{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T10:31:34Z","timestamp":1740133894037,"version":"3.37.3"},"reference-count":20,"publisher":"BMJ","issue":"5","funder":[{"name":"\"Trauma Audit and Research Network - www.tarn.ac.uk\"","award":["Grant Number Not Applicable\/NA"],"award-info":[{"award-number":["Grant Number Not Applicable\/NA"]}]},{"DOI":"10.13039\/100011272","name":"FP7 Health","doi-asserted-by":"publisher","award":["EC grant 602150"],"award-info":[{"award-number":["EC grant 602150"]}],"id":[{"id":"10.13039\/100011272","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000659","name":"Research Trainees Coordinating Centre","doi-asserted-by":"publisher","award":["CDF-2015-08-11","CM is a NIHR CL Grant Number Not Applicable\/NA"],"award-info":[{"award-number":["CDF-2015-08-11","CM is a NIHR CL Grant Number Not Applicable\/NA"]}],"id":[{"id":"10.13039\/501100000659","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["Emerg Med J"],"published-print":{"date-parts":[[2022,5]]},"abstract":"<jats:sec><jats:title>Background<\/jats:title><jats:p>Patients with mild traumatic brain injury on CT scan are routinely admitted for inpatient observation. Only a small proportion of patients require clinical intervention. We recently developed a decision rule using traditional statistical techniques that found neurologically intact patients with isolated simple skull fractures or single bleeds &lt;5\u2009mm with no preinjury antiplatelet or anticoagulant use may be safely discharged from the emergency department. The decision rule achieved a sensitivity of 99.5% (95% CI 98.1% to 99.9%) and specificity of 7.4% (95% CI 6.0% to 9.1%) to clinical deterioration. We aimed to transparently report a machine learning approach to assess if predictive accuracy could be improved.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>We used data from the same retrospective cohort of 1699 initial Glasgow Coma Scale (GCS) 13\u201315 patients with injuries identified by CT who presented to three English Major Trauma Centres between 2010 and 2017 as in our original study. We assessed the ability of machine learning to predict the same composite outcome measure of deterioration (indicating need for hospital admission). Predictive models were built using gradient boosted decision trees which consisted of an ensemble of decision trees to optimise model performance.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>The final algorithm reported a mean positive predictive value of 29%, mean negative predictive value of 94%, mean area under the curve (C-statistic) of 0.75, mean sensitivity of 99% and mean specificity of 7%. As with logistic regression, GCS, severity and number of brain injuries were found to be important predictors of deterioration.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>We found no clear advantages over the traditional prediction methods, although the models were, effectively, developed using a smaller data set, due to the need to divide it into training, calibration and validation sets. Future research should focus on developing models that provide clear advantages over existing classical techniques in predicting outcomes in this population.<\/jats:p><\/jats:sec>","DOI":"10.1136\/emermed-2020-210776","type":"journal-article","created":{"date-parts":[[2021,4,8]],"date-time":"2021-04-08T17:41:23Z","timestamp":1617903683000},"page":"394-401","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":6,"title":["Predicting need for hospital admission in patients with traumatic brain injury or skull fractures identified on CT imaging: a machine learning approach"],"prefix":"10.1136","volume":"39","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3043-7564","authenticated-orcid":false,"given":"Carl","family":"Marincowitz","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Lewis","family":"Paton","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fiona","family":"Lecky","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Paul","family":"Tiffin","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"239","published-online":{"date-parts":[[2021,4,8]]},"reference":[{"key":"2023100106350660000_39.5.394.1","unstructured":"NICE . CG 176 head injury triage, assessment, investigation and early management of head injury in children, young people and adult. In: Department of Health, ed, 2014."