An Automated Machine Learning-based Model Predicts Postoperative Mortality Using Readily-Extractable Preoperative Electronic Health Record Data

Authors

Brian L. Hill, University of California, Los Angeles
Robert Brown, University of California, Los Angeles
Eilon Gabel, University of California, Los Angeles
Nadav Rakocz, University of California, Los Angeles
Christine Lee, University of California, Los Angeles
Maxime Cannesson, University of California, Los Angeles
Pierre Baldi, University of California, Irvine
Loes Olde Loohuis, University of California, Los Angeles
Ruth Johnson, University of California, Los Angeles
Brandon Jew, University of California, Los Angeles
Uri Maoz, Chapman UniversityFollow
Aman Mahajan, University of Pittsburgh
Sriran Sankararaman, University of California, Los Angeles
Ira Hofer, University of California, Los Angeles
Eran Halperin, University of California, Los Angeles

Document Type

Article

Publication Date

10-15-2019

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.

Comments

This article was originally published in British Journal of Anaesthesia, volume 123, issue 6, in 2019. https://doi.org/10.1016/j.bja.2019.07.030

Recommended Citation

Hill BL, Brown R, Gabel E, et al. An automated machine learning-based model predicts postoperative mortality using readily-extractable preoperative electronic health record data. Br J Anaesth. 2019;123(6):877-886. https://doi.org/10.1016/j.bja.2019.07.030

Copyright

Elsevier