{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,14]],"date-time":"2026-11-14T19:54:31Z","timestamp":1794686071227,"version":"4.9.1"},"reference-count":28,"publisher":"American Medical Association (AMA)","issue":"1","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JAMA Surg"],"published-print":{"date-parts":[[2024,1,1]]},"abstract":"<jats:sec id=\"ab-soi230068-4\"><jats:title>Importance<\/jats:title><jats:p>Despite the unmet need, many deceased-donor kidneys are discarded or not recovered. Inefficient allocation and prolonged ischemia time are contributing factors, and early detection of high-risk donors may reduce organ loss.<\/jats:p><\/jats:sec><jats:sec id=\"ab-soi230068-5\"><jats:title>Objective<\/jats:title><jats:p>To evaluate the feasibility of machine learning (ML) and natural language processing (NLP) classification of donors with kidneys that are used vs not used for organ transplant.<\/jats:p><\/jats:sec><jats:sec id=\"ab-soi230068-6\"><jats:title>Design, Setting, and Participants<\/jats:title><jats:p>This retrospective cohort study used donor information (structured donor characteristics and unstructured donor narratives) from the United Network for Organ Sharing (UNOS). All donor offers to a single transplant center between January 2015 and December 2020 were used to train and validate ML models to predict donors who had at least 1 kidney transplanted (at our center or another center). The donor data from 2021 were used to test each model.<\/jats:p><\/jats:sec><jats:sec id=\"ab-soi230068-7\"><jats:title>Exposures<\/jats:title><jats:p>Donor information was provided by UNOS to the transplant centers with potential transplant candidates. Each center evaluated the donor and decided within an allotted time whether to accept the kidney for organ transplant.<\/jats:p><\/jats:sec><jats:sec id=\"ab-soi230068-8\"><jats:title>Main Outcomes and Measures<\/jats:title><jats:p>Outcome metrics of the test cohort included area under the receiver operating characteristic curve (AUROC), <jats:italic>F<\/jats:italic><jats:sub>1<\/jats:sub> score, accuracy, precision, and recall of each ML classifier. Feature importance and Shapley additive explanation (SHAP) summaries were assessed for model explainability.<\/jats:p><\/jats:sec><jats:sec id=\"ab-soi230068-9\"><jats:title>Results<\/jats:title><jats:p>The training\/validation cohort included 9555 donors (median [IQR] age, 50 [36-58] years; 5571 male [58.3%]), and the test cohort included 2481 donors (median [IQR] age, 52 [40-59] years; 1496 male [60.3%]). Only 20% to 30% of potential donors had at least 1 kidney transplanted. The ML model with a single variable (Kidney Donor Profile Index) showed an AUROC of 0.69, <jats:italic>F<\/jats:italic><jats:sub>1<\/jats:sub> score of 0.42, and accuracy of 0.64. Multivariable ML models based on basic a priori structured donor data showed similar metrics (logistic regression: AUROC\u2009=\u20090.70; <jats:italic>F<\/jats:italic><jats:sub>1<\/jats:sub> score\u2009=\u20090.42; accuracy\u2009=\u20090.62; random forest classifier: AUROC\u2009=\u20090.69; <jats:italic>F<\/jats:italic><jats:sub>1<\/jats:sub> score\u2009=\u20090.42; accuracy\u2009=\u20090.64). The classic NLP model (bag-of-words model) showed its best metrics (AUROC\u2009=\u20090.60; <jats:italic>F<\/jats:italic><jats:sub>1<\/jats:sub> score\u2009=\u20090.35; accuracy\u2009=\u20090.59) by the logistic regression classifier. The advanced Bidirectional Encoder Representations From Transformers model showed comparable metrics (AUROC\u2009=\u20090.62; <jats:italic>F<\/jats:italic><jats:sub>1<\/jats:sub> score\u2009=\u20090.39; accuracy\u2009=\u20090.69) only after appending basic donor information. Feature importance and SHAP detected the variables (and words) that affected the models most.<\/jats:p><\/jats:sec><jats:sec id=\"ab-soi230068-10\"><jats:title>Conclusions and Relevance<\/jats:title><jats:p>Results of this cohort study suggest that models using ML can be applied to predict donors with high-risk kidneys not used for organ transplant, but the models still need further elaboration. The use of unstructured data is likely to expand the possibilities; further exploration of new approaches will be necessary to develop models with better predictive metrics.