"crossref","unstructured":"Chen T, Guestrin C. XGBoost: A Scalable Tree Boosting System. arXiv:160302754. 2016.","DOI":"10.1145\/2939672.2939785"},{"issue":"1","key":"pone.0346896.ref026","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random Forests","volume":"45","author":"L Breiman","year":"2001","journal-title":"Machine Learning"},{"issue":"7","key":"pone.0346896.ref027","doi-asserted-by":"crossref","first-page":"1545","DOI":"10.1162\/neco.1997.9.7.1545","article-title":"Shape Quantization and Recognition with Randomized Trees","volume":"9","author":"Y Amit","year":"1997","journal-title":"Neural Comput"},{"issue":"7553","key":"pone.0346896.ref028","doi-asserted-by":"crossref","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"Y LeCun","year":"2015","journal-title":"Nature"},{"key":"pone.0346896.ref029","volume-title":"Deep Learning","author":"I Goodfellow","year":"2017"},{"issue":"4","key":"pone.0346896.ref030","article-title":"Can machine-learning improve cardiovascular risk prediction using routine clinical data?","volume":"12","author":"SF Weng","year":"2017","journal-title":"PLoS One"},{"issue":"3","key":"pone.0346896.ref031","article-title":"A deep learning semiparametric regression for adjusting complex confounding structures","volume":"15","author":"X Mi","year":"2021","journal-title":"Ann Appl Stat"},{"key":"pone.0346896.ref032","unstructured":"Holzinger A, Biemann C, Pattichis CS, Kell DB. What do we need to build explainable AI systems for the medical domain? arXiv preprint arXiv:171209923. 2017."},{"key":"pone.0346896.ref033","first-page":"1","article-title":"Explainability for artificial intelligence in healthcare: a multidisciplinary perspective","volume":"20","author":"J Amann","year":"2020","journal-title":"BMC Med Inf Dec Mak"},{"issue":"5","key":"pone.0346896.ref034","doi-asserted-by":"crossref","DOI":"10.1002\/widm.1379","article-title":"Interpretability of machine learning\u2010based prediction models in healthcare","volume":"10","author":"G Stiglic","year":"2020","journal-title":"WIREs Data Min Knowl"},{"key":"pone.0346896.ref035","unstructured":"Zhou Y, Kantarcioglu M. On transparency of machine learning models: A position paper. In: AI for Social Good Workshop. 2020."},{"issue":"1","key":"pone.0346896.ref036","doi-asserted-by":"crossref","first-page":"3008","DOI":"10.1038\/s41467-021-22756-2","article-title":"Permutation-based identification of important biomarkers for complex diseases via machine learning models","volume":"12","author":"X Mi","year":"2021","journal-title":"Nat Commun"},{"key":"pone.0346896.ref037","doi-asserted-by":"crossref","first-page":"1213411","DOI":"10.3389\/fmed.2023.1213411","article-title":"Machine learning-based clinical decision support for infection risk prediction","volume":"10","author":"T Feng","year":"2023","journal-title":"Front Med (Lausanne)"},{"issue":"3","key":"pone.0346896.ref038","doi-asserted-by":"crossref","first-page":"458","DOI":"10.1097\/DCR.0000000000002559","article-title":"Improved Prediction of Surgical-Site Infection After Colorectal Surgery Using Machine Learning","volume":"66","author":"KA Chen","year":"2023","journal-title":"Dis Colon Rectum"},{"issue":"6","key":"pone.0346896.ref039","doi-asserted-by":"crossref","first-page":"140","DOI":"10.11648\/j.cmr.20130206.15","article-title":"Associated risk factors for postoperative nosocomial infections among patients admitted at Felege Hiwot Referral Hospital, Bahir Dar, Northwest Ethiopia","volume":"2","author":"W Mulu","year":"2013","journal-title":"Clin Med Res"},{"key":"pone.0346896.ref040","doi-asserted-by":"crossref","first-page":"1050698","DOI":"10.3389\/fcvm.2023.1050698","article-title":"Identification of risk factors for infection after mitral valve surgery through machine learning approaches","volume":"10","author":"N Zhang","year":"2023","journal-title":"Front Cardiovasc Med"},{"key":"pone.0346896.ref041","article-title":"MIMIC-III Clinical Database (version 1.4)","author":"AEW Johnson","year":"2016"},{"key":"pone.0346896.ref042","unstructured":"Centers for Disease Control and Prevention. ICD-9-CM Official Guidelines for Coding and Reporting. 