{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T11:08:08Z","timestamp":1784113688469,"version":"3.55.0"},"reference-count":31,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,6,12]],"date-time":"2026-06-12T00:00:00Z","timestamp":1781222400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["clinicalkey.com","clinicalkey.com.au","clinicalkey.es","clinicalkey.fr","clinicalkey.jp","elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["International Journal of Medical Informatics"],"published-print":{"date-parts":[[2026,10]]},"DOI":"10.1016\/j.ijmedinf.2026.106538","type":"journal-article","created":{"date-parts":[[2026,6,11]],"date-time":"2026-06-11T20:08:47Z","timestamp":1781208527000},"page":"106538","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":1,"special_numbering":"C","title":["Why almost all ML models for medicine are wrong-and what we need for evidence-based medical AI"],"prefix":"10.1016","volume":"219","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4065-3415","authenticated-orcid":false,"given":"Federico","family":"Cabitza","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Giuseppe","family":"Jurman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Filippo","family":"Molinari","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Riccardo","family":"Bellazzi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"issue":"1","key":"10.1016\/j.ijmedinf.2026.106538_bib0005","doi-asserted-by":"crossref","first-page":"44","DOI":"10.1038\/s41591-018-0300-7","article-title":"High-performance medicine: the convergence of human and artificial intelligence","volume":"25","author":"Topol","year":"2019","journal-title":"Nat. Med."},{"issue":"1","key":"10.1016\/j.ijmedinf.2026.106538_bib0010","doi-asserted-by":"crossref","first-page":"125","DOI":"10.1038\/s41746-024-01103-x","article-title":"A systematic review and meta-analysis of artificial intelligence versus clinicians for skin cancer diagnosis","volume":"7","author":"Salinas","year":"2024","journal-title":"NPJ Digit. Med."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0015","doi-asserted-by":"crossref","first-page":"m689","DOI":"10.1136\/bmj.m689","article-title":"Artificial intelligence versus clinicians: systematic review of design, reporting standards, and claims of deep learning studies","volume":"368","author":"Nagendran","year":"2020","journal-title":"BMJ"},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0020","doi-asserted-by":"crossref","DOI":"10.1016\/j.artmed.2019.101785","article-title":"The impact of machine learning on patient care: a systematic review","volume":"103","author":"Ben-Israel","year":"2020","journal-title":"Artif. Intell. Med."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0025","series-title":"Proceedings of the ACM Conference on Health, Inference, and Learning","first-page":"151","article-title":"Hidden stratification causes clinically meaningful failures in machine learning for medical imaging","author":"Oakden-Rayner","year":"2020"},{"issue":"1","key":"10.1016\/j.ijmedinf.2026.106538_bib0030","doi-asserted-by":"crossref","first-page":"219","DOI":"10.1186\/s12911-020-01224-9","article-title":"As if sand were stone. New concepts and metrics to probe the ground on which to build trustable AI","volume":"20","author":"Cabitza","year":"2020","journal-title":"BMC Med. Inform. Decis. Mak."},{"issue":"8","key":"10.1016\/j.ijmedinf.2026.106538_bib0035","first-page":"1047","article-title":"The EFLM strategy for harmonization of the preanalytical phase. A joint recommendation by the EFLM working group for preanalytical phase (WG-PRE) and the latin American working group for preanalytical phase (WG-PRE-LATAM)","volume":"55","author":"Lippi","year":"2017","journal-title":"Clin. Chem. Lab. Med."