Machine Learning to Predict Outcomes and Cost by Phase of Care After Coronary Artery Bypass Grafting.
Journal
The Annals of thoracic surgery
ISSN: 1552-6259
Titre abrégé: Ann Thorac Surg
Pays: Netherlands
ID NLM: 15030100R
Informations de publication
Date de publication:
09 2022
09 2022
Historique:
received:
31
01
2021
revised:
09
08
2021
accepted:
16
08
2021
pubmed:
29
9
2021
medline:
1
9
2022
entrez:
28
9
2021
Statut:
ppublish
Résumé
Machine learning may enhance prediction of outcomes after coronary artery bypass grafting (CABG). We sought to develop and validate a dynamic machine learning model to predict CABG outcomes at clinically relevant pre- and postoperative time points. The Society of Thoracic Surgeons (STS) registry data elements from 2086 isolated CABG patients were divided into training and testing datasets and input into Extreme Gradient Boosting decision-tree machine learning algorithms. Two prediction models were developed based on data from preoperative (80 parameters) and postoperative (125 parameters) phases of care. Outcomes included operative mortality, major morbidity or mortality, high cost, and 30-day readmission. Machine learning and STS model performance were assessed using accuracy and the area under the precision-recall curve (AUC-PR). Preoperative machine learning models predicted mortality (accuracy, 98%; AUC-PR = 0.16; F1 = 0.24), major morbidity or mortality (accuracy, 75%; AUC-PR = 0.33; F1 = 0.42), high cost (accuracy, 83%; AUC-PR = 0.51; F1 = 0.52), and 30-day readmission (accuracy, 70%; AUC-PR = 0.47; F1 = 0.49) with high accuracy. Preoperative machine learning models performed similarly to the STS for prediction of mortality (STS AUC-PR = 0.11; P = .409) and outperformed STS for prediction of mortality or major morbidity (STS AUC-PR = 0.28; P < .001). Addition of intraoperative parameters further improved machine learning model performance for major morbidity or mortality (AUC-PR = 0.39; P < .01) and high cost (AUC-PR = 0.64; P < .01), with cross-clamp and bypass times emerging as important additive predictive parameters. Machine learning can predict mortality, major morbidity, high cost, and readmission after isolated CABG. Prediction based on the phase of care allows for dynamic risk assessment through the hospital course, which may benefit quality assessment and clinical decision-making.
Sections du résumé
BACKGROUND
Machine learning may enhance prediction of outcomes after coronary artery bypass grafting (CABG). We sought to develop and validate a dynamic machine learning model to predict CABG outcomes at clinically relevant pre- and postoperative time points.
METHODS
The Society of Thoracic Surgeons (STS) registry data elements from 2086 isolated CABG patients were divided into training and testing datasets and input into Extreme Gradient Boosting decision-tree machine learning algorithms. Two prediction models were developed based on data from preoperative (80 parameters) and postoperative (125 parameters) phases of care. Outcomes included operative mortality, major morbidity or mortality, high cost, and 30-day readmission. Machine learning and STS model performance were assessed using accuracy and the area under the precision-recall curve (AUC-PR).
RESULTS
Preoperative machine learning models predicted mortality (accuracy, 98%; AUC-PR = 0.16; F1 = 0.24), major morbidity or mortality (accuracy, 75%; AUC-PR = 0.33; F1 = 0.42), high cost (accuracy, 83%; AUC-PR = 0.51; F1 = 0.52), and 30-day readmission (accuracy, 70%; AUC-PR = 0.47; F1 = 0.49) with high accuracy. Preoperative machine learning models performed similarly to the STS for prediction of mortality (STS AUC-PR = 0.11; P = .409) and outperformed STS for prediction of mortality or major morbidity (STS AUC-PR = 0.28; P < .001). Addition of intraoperative parameters further improved machine learning model performance for major morbidity or mortality (AUC-PR = 0.39; P < .01) and high cost (AUC-PR = 0.64; P < .01), with cross-clamp and bypass times emerging as important additive predictive parameters.
CONCLUSIONS
Machine learning can predict mortality, major morbidity, high cost, and readmission after isolated CABG. Prediction based on the phase of care allows for dynamic risk assessment through the hospital course, which may benefit quality assessment and clinical decision-making.
Identifiants
pubmed: 34582751
pii: S0003-4975(21)01649-0
doi: 10.1016/j.athoracsur.2021.08.040
pmc: PMC9703607
mid: NIHMS1754047
pii:
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
711-719Subventions
Organisme : NHLBI NIH HHS
ID : T32 HL139430
Pays : United States
Informations de copyright
Copyright © 2022 The Society of Thoracic Surgeons. Published by Elsevier Inc. All rights reserved.
Références
Platelets. 2019;30(4):480-486
pubmed: 29676943
J Thorac Cardiovasc Surg. 2022 Jun;163(6):2075-2087.e9
pubmed: 32900480
J Thorac Cardiovasc Surg. 2022 Sep;164(3):943-955.e7
pubmed: 33127082
Interact Cardiovasc Thorac Surg. 2014 Jul;19(1):21-6
pubmed: 24722513
Ann Thorac Surg. 2019 Jun;107(6):1782-1789
pubmed: 30553740
Ann Thorac Surg. 2020 May;109(5):1323-1329
pubmed: 31706869
Cardiol Rev. 2020 Mar/Apr;28(2):53-64
pubmed: 32022759
Springerplus. 2013 May 14;2(1):222
pubmed: 23853744
Ann Thorac Surg. 1997 Jun;63(6):1635-43
pubmed: 9205161
Eur J Cardiothorac Surg. 2012 Apr;41(4):734-44; discussion 744-5
pubmed: 22378855
Lancet Digit Health. 2020 Apr;2(4):e179-e191
pubmed: 33328078
Ann Thorac Surg. 2016 Aug;102(2):573-9
pubmed: 27112651
J Thorac Cardiovasc Surg. 2022 Jun;163(6):2088-2090
pubmed: 33051071
Ann Thorac Surg. 2020 Jun;109(6):1811-1819
pubmed: 31706872
J Cardiothorac Vasc Anesth. 2008 Dec;22(6):814-22
pubmed: 18948034
Ann Thorac Surg. 2021 Feb;111(2):488-494
pubmed: 32585200
Ann Thorac Surg. 2018 May;105(5):1419-1428
pubmed: 29577924
Ann Thorac Surg. 2020 Jan;109(1):10-13
pubmed: 31758929
J Thorac Dis. 2021 Feb;13(2):735-742
pubmed: 33717545
Eur Heart J Qual Care Clin Outcomes. 2018 Oct 1;4(4):246-257
pubmed: 29939246
J Thorac Cardiovasc Surg. 2019 Oct;158(4):1120-1126
pubmed: 31537256