Brian L Hill1, Robert Brown1, Eilon Gabel2, Nadav Rakocz1, Christine Lee3, Maxime Cannesson2, Pierre Baldi4, Loes Olde Loohuis5, Ruth Johnson1, Brandon Jew6, Uri Maoz7, Aman Mahajan8, Sriram Sankararaman9, Ira Hofer10, Eran Halperin11. 1. Department of Computer Science, University of California, Los Angeles, CA, USA. 2. Department of Anaesthesiology and Perioperative Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA. 3. Department of Anaesthesiology and Perioperative Care, University of California, Irvine, CA, USA; Department of Biomedical Engineering, University of California, Irvine, CA, USA. 4. Department of Biomedical Engineering, University of California, Irvine, CA, USA; Department of Computer Science, University of California, Irvine, CA, USA. 5. Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behaviour, University of California, Los Angeles, CA, USA. 6. Bioinformatics Interdepartmental Program, University of California, Los Angeles, CA, USA. 7. Department of Anaesthesiology and Perioperative Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA; Crean College of Health and Behavioural Sciences, Chapman University, Orange, CA, USA; Schmid College of Science and Technology, Chapman University, Orange, CA, USA; Institute for Interdisciplinary Brain and Behavioural Sciences, Chapman University, Orange, CA, USA; Anderson School of Management at UCLA, Los Angeles, CA, USA; Biology and Bioengineering, California Institute of Technology, Pasadena, CA, USA. 8. Department of Anaesthesiology and Perioperative Medicine, University of Pittsburgh, Pittsburgh, PA, USA. 9. Department of Computer Science, University of California, Los Angeles, CA, USA; Department of Human Genetics, University of California, Los Angeles, CA, USA. 10. Department of Anaesthesiology and Perioperative Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA. Electronic address: ihofer@mednet.ucla.edu. 11. Department of Computer Science, University of California, Los Angeles, CA, USA; Department of Anaesthesiology and Perioperative Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA; Department of Human Genetics, University of California, Los Angeles, CA, USA; Department of Biomathematics, University of California, Los Angeles, CA, USA.
Abstract
BACKGROUND: Rapid, preoperative identification of patients with the highest risk for medical complications is necessary to ensure that limited infrastructure and human resources are directed towards those most likely to benefit. Existing risk scores either lack specificity at the patient level or utilise the American Society of Anesthesiologists (ASA) physical status classification, which requires a clinician to review the chart. METHODS: We report on the use of machine learning algorithms, specifically random forests, to create a fully automated score that predicts postoperative in-hospital mortality based solely on structured data available at the time of surgery. Electronic health record data from 53 097 surgical patients (2.01% mortality rate) who underwent general anaesthesia between April 1, 2013 and December 10, 2018 in a large US academic medical centre were used to extract 58 preoperative features. RESULTS: Using a random forest classifier we found that automatically obtained preoperative features (area under the curve [AUC] of 0.932, 95% confidence interval [CI] 0.910-0.951) outperforms Preoperative Score to Predict Postoperative Mortality (POSPOM) scores (AUC of 0.660, 95% CI 0.598-0.722), Charlson comorbidity scores (AUC of 0.742, 95% CI 0.658-0.812), and ASA physical status (AUC of 0.866, 95% CI 0.829-0.897). Including the ASA physical status with the preoperative features achieves an AUC of 0.936 (95% CI 0.917-0.955). CONCLUSIONS: This automated score outperforms the ASA physical status score, the Charlson comorbidity score, and the POSPOM score for predicting in-hospital mortality. Additionally, we integrate this score with a previously published postoperative score to demonstrate the extent to which patient risk changes during the perioperative period.
BACKGROUND: Rapid, preoperative identification of patients with the highest risk for medical complications is necessary to ensure that limited infrastructure and human resources are directed towards those most likely to benefit. Existing risk scores either lack specificity at the patient level or utilise the American Society of Anesthesiologists (ASA) physical status classification, which requires a clinician to review the chart. METHODS: We report on the use of machine learning algorithms, specifically random forests, to create a fully automated score that predicts postoperative in-hospital mortality based solely on structured data available at the time of surgery. Electronic health record data from 53 097 surgical patients (2.01% mortality rate) who underwent general anaesthesia between April 1, 2013 and December 10, 2018 in a large US academic medical centre were used to extract 58 preoperative features. RESULTS: Using a random forest classifier we found that automatically obtained preoperative features (area under the curve [AUC] of 0.932, 95% confidence interval [CI] 0.910-0.951) outperforms Preoperative Score to Predict Postoperative Mortality (POSPOM) scores (AUC of 0.660, 95% CI 0.598-0.722), Charlson comorbidity scores (AUC of 0.742, 95% CI 0.658-0.812), and ASA physical status (AUC of 0.866, 95% CI 0.829-0.897). Including the ASA physical status with the preoperative features achieves an AUC of 0.936 (95% CI 0.917-0.955). CONCLUSIONS: This automated score outperforms the ASA physical status score, the Charlson comorbidity score, and the POSPOM score for predicting in-hospital mortality. Additionally, we integrate this score with a previously published postoperative score to demonstrate the extent to which patient risk changes during the perioperative period.
Authors: Jarrod E Dalton; Andrea Kurz; Alparslan Turan; Edward J Mascha; Daniel I Sessler; Leif Saager Journal: Anesthesiology Date: 2011-06 Impact factor: 7.892
Authors: Karl Y Bilimoria; Yaoming Liu; Jennifer L Paruch; Lynn Zhou; Thomas E Kmiecik; Clifford Y Ko; Mark E Cohen Journal: J Am Coll Surg Date: 2013-09-18 Impact factor: 6.113
Authors: Ira L Leeds; Brindusa Truta; Alyssa M Parian; Sophia Y Chen; Jonathan E Efron; Susan L Gearhart; Bashar Safar; Sandy H Fang Journal: J Gastrointest Surg Date: 2017-08-17 Impact factor: 3.452
Authors: Daniel I Sessler; Jeffrey C Sigl; Paul J Manberg; Scott D Kelley; Armin Schubert; Nassib G Chamoun Journal: Anesthesiology Date: 2010-11 Impact factor: 7.892
Authors: Rupert M Pearse; David A Harrison; Philip James; David Watson; Charles Hinds; Andrew Rhodes; R Michael Grounds; E David Bennett Journal: Crit Care Date: 2006-06-02 Impact factor: 9.097
Authors: Xinyu Yan; Jeff Goldsmith; Sumit Mohan; Zachary A Turnbull; Robert E Freundlich; Frederic T Billings; Ravi P Kiran; Guohua Li; Minjae Kim Journal: Anesth Analg Date: 2022-01-01 Impact factor: 5.108
Authors: Andrew Bishara; Catherine Chiu; Elizabeth L Whitlock; Vanja C Douglas; Sei Lee; Atul J Butte; Jacqueline M Leung; Anne L Donovan Journal: BMC Anesthesiol Date: 2022-01-03 Impact factor: 2.376
Authors: Annick Stolze; Ewoudt M W van de Garde; Linda M Posthuma; Markus W Hollmann; Dianne de Korte-de Boer; Valérie M Smit-Fun; Wolfgang F F A Buhre; Christa Boer; Peter G Noordzij Journal: BMC Anesthesiol Date: 2022-03-03 Impact factor: 2.217