},{"issue":"4","key":"10.1016\/j.ijmedinf.2026.106538_bib0040","doi-asserted-by":"crossref","first-page":"684","DOI":"10.1515\/cclm-2024-0971","volume":"63","author":"Padoan","year":"2025","journal-title":"Clin. Chem. Lab. Med. (CCLM)"},{"issue":"11","key":"10.1016\/j.ijmedinf.2026.106538_bib0045","doi-asserted-by":"crossref","first-page":"1122","DOI":"10.1001\/jama.2015.1405","article-title":"Diagnostic concordance among pathologists interpreting breast biopsy specimens","volume":"313","author":"Elmore","year":"2015","journal-title":"JAMA"},{"issue":"11","key":"10.1016\/j.ijmedinf.2026.106538_bib0050","doi-asserted-by":"crossref","first-page":"4014","DOI":"10.3390\/app10114014","article-title":"The elephant in the machine: proposing a new metric of data reliability and its application to a medical case to assess classification reliability","volume":"10","author":"Cabitza","year":"2020","journal-title":"Appl. Sci."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0055","doi-asserted-by":"crossref","first-page":"i6","DOI":"10.1136\/bmj.i6","article-title":"Net benefit approaches to the evaluation of prediction models, molecular markers, and diagnostic tests","volume":"352","author":"Vickers","year":"2016","journal-title":"BMJ"},{"issue":"4","key":"10.1016\/j.ijmedinf.2026.106538_bib0060","doi-asserted-by":"crossref","first-page":"932","DOI":"10.1007\/s00330-014-3487-0","article-title":"Disadvantages of using the area under the receiver operating characteristic curve to assess imaging tests: a discussion and proposal for an alternative approach","volume":"25","author":"Halligan","year":"2015","journal-title":"European radiology"},{"issue":"20","key":"10.1016\/j.ijmedinf.2026.106538_bib0065","doi-asserted-by":"crossref","first-page":"1109","DOI":"10.1056\/NEJM198005153022003","article-title":"The threshold approach to clinical decision making","volume":"302","author":"Pauker","year":"1980","journal-title":"New England Journal of Medicine"},{"issue":"6","key":"10.1016\/j.ijmedinf.2026.106538_bib0070","doi-asserted-by":"crossref","first-page":"796","DOI":"10.1016\/j.eururo.2018.08.038","article-title":"Reporting and interpreting decision curve analysis: a guide for investigators","volume":"74","author":"Van Calster","year":"2018","journal-title":"Eur. Urol."},{"issue":"10","key":"10.1016\/j.ijmedinf.2026.106538_bib0075","doi-asserted-by":"crossref","first-page":"3962","DOI":"10.1038\/s41591-025-04050-6","article-title":"The missing value of medical artificial intelligence","volume":"12","author":"Kohane","year":"2025","journal-title":"Nat. Med."},{"issue":"11","key":"10.1016\/j.ijmedinf.2026.106538_bib0080","doi-asserted-by":"crossref","first-page":"E537","DOI":"10.1503\/cmaj.121286","article-title":"Variation of a test\u2019s sensitivity and specificity with disease prevalence","volume":"185","author":"Leeflang","year":"2013","journal-title":"CMAJ"},{"issue":"29","key":"10.1016\/j.ijmedinf.2026.106538_bib0085","doi-asserted-by":"crossref","first-page":"1925","DOI":"10.1093\/eurheartj\/ehu207","article-title":"Towards better clinical prediction models: seven steps for development and an ABCD for validation","volume":"35","author":"Steyerberg","year":"2014","journal-title":"Eur. Heart J."},{"issue":"6","key":"10.1016\/j.ijmedinf.2026.106538_bib0090","doi-asserted-by":"crossref","first-page":"565","DOI":"10.1177\/0272989X06295361","article-title":"Decision curve analysis: a novel method for evaluating prediction models","volume":"26","author":"Vickers","year":"2006","journal-title":"Med. Decis. Mak."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0095","doi-asserted-by":"crossref","DOI":"10.1016\/j.jbi.2022.103996","article-title":"Evaluating pointwise reliability of machine learning prediction","author":"Nicora","year":"2022","journal-title":"J. Biomed. Inform."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0100","doi-asserted-by":"crossref","DOI":"10.1016\/j.ijmedinf.2025.105857","article-title":"RelAI: an automated approach to judge pointwise ML prediction reliability","volume":"197","author":"Peracchio","year":"2025","journal-title":"Int. J. Med. Inform."},{"issue":"1","key":"10.1016\/j.ijmedinf.2026.106538_bib0105","doi-asserted-by":"crossref","first-page":"146","DOI":"10.1186\/s12874-024-02268-5","article-title":"An evaluation of sample size requirements for developing risk prediction models with binary outcomes","volume":"24","author":"Pavlou","year":"2024","journal-title":"BMC Med. Res. Methodol."