{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T17:03:26Z","timestamp":1790269406557,"version":"4.1.0"},"reference-count":18,"publisher":"SAGE Publications","license":[{"start":{"date-parts":[[2026,9,2]],"date-time":"2026-09-02T00:00:00Z","timestamp":1788307200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["Ann Clin Biochem"],"abstract":"\n Background<\/jats:title>\n \n K\n 2<\/jats:sub>\n EDTA contamination of serum samples is a preanalytical error that poses a risk to patient safety. Contamination is most frequently mild to moderate, which standard detection procedures often miss. We examined whether machine learning models could improve the detection of K\n 2<\/jats:sub>\n EDTA contamination.\n <\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n \n Artificial neural network, decision tree (both simple and complex), extreme gradient boosting, k-nearest neighbours, logistic regression, na\u00efve Bayes, random forest, and support vector machine models were developed. Models were trained using extracted patient results for electrolytes, urea, creatinine, albumin-adjusted calcium, magnesium, and phosphate, with K\n 2<\/jats:sub>\n EDTA contamination errors simulated in silico. Model performance was evaluated on 300 real-world samples, half of which were intentionally contaminated with mild to moderate amounts of K\n 2<\/jats:sub>\n EDTA. Model performance was compared with that of limit checks, multi-analyte rules and two novel parameters, the potassium\/calcium ratio and the potassium\/magnesium ratio.\n <\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n All nine machine learning models identified K\n 2<\/jats:sub>\n EDTA contamination more accurately than standard approaches (\n p<\/jats:italic>\n -values <0.05). Seven models performed similarly, with accuracies of 91.3\u201393.3%, sensitivities of 88.0\u201390.0%, specificities of 94.7\u201396.7%, and area under the receiver operating characteristic curve (AUROC) of 0.9727\u20130.9787. The simple decision tree and na\u00efve Bayes models performed slightly worse. The potassium\/calcium ratio was the most effective of the standard approaches, with an accuracy of 79.7%, sensitivity of 68.7%, specificity of 90.7%, and AUROC of 0.9233.\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n \n Using machine learning models would enable better detection of mild to moderate K\n 2<\/jats:sub>\n EDTA serum contamination and improve patient safety. For laboratories unable to implement machine learning models, the best alternative is the potassium\/calcium ratio.\n <\/jats:p>\n <\/jats:sec>","DOI":"10.1177\/00045632261488751","type":"journal-article","created":{"date-parts":[[2026,9,3]],"date-time":"2026-09-03T01:22:38Z","timestamp":1788398558000},"update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":0,"title":["Machine learning for the identification of mild to moderate EDTA contamination of serum samples"],"prefix":"10.1177","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4511-3095","authenticated-orcid":false,"given":"Christopher-John L.","family":"Farrell","sequence":"first","affiliation":[{"name":"NSW Health Pathology"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Julie","family":"Sherfan","sequence":"additional","affiliation":[{"name":"NSW Health Pathology"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4846-9134","authenticated-orcid":false,"given":"Tony","family":"Badrick","sequence":"additional","affiliation":[{"name":"Royal College of Pathologists of Australasian Quality Assurance Programs, Sydney, NSW, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"179","published-online":{"date-parts":[[2026,9,2]]},"reference":[{"key":"e_1_3_3_2_2","doi-asserted-by":"publisher","DOI":"10.1111\/j.1742-1241.2008.01981.x"},{"key":"e_1_3_3_3_2","doi-asserted-by":"publisher","DOI":"10.1258\/0004563021901973"},{"key":"e_1_3_3_4_2","doi-asserted-by":"publisher","DOI":"10.1258\/acb.2008.007241"},{"key":"e_1_3_3_5_2","doi-asserted-by":"publisher","DOI":"10.1515\/cclm-2016-0426"},{"key":"e_1_3_3_6_2","doi-asserted-by":"publisher","DOI":"10.1177\/0004563219878463"},{"key":"e_1_3_3_7_2","doi-asserted-by":"publisher","DOI":"10.1515\/cclm-2014-0944"},{"key":"e_1_3_3_8_2","doi-asserted-by":"publisher","DOI":"10.1177\/0004563218812500"},{"key":"e_1_3_3_9_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.clinbiochem.2019.08.002"},{"key":"e_1_3_3_10_2","first-page":"597","article-title":"Multi-centre observational study of spurious hyperkalaemia due to EDTA contamination","volume":"56","author":"Cornes MP","year":"2010","unstructured":"Cornes MP, Davidson F, Darwin L, et al. Multi-centre observational study of spurious hyperkalaemia due to EDTA contamination. Clin Lab 2010; 56: 597\u2013599.","journal-title":"Clin Lab"},{"key":"e_1_3_3_11_2","doi-asserted-by":"publisher","DOI":"10.1093\/clinchem\/hvad100"},{"key":"e_1_3_3_12_2","doi-asserted-by":"publisher","DOI":"10.1093\/clinchem\/hvad207"},{"key":"e_1_3_3_13_2","doi-asserted-by":"publisher","DOI":"10.1177\/00045632221140989"},{"key":"e_1_3_3_14_2","doi-asserted-by":"publisher","DOI":"10.21037\/jlpm-24-42"},{"key":"e_1_3_3_15_2","volume-title":"R: a language and environmental for statistical computing","author":"R Core Team","year":"2022","unstructured":"R Core Team. R: a language and environmental for statistical computing. R Foundation for Statistical Computing, 2022."},{"key":"e_1_3_3_16_2","first-page":"213","article-title":"Harmonising adult and paediatric reference intervals in Australia and New Zealand: an evidence-based approach for establishing a first panel of chemistry analytes","volume":"35","author":"Tate JR","year":"2014","unstructured":"Tate JR, Sikaris KA, Jones GR, et al. Harmonising adult and paediatric reference intervals in Australia and New Zealand: an evidence-based approach for establishing a first panel of chemistry analytes. Clin Biochem Rev 2014; 35: 213\u2013235.","journal-title":"Clin Biochem Rev"},{"key":"e_1_3_3_17_2","doi-asserted-by":"publisher","DOI":"10.1373\/jalm.2018.027920"},{"key":"e_1_3_3_18_2","doi-asserted-by":"publisher","DOI":"10.1093\/ajcp\/aqw104"},{"key":"e_1_3_3_19_2","doi-asserted-by":"publisher","DOI":"10.1515\/CCLM.2006.123"}],"container-title":["Annals of Clinical Biochemistry: International Journal of Laboratory Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/00045632261488751","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.1177\/00045632261488751","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/00045632261488751","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T16:22:30Z","timestamp":1790266950000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.1177\/00045632261488751"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,2]]},"references-count":18,"alternative-id":["10.1177\/00045632261488751"],"URL":"https:\/\/doi.org\/10.1177\/00045632261488751","relation":{},"ISSN":["0004-5632","1758-1001"],"issn-type":[{"value":"0004-5632","type":"print"},{"value":"1758-1001","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,2]]},"article-number":"00045632261488751"}}