},{"key":"2023100106350660000_39.5.394.2","doi-asserted-by":"publisher","DOI":"10.1007\/BF01743065"},{"key":"2023100106350660000_39.5.394.3","doi-asserted-by":"crossref","first-page":"100","DOI":"10.1056\/NEJM200007133430204","article-title":"Indications for computed tomography in patients with minor head injury","volume":"343","author":"Haydel","year":"2000","journal-title":"N Engl J Med Overseas Ed"},{"key":"2023100106350660000_39.5.394.4","doi-asserted-by":"publisher","DOI":"10.1089\/neu.2017.5259"},{"key":"2023100106350660000_39.5.394.5","doi-asserted-by":"publisher","DOI":"10.1016\/j.jamcollsurg.2009.12.039"},{"key":"2023100106350660000_39.5.394.6","doi-asserted-by":"crossref","first-page":"965","DOI":"10.1097\/TA.0000000000000161","article-title":"The big (brain injury guidelines) project: defining the management of traumatic brain injury by acute care surgeons","volume":"76","author":"Joseph","year":"2014","journal-title":"J Trauma Acute Care Surg"},{"key":"2023100106350660000_39.5.394.7","doi-asserted-by":"crossref","first-page":"6","DOI":"10.1186\/s41512-018-0027-4","article-title":"A protocol for the development of a prediction model in mild traumatic brain injury with CT scan abnormality: which patients are safe for discharge?","volume":"2","author":"Marincowitz","year":"2018","journal-title":"Diagn Progn Res"},{"key":"2023100106350660000_39.5.394.8","doi-asserted-by":"crossref","first-page":"324","DOI":"10.1089\/neu.2019.6652","article-title":"Development of a clinical decision rule for the early safe discharge of patients with mild traumatic brain injury and findings on computed tomography brain scan: a retrospective cohort study","volume":"37","author":"Marincowitz","year":"2020","journal-title":"J Neurotrauma"},{"key":"2023100106350660000_39.5.394.9","doi-asserted-by":"publisher","DOI":"10.1016\/j.jclinepi.2019.02.004"},{"key":"2023100106350660000_39.5.394.10","doi-asserted-by":"publisher","DOI":"10.1016\/s0140-6736(19)30037-6"},{"key":"2023100106350660000_39.5.394.11","unstructured":"CatBoost . CatBoost. Available: https:\/\/catboost.ai\/ [Accessed 14th April 2020]."},{"key":"2023100106350660000_39.5.394.12","unstructured":"R Foundation for Statistical Computing . R: A Language and Environment for Statistical Computing [program], 2019."},{"key":"2023100106350660000_39.5.394.13","unstructured":"Kaggle Forum . Ranking of Kaggle algorithms by competitions won 2016, 2018. Available: https:\/\/www.kaggle.com\/general\/25913 [Accessed 20th August 2018]."},{"key":"2023100106350660000_39.5.394.14","doi-asserted-by":"publisher","DOI":"10.1613\/jair.953"},{"key":"2023100106350660000_39.5.394.15","unstructured":"Niculescu-Mizil A , Caruana R . Obtaining calibrated probabilities from boosting. UAI, 2005: 413."},{"key":"2023100106350660000_39.5.394.16"},{"key":"2023100106350660000_39.5.394.17","doi-asserted-by":"publisher","DOI":"10.1007\/BF02246500"},{"key":"2023100106350660000_39.5.394.18","doi-asserted-by":"crossref","first-page":"1342","DOI":"10.1038\/s41591-018-0107-6","article-title":"Clinically applicable deep learning for diagnosis and referral in retinal disease","volume":"24","author":"De Fauw","year":"2018","journal-title":"Nat Med"},{"key":"2023100106350660000_39.5.394.19","doi-asserted-by":"crossref","DOI":"10.1186\/s13326-019-0214-4","article-title":"Natural language processing for disease phenotyping in UK primary care records for research: a pilot study in myocardial infarction and death","volume":"10","author":"Shah","year":"2019","journal-title":"J Biomed Semantics"},{"key":"2023100106350660000_39.5.394.20","doi-asserted-by":"crossref","DOI":"10.1186\/s12916-018-1099-2","article-title":"Poor reporting of multivariable prediction model studies: towards a targeted implementation strategy of the TRIPOD statement","volume":"16","author":"Heus","year":"2018","journal-title":"BMC Med"}],"container-title":["Emergency Medicine Journal"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/emermed-2020-210776","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,10,1]],"date-time":"2023-10-01T13:36:30Z","timestamp":1696167390000},"score":1,"resource":{"primary":{"URL":"https:\/\/emj.bmj.com\/lookup\/doi\/10.1136\/emermed-2020-210776"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,4,8]]},"references-count":20,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2022,4,21]]},"published-print":{"date-parts":[[2022,5]]}},"alternative-id":["10.1136\/emermed-2020-210776"],"URL":"https:\/\/doi.org\/10.1136\/emermed-2020-210776","relation":{},"ISSN":["1472-0205","1472-0213"],"issn-type":[{"type":"print","value":"1472-0205"},{"type":"electronic","value":"1472-0213"}],"subject":[],"published":{"date-parts":[[2021,4,8]]}}}