<\/jats:p><\/jats:sec>","DOI":"10.1001\/jamasurg.2023.4679","type":"journal-article","created":{"date-parts":[[2023,11,1]],"date-time":"2023-11-01T16:18:12Z","timestamp":1698855492000},"page":"60","source":"Crossref","is-referenced-by-count":12,"title":["Prediction of High-Risk Donors for Kidney Discard and Nonrecovery Using Structured Donor Characteristics and Unstructured Donor Narratives"],"prefix":"10.1001","volume":"159","author":[{"given":"Junichiro","family":"Sageshima","sequence":"first","affiliation":[{"name":"Department of Surgery, University of California, Davis Health, Sacramento"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter","family":"Than","sequence":"additional","affiliation":[{"name":"Department of Surgery, University of California, Davis Health, Sacramento"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Naeem","family":"Goussous","sequence":"additional","affiliation":[{"name":"Department of Surgery, University of California, Davis Health, Sacramento"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Neal","family":"Mineyev","sequence":"additional","affiliation":[{"name":"Department of Surgery, University of California, Davis Health, Sacramento"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Richard","family":"Perez","sequence":"additional","affiliation":[{"name":"Department of Surgery, University of California, Davis Health, Sacramento"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"10","reference":[{"key":"soi230068r1","doi-asserted-by":"publisher","first-page":"21","DOI":"10.1111\/ajt.16982","article-title":"OPTN\/SRTR 2020 annual data report: kidney.","volume":"22","author":"Lentine","year":"2022","journal-title":"Am J Transplant"},{"issue":"9","key":"soi230068r2","doi-asserted-by":"publisher","first-page":"1353","DOI":"10.2215\/CJN.01600222","article-title":"Implications of accumulated cold time for US kidney transplantation offer acceptance.","volume":"17","author":"Barah","year":"2022","journal-title":"Clin J Am Soc Nephrol"},{"issue":"7","key":"soi230068r3","doi-asserted-by":"publisher","first-page":"1613","DOI":"10.1111\/j.1600-6143.2010.03163.x","article-title":"Improving distribution efficiency of hard-to-place deceased donor kidneys: predicting probability of discard or delay.","volume":"10","author":"Massie","year":"2010","journal-title":"Am J Transplant"},{"issue":"7","key":"soi230068r4","doi-asserted-by":"publisher","first-page":"1690","DOI":"10.1097\/TP.0000000000001238","article-title":"Predictors of deceased donor kidney discard in the US.","volume":"101","author":"Marrero","year":"2017","journal-title":"Transplantation"},{"issue":"2","key":"soi230068r5","doi-asserted-by":"publisher","first-page":"391","DOI":"10.1111\/ajt.14449","article-title":"Kidney allograft offers: predictors of turndown and the impact of late organ acceptance on allograft survival.","volume":"18","author":"Cohen","year":"2018","journal-title":"Am J Transplant"},{"issue":"11","key":"soi230068r6","doi-asserted-by":"publisher","first-page":"2708","DOI":"10.1111\/ajt.14712","article-title":"Hard-to-place kidney offers: donor- and system-level predictors of discard.","volume":"18","author":"Narvaez","year":"2018","journal-title":"Am J Transplant"},{"issue":"12","key":"soi230068r7","doi-asserted-by":"publisher","first-page":"2646","DOI":"10.1097\/TP.0000000000003640","article-title":"Automated en masse machine learning model generation shows comparable performance as classic regression models for predicting delayed graft function in renal allografts.","volume":"105","author":"Jen","year":"2021","journal-title":"Transplantation"},{"issue":"6","key":"soi230068r8","doi-asserted-by":"publisher","first-page":"633","DOI":"10.1001\/jamacardio.2021.0122","article-title":"Use of machine learning models to predict death after acute myocardial infarction.","volume":"6","author":"Khera","year":"2021","journal-title":"JAMA Cardiol"},{"issue":"12","key":"soi230068r9","doi-asserted-by":"publisher","first-page":"2821","DOI":"10.1111\/ajt.17192","article-title":"Machine learning-supported interpretation of kidney graft elementary lesions in combination with clinical data.","volume":"22","author":"Labriffe","year":"2022","journal-title":"Am J Transplant"},{"issue":"1","key":"soi230068r10","doi-asserted-by":"publisher","first-page":"215","DOI":"10.1186\/s12967-022-03364-0","article-title":"Machine learning for the prediction of acute kidney injury in patients with sepsis.","volume":"20","author":"Yue","year":"2022","journal-title":"J Transl Med"},{"issue":"7","key":"soi230068r11","doi-asserted-by":"publisher","first-page":"1292","DOI":"10.1093\/jamia\/ocac058","article-title":"Use of unstructured text in prognostic clinical prediction models: a systematic review.","volume":"29","author":"Seinen","year":"2022","journal-title":"J Am Med Inform Assoc"},{"issue":"4","key":"soi230068r12","doi-asserted-by":"publisher","first-page":"1095","DOI":"10.1111\/ajt.15705","article-title":"Can donor narratives yield insights? a natural language processing proof of concept to facilitate kidney allocation.","volume":"20","author":"Placona","year":"2020","journal-title":"Am J Transplant"},{"issue":"12","key":"soi230068r15","doi-asserted-by":"publisher","first-page":"1935","DOI":"10.1093\/jamia\/ocaa189","article-title":"Clinical concept extraction using transformers.","volume":"27","author":"Yang","year":"2020","journal-title":"J Am Med Inform Assoc"},{"issue":"6","key":"soi230068r16","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbac409","article-title":"BioGPT: generative pretrained transformer for biomedical text generation and