Association for Academic Surgery Using artificial intelligence to predict prolonged mechanical ventilation and tracheostomy placement Joshua Parreco, MD, a Antonio Hidalgo, MS, a Jonathan J. Parks, MD, a Robert Kozol, MD, a and Rishi Rattan, MD b, * a DeWitt Daughtry Family Department of Surgery, University of Miami, Miller School of Medicine, Miami, Florida b Division of Trauma Surgery and Surgical Critical Care, DeWitt Daughtry Family Department of Surgery, University of Miami, Miller School of Medicine, Miami, Florida article info Article history: Received 30 November 2017 Received in revised form 7 February 2018 Accepted 14 March 2018 Available online xxx Keywords: Tracheostomy Prolonged mechanical ventilation Machine learning Artificial intelligence Critical care abstract Background: Early identification of critically ill patients who will require prolonged mechanical ventilation (PMV) has proven to be difficult. The purpose of this study was to use machine learning to identify patients at risk for PMV and tracheostomy placement. Materials and methods: The Multiparameter Intelligent Monitoring in Intensive Care III database was queried for all intensive care unit (ICU) stays with mechanical ventilation. PMV was defined as ventilation >7 d. Classifiers with a gradient-boosted decision trees algorithm were created for the outcomes of PMV and tracheostomy placement. The variables used were six different severity-of-illness scores calculated on the first day of ICU admission including their components and 30 comorbidities. Mean receiver operating characteristic curves were calculated for the outcomes, and variable importance was quantified. Results: There were 20,262 ICU stays identified. PMV was required in 13.6%, and tracheos- tomy was performed in 6.6% of patients. The classifier for predicting PMV was able to achieve a mean area under the curve (AUC) of 0.820  0.016, and tracheostomy was pre- dicted with an AUC of 0.830  0.011. There were 60.7% patients admitted to a surgical ICU, and the classifiers for these patients predicted PMV with an AUC of 0.852  0.017 and tracheostomy with an AUC of 0.869  0.015. The variable with the highest importance for predicting PMV was the logistic organ dysfunction score pulmonary component (13%), and the most important comorbidity in predicting tracheostomy was cardiac arrhythmia (12%). Conclusions: This study demonstrates the use of artificial intelligence through machine- learning classifiers for the early identification of patients at risk for PMV and tracheos- tomy. Application of these identification techniques could lead to improved outcomes by allowing for early intervention. ª 2018 Elsevier Inc. All rights reserved. * Corresponding author. Division of Trauma Surgery and Surgical Critical Care, DeWitt Daughtry Family Department of Surgery, Uni- versity of Miami, Miller School of Medicine, 1800 NW 10th Avenue, T215 (D-40), Miami, FL 33136. Tel.: þ1 305 585 1822; fax: þ1 305 326 7065. E-mail address: rrattan@med.miami.edu (R. Rattan). Available online at www.sciencedirect.com ScienceDirect journal homepage: www.JournalofSurgicalResearch.com journal of surgical research  august 2018 (228) 179 e187 0022-4804/$ e see front matter ª 2018 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jss.2018.03.028