{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,3]],"date-time":"2026-03-03T16:44:34Z","timestamp":1772556274087,"version":"3.50.1"},"reference-count":22,"publisher":"Georg Thieme Verlag KG","issue":"01","funder":[{"name":"National Library of Medicine Training","award":["5T15LM007450-13"],"award-info":[{"award-number":["5T15LM007450-13"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Appl Clin Inform"],"published-print":{"date-parts":[[2016,1]]},"abstract":"<jats:title>Summary<\/jats:title><jats:p>Discharging patients from the Neonatal Intensive Care Unit (NICU) can be delayed for non-medical reasons including the procurement of home medical equipment, parental education, and the need for children\u2019s services. We previously created a model to identify patients that will be medically ready for discharge in the subsequent 2\u201310 days. In this study we use Natural Language Processing to improve upon that model and discern why the model performed poorly on certain patients.<\/jats:p><jats:p>We retrospectively examined the text of the Assessment and Plan section from daily progress notes of 4,693 patients (103,206 patient-days) from the NICU of a large, academic children\u2019s hospital. A matrix was constructed using words from NICU notes (single words and bigrams) to train a supervised machine learning algorithm to determine the most important words differentiating poorly performing patients compared to well performing patients in our original discharge prediction model.<\/jats:p><jats:p>NLP using a bag of words (BOW) analysis revealed several cohorts that performed poorly in our original model. These included patients with surgical diagnoses, pulmonary hypertension, retinopathy of prematurity, and psychosocial issues.<\/jats:p><jats:p>The BOW approach aided in cohort discovery and will allow further refinement of our original discharge model prediction. Adequately identifying patients discharged home on g-tube feeds alone could improve the AUC of our original model by 0.02. Additionally, this approach identified social issues as a major cause for delayed discharge.<\/jats:p><jats:p>A BOW analysis provides a method to improve and refine our NICU discharge prediction model and could potentially avoid over 900 (0.9%) hospital days.<\/jats:p><jats:p>AUC \u2013 Area under the Curve, CART -- Classification And Regression Trees, DTD \u2013 Days to Dis- charge, GI \u2013 Gastrointestinal, LOS \u2013 Length of Stay, NICU \u2013 Neonatal Intensive Care Unit, NS \u2013 Neurosurgery, RF \u2013 Random Forest.<\/jats:p>","DOI":"10.4338\/aci-2015-09-ra-0114","type":"journal-article","created":{"date-parts":[[2016,2,24]],"date-time":"2016-02-24T07:48:36Z","timestamp":1456300116000},"page":"101-115","source":"Crossref","is-referenced-by-count":14,"title":["Natural Language Processing for Cohort Discovery in a Discharge Prediction Model for the Neonatal