mining.","volume":"23","author":"Luo","year":"2022","journal-title":"Brief Bioinform"},{"issue":"4","key":"soi230068r17","doi-asserted-by":"publisher","first-page":"1234","DOI":"10.1093\/bioinformatics\/btz682","article-title":"BioBERT: a pretrained biomedical language representation model for biomedical text mining.","volume":"36","author":"Lee","year":"2020","journal-title":"Bioinformatics"},{"issue":"2","key":"soi230068r18","doi-asserted-by":"publisher","DOI":"10.2196\/32875","article-title":"Operationalizing and implementing pretrained, large artificial intelligence linguistic models in the US health care system: outlook of generative pretrained transformer 3 (GPT-3) as a service model.","volume":"10","author":"Sezgin","year":"2022","journal-title":"JMIR Med Inform"},{"key":"soi230068r19","doi-asserted-by":"publisher","first-page":"18","DOI":"10.3233\/SHTI210848","article-title":"An evaluation of pretrained BERT models for comparing semantic similarity across unstructured clinical trial texts.","volume":"289","author":"Patricoski","year":"2022","journal-title":"Stud Health Technol Inform"},{"issue":"2","key":"soi230068r20","doi-asserted-by":"publisher","DOI":"10.2196\/22795","article-title":"Adapting Bidirectional Encoder Representations From Transformers (BERT) to assess clinical semantic textual similarity: algorithm development and validation study.","volume":"9","author":"Kades","year":"2021","journal-title":"JMIR Med Inform"},{"issue":"10","key":"soi230068r21","doi-asserted-by":"publisher","first-page":"749","DOI":"10.1038\/s41551-018-0304-0","article-title":"Explainable machine-learning predictions for the prevention of hypoxaemia during surgery.","volume":"2","author":"Lundberg","year":"2018","journal-title":"Nat Biomed Eng"},{"issue":"4","key":"soi230068r22","doi-asserted-by":"publisher","first-page":"764","DOI":"10.1097\/TP.0000000000002362","article-title":"Prospective validation of prediction model for kidney discard.","volume":"103","author":"Zhou","year":"2019","journal-title":"Transplantation"},{"issue":"4","key":"soi230068r23","doi-asserted-by":"publisher","first-page":"783","DOI":"10.1111\/j.1600-6143.2008.02157.x","article-title":"Determinants of discard of expanded criteria donor kidneys: impact of biopsy and machine perfusion.","volume":"8","author":"Sung","year":"2008","journal-title":"Am J Transplant"},{"issue":"11","key":"soi230068r24","doi-asserted-by":"publisher","first-page":"e745","DOI":"10.1016\/S2589-7500(21)00208-9","article-title":"The false hope of current approaches to explainable artificial intelligence in health care.","volume":"3","author":"Ghassemi","year":"2021","journal-title":"Lancet Digit Health"},{"issue":"12","key":"soi230068r25","doi-asserted-by":"publisher","DOI":"10.2196\/26611","article-title":"Population preferences for performance and explainability of artificial intelligence in health care: choice-based conjoint survey.","volume":"23","author":"Ploug","year":"2021","journal-title":"J Med Internet Res"},{"issue":"1","key":"soi230068r26","doi-asserted-by":"publisher","first-page":"310","DOI":"10.1186\/s12911-020-01332-6","article-title":"Explainability for artificial intelligence in health care: a multidisciplinary perspective.","volume":"20","author":"Amann","year":"2020","journal-title":"BMC Med Inform Decis Mak"},{"key":"soi230068r27","doi-asserted-by":"publisher","DOI":"10.3389\/fdgth.2021.810260","article-title":"Predicting clinical events based on raw text: from bag-of-words to attention-based transformers.","volume":"3","author":"Roussinov","year":"2022","journal-title":"Front Digit Health"},{"issue":"1","key":"soi230068r28","doi-asserted-by":"publisher","first-page":"171","DOI":"10.1038\/s41746-022-00712-8","article-title":"Multimodal machine learning in precision health: a scoping review.","volume":"5","author":"Kline","year":"2022","journal-title":"NPJ Digit Med"},{"key":"soi230068r14","first-page":"4171","volume-title":"Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","author":"Devlin","year":"2019"},{"key":"soi230068r13","unstructured":"Vaswani? A, Shazeer? N, Parmar? N, . Attention is all you need.? arXiv. Preprint posted online June 12, 2017. doi:10.48550\/arXiv.1706.03762"}],"container-title":["JAMA Surgery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/jamanetwork.com\/journals\/jamasurgery\/articlepdf\/2811385\/jamasurgery_sageshima_2023_oi_230068_1703714476.71652.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,1,10]],"date-time":"2024-01-10T16:02:14Z","timestamp":1704902534000},"score":1,"resource":{"primary":{"URL":"https:\/\/jamanetwork.com\/journals\/jamasurgery\/fullarticle\/2811385"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,1,1]]},"references-count":28,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2024,1,1]]}},"URL":"https:\/\/doi.org\/10.1001\/jamasurg.2023.4679","relation":{},"ISSN":["2168-6254"],"issn-type":[{"value":"2168-6254","type":"print"}],"subject":[],"published":{"date-parts":[[2024,1,1]]}}}