2024. https:\/\/www.cdc.gov\/nchs\/icd\/icd9cm_addenda_guidelines.htm"},{"key":"pone.0346896.ref043","doi-asserted-by":"crossref","first-page":"321","DOI":"10.1613\/jair.953","article-title":"SMOTE: Synthetic Minority Over-sampling Technique","volume":"16","author":"NV Chawla","year":"2002","journal-title":"JAIR"},{"key":"pone.0346896.ref044","doi-asserted-by":"crossref","DOI":"10.1016\/0002-9343(91)90362-2","article-title":"Risk factors for postoperative infection","volume":"91","author":"RA Garibaldi","year":"1991","journal-title":"Am J Med"},{"issue":"6","key":"pone.0346896.ref045","doi-asserted-by":"crossref","first-page":"347","DOI":"10.1086\/501911","article-title":"Factors associated with postoperative infection","volume":"22","author":"JD Scott","year":"2001","journal-title":"Infect Control Hosp Epidemiol"},{"issue":"10","key":"pone.0346896.ref046","doi-asserted-by":"crossref","first-page":"963","DOI":"10.1016\/j.ajic.2012.01.012","article-title":"Epidemiology and outcome of major postoperative infections following cardiac surgery: risk factors and impact of pathogen type","volume":"40","author":"LF Chen","year":"2012","journal-title":"Am J Infect Control"},{"issue":"7","key":"pone.0346896.ref047","doi-asserted-by":"crossref","first-page":"1056","DOI":"10.1086\/428626","article-title":"The effect of increasing age on the risk of surgical site infection","volume":"191","author":"KS Kaye","year":"2005","journal-title":"J Infect Dis"},{"issue":"4","key":"pone.0346896.ref048","doi-asserted-by":"crossref","first-page":"331","DOI":"10.1177\/000313481608200418","article-title":"Obesity and the Risk for Surgical Site Infection in Abdominal Surgery","volume":"82","author":"RD Winfield","year":"2016","journal-title":"Am Surg"},{"issue":"3","key":"pone.0346896.ref049","doi-asserted-by":"crossref","DOI":"10.1093\/ofid\/ofz083","article-title":"History of Depression and Increased Risk of Sternal Wound Infection After Cardiothoracic Surgery: A Novel and Potentially Modifiable Risk Factor","volume":"6","author":"DA Theodore","year":"2019","journal-title":"Open Forum Infect Dis"},{"issue":"1","key":"pone.0346896.ref050","doi-asserted-by":"crossref","first-page":"837","DOI":"10.1186\/s12879-020-05567-6","article-title":"Incidence and risk factors of surgical site infection following colorectal surgery in China: a national cross-sectional study","volume":"20","author":"X Zhang","year":"2020","journal-title":"BMC Infect Dis"},{"issue":"13","key":"pone.0346896.ref051","doi-asserted-by":"crossref","first-page":"1216","DOI":"10.1056\/NEJMp1606181","article-title":"Predicting the Future - Big Data, Machine Learning, and Clinical Medicine","volume":"375","author":"Z Obermeyer","year":"2016","journal-title":"N Engl J Med"},{"issue":"4","key":"pone.0346896.ref052","doi-asserted-by":"crossref","first-page":"373","DOI":"10.1001\/archsurg.2012.5","article-title":"Wound healing and infection in surgery. The clinical impact of smoking and smoking cessation: a systematic review and meta-analysis","volume":"147","author":"LT S\u00f8rensen","year":"2012","journal-title":"Arch Surg"},{"issue":"7","key":"pone.0346896.ref053","doi-asserted-by":"crossref","first-page":"2218","DOI":"10.1016\/j.arth.2018.02.069","article-title":"Is There an Association Between Smoking Status and Prosthetic Joint Infection After Primary Total Joint Arthroplasty?","volume":"33","author":"AI Gonzalez","year":"2018","journal-title":"J Arthroplasty"},{"issue":"4","key":"pone.0346896.ref054","first-page":"1086","article-title":"Nutritional risk and status of surgical patients; the relevance of nutrition training of medical students","volume":"27","author":"C Ferreira","year":"2012","journal-title":"Nutr Hosp"},{"issue":"4","key":"pone.0346896.ref055","doi-asserted-by":"crossref","first-page":"117","DOI":"10.4103\/2231-0770.165122","article-title":"Nutritional status as a predictive marker for surgical site infection in total joint arthroplasty","volume":"5","author":"R Alfargieny","year":"2015","journal-title":"Avicenna J Med"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0346896","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T17:40:24Z","timestamp":1775756424000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0346896"}},"subtitle":[],"editor":[{"given":"Young-Seob","family":"Jeong","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,4,9]]},"references-count":55,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2026,4,9]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0346896","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,9]]}}}