Please use this identifier to cite or link to this item:
Title: Value of laboratory results in addition to vital signs in a machine learning algorithm to predict in-hospital cardiac arrest: A single-center retrospective cohort study.
Austin Authors: Ueno, Ryo;Xu, Liyuan;Uegami, Wataru;Matsui, Hiroki;Okui, Jun;Hayashi, Hiroshi;Miyajima, Toru;Hayashi, Yoshiro;Pilcher, David;Jones, Daryl A 
Affiliation: Anatomical Pathology, Kameda Medical Center, Chiba, Japan
Department of Computer Science, Graduate School of Information Science and Technology, The University of Tokyo, Tokyo, Japan
Department of Intensive Care Medicine, Kameda Medical Center, Chiba, Japan
Australian and New Zealand Intensive Care Research Center, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia
Department of Intensive Care, Austin Health, Heidelberg, Victoria, Australia
Clinical Research Support Division, Kameda Medical Center, Chiba, Japan
Post-Graduate Education Center, Kameda Medical Center, Chiba, Japan
Department of Intensive Care Medicine, Kameda Medical Center, Chiba, Japan
Issue Date: 13-Jul-2020
Date: 2020-07-13
Publication information: PLoS One 2020; 15(7): e0235835
Abstract: Although machine learning-based prediction models for in-hospital cardiac arrest (IHCA) have been widely investigated, it is unknown whether a model based on vital signs alone (Vitals-Only model) can perform similarly to a model that considers both vital signs and laboratory results (Vitals+Labs model). All adult patients hospitalized in a tertiary care hospital in Japan between October 2011 and October 2018 were included in this study. Random forest models with/without laboratory results (Vitals+Labs model and Vitals-Only model, respectively) were trained and tested using chronologically divided datasets. Both models use patient demographics and eight-hourly vital signs collected within the previous 48 hours. The primary and secondary outcomes were the occurrence of IHCA in the next 8 and 24 hours, respectively. The area under the receiver operating characteristic curve (AUC) was used as a comparative measure. Sensitivity analyses were performed under multiple statistical assumptions. Of 141,111 admitted patients (training data: 83,064, test data: 58,047), 338 had an IHCA (training data: 217, test data: 121) during the study period. The Vitals-Only model and Vitals+Labs model performed comparably when predicting IHCA within the next 8 hours (Vitals-Only model vs Vitals+Labs model, AUC = 0.862 [95% confidence interval (CI): 0.855-0.868] vs 0.872 [95% CI: 0.867-0.878]) and 24 hours (Vitals-Only model vs Vitals+Labs model, AUC = 0.830 [95% CI: 0.825-0.835] vs 0.837 [95% CI: 0.830-0.844]). Both models performed similarly well on medical, surgical, and ward patient data, but did not perform well for intensive care unit patients. In this single-center study, the machine learning model predicted IHCAs with good discrimination. The addition of laboratory values to vital signs did not significantly improve its overall performance.
DOI: 10.1371/journal.pone.0235835
ORCID: 0000-0003-1681-0107
Journal: PLoS One
PubMed URL: 32658901
Type: Journal Article
Appears in Collections:Journal articles

Show full item record

Page view(s)

checked on Nov 29, 2023

Google ScholarTM


Items in AHRO are protected by copyright, with all rights reserved, unless otherwise indicated.