ICU"],"prefix":"10.4338","volume":"07","author":[{"given":"Christoph","family":"Lehmann","sequence":"first","affiliation":[]},{"given":"Daniel","family":"Fabbri","sequence":"first","affiliation":[]},{"given":"Michael","family":"Temple","sequence":"additional","affiliation":[]}],"member":"194","published-online":{"date-parts":[[2017,12,16]]},"reference":[{"key":"10.4338\/ACI-2015-09-RA-0114-1","doi-asserted-by":"publisher","DOI":"10.1038\/jp.2013.136"},{"key":"10.4338\/ACI-2015-09-RA-0114-2","doi-asserted-by":"publisher","DOI":"10.1002\/gps.3983"},{"key":"10.4338\/ACI-2015-09-RA-0114-3","doi-asserted-by":"publisher","DOI":"10.1542\/peds.2015-0456"},{"key":"10.4338\/ACI-2015-09-RA-0114-4","doi-asserted-by":"publisher","DOI":"10.1542\/peds.2012-2189"},{"key":"10.4338\/ACI-2015-09-RA-0114-5","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.319.7217.1093"},{"key":"10.4338\/ACI-2015-09-RA-0114-6","doi-asserted-by":"publisher","DOI":"10.1542\/peds.2009-0810"},{"key":"10.4338\/ACI-2015-09-RA-0114-7","doi-asserted-by":"publisher","DOI":"10.1197\/jamia.M3096"},{"key":"10.4338\/ACI-2015-09-RA-0114-8","doi-asserted-by":"publisher","DOI":"10.1136\/jamia.2010.003863"},{"key":"10.4338\/ACI-2015-09-RA-0114-9","doi-asserted-by":"publisher","DOI":"10.1136\/amiajnl-2011-000163"},{"key":"10.4338\/ACI-2015-09-RA-0114-10","doi-asserted-by":"publisher","DOI":"10.1136\/amiajnl-2012-001576"},{"key":"10.4338\/ACI-2015-09-RA-0114-11","first-page":"1191","volume":"2012","author":"Cui","year":"2012","journal-title":"AMIA Annual Symposium proceedings \/ AMIA Symposium AMIA Symposium"},{"key":"10.4338\/ACI-2015-09-RA-0114-12","first-page":"103","volume":"2013","author":"Bejan","year":"2013","journal-title":"AMIA Annual Symposium proceedings\/AMIA Symposium AMIA Symposium"},{"key":"10.4338\/ACI-2015-09-RA-0114-13","doi-asserted-by":"publisher","DOI":"10.1136\/amiajnl-2013-002463"},{"key":"10.4338\/ACI-2015-09-RA-0114-14","doi-asserted-by":"publisher","DOI":"10.1136\/amiajnl-2013-001924"},{"key":"10.4338\/ACI-2015-09-RA-0114-15","doi-asserted-by":"publisher","DOI":"10.1136\/amiajnl-2013-002601"},{"key":"10.4338\/ACI-2015-09-RA-0114-16","doi-asserted-by":"publisher","DOI":"10.1016\/j.jbi.2014.02.003"},{"key":"10.4338\/ACI-2015-09-RA-0114-17","unstructured":"http:\/\/www.nltk.org"},{"key":"10.4338\/ACI-2015-09-RA-0114-18","unstructured":"http:\/\/scikit-learn.org\/stable\/index.html"},{"key":"10.4338\/ACI-2015-09-RA-0114-19","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpedsurg.2014.06.010"},{"key":"10.4338\/ACI-2015-09-RA-0114-20","doi-asserted-by":"publisher","DOI":"10.1038\/sj.jp.7210879"},{"key":"10.4338\/ACI-2015-09-RA-0114-21","doi-asserted-by":"publisher","DOI":"10.1038\/jp.2011.19"},{"key":"10.4338\/ACI-2015-09-RA-0114-22","doi-asserted-by":"crossref","unstructured":"Chapman WW, Nadkarni PM, Hirschman L, D\u2019Avolio LW, Savova GK, Uzuner O. Overcoming barriers to NLP for clinical text: the role of shared tasks and the need for additional creative solutions 2011; 2011-09-01 00:00:00. 540-3 p","DOI":"10.1136\/amiajnl-2011-000465"}],"container-title":["Applied Clinical Informatics"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/www.thieme-connect.de\/products\/ejournals\/pdf\/10.4338\/ACI-2015-09-RA-0114.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2019,9,4]],"date-time":"2019-09-04T22:56:58Z","timestamp":1567637818000},"score":1,"resource":{"primary":{"URL":"http:\/\/www.thieme-connect.de\/DOI\/DOI?10.4338\/ACI-2015-09-RA-0114"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2016,1]]},"references-count":22,"journal-issue":{"issue":"01","published-online":{"date-parts":[[2017,12,16]]},"published-print":{"date-parts":[[2016]]}},"URL":"https:\/\/doi.org\/10.4338\/aci-2015-09-ra-0114","relation":{},"ISSN":["1869-0327"],"issn-type":[{"value":"1869-0327","type":"electronic"}],"subject":[],"published":{"date-parts":[[2016,1]]}}}