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0110","doi-asserted-by":"crossref","DOI":"10.1136\/bmj.g7594","article-title":"Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement","volume":"350","author":"Collins","year":"2015","journal-title":"BMJ"},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0115","doi-asserted-by":"crossref","DOI":"10.1016\/j.ijmedinf.2025.105970","article-title":"Show and tell: a critical review on robustness and uncertainty for a more responsible medical AI","author":"Marconi","year":"2025","journal-title":"Int. J. Med. Inform."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0120","doi-asserted-by":"crossref","DOI":"10.1016\/j.cmpb.2021.106288","article-title":"The importance of being external. Methodological insights for the external validation of machine learning models in medicine","volume":"208","author":"Cabitza","year":"2021","journal-title":"Comput. Methods Programs Biomed."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0125","article-title":"Prediction models for diagnosis and prognosis of COVID-19: systematic review and critical appraisal","volume":"369","author":"Wynants","year":"2020","journal-title":"BMJ"},{"issue":"6","key":"10.1016\/j.ijmedinf.2026.106538_bib0130","doi-asserted-by":"crossref","first-page":"1052","DOI":"10.1093\/jamia\/ocx030","article-title":"Calibration drift in regression and machine learning models for acute kidney injury","volume":"24","author":"Davis","year":"2017","journal-title":"J. Am. Med. Inform. Assoc."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0135","article-title":"Maturity framework for operationalizing machine learning applications in health care: scoping review","author":"Li","year":"2025","journal-title":"J. Med. Internet Res."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0140","series-title":"Proceedings of the AAAI Conference on Artificial Intelligence, 37","first-page":"6860","article-title":"Toward a perspectivist turn in ground truthing for predictive computing","author":"Cabitza","year":"2023"},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0145","doi-asserted-by":"crossref","DOI":"10.1016\/j.ijmedinf.2021.104510","article-title":"The need to separate the wheat from the chaff in medical informatics: introducing a comprehensive checklist for the (self)-assessment of medical AI studies","volume":"153","author":"Cabitza","year":"2021","journal-title":"Int. J. Med. Inform."},{"key":"10.1016\/j.ijmedinf.2026.106538_bib0150","article-title":"TRIPOD+ AI: an updated reporting guideline for clinical prediction models","volume":"385","author":"Cohen","year":"2024","journal-title":"BMJ"},{"issue":"March 2024","key":"10.1016\/j.ijmedinf.2026.106538_bib0155","article-title":"Evidence-based XAI: an empirical approach to design more effective and explainable decision support systems","volume":"170","author":"Famiglini","year":"2024","journal-title":"Comput. Biol. Med."}],"container-title":["International Journal of Medical Informatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S1386505626002789?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S1386505626002789?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T10:52:27Z","timestamp":1784112747000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S1386505626002789"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10]]},"references-count":31,"alternative-id":["S1386505626002789"],"URL":"https:\/\/doi.org\/10.1016\/j.ijmedinf.2026.106538","relation":{},"ISSN":["1386-5056"],"issn-type":[{"value":"1386-5056","type":"print"}],"subject":[],"published":{"date-parts":[[2026,10]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Why almost all ML models for medicine are wrong-and what we need for evidence-based medical AI","name":"articletitle","label":"Article Title"},{"value":"International Journal of Medical Informatics","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.ijmedinf.2026.106538","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 The Authors. Published by Elsevier B.V.","name":"copyright","label":"Copyright"}],"article-number":"106538"}}