Clinically applicable approach for predicting mechanical ventilation in patients with COVID-19.
COVID-19
critical care medicine
machine learning
mechanical ventilation
predictive models
respiratory failure
respiratory insufficiency
Journal
British journal of anaesthesia
ISSN: 1471-6771
Titre abrégé: Br J Anaesth
Pays: England
ID NLM: 0372541
Informations de publication
Date de publication:
03 2021
03 2021
Historique:
received:
16
09
2020
revised:
19
10
2020
accepted:
03
11
2020
pubmed:
18
1
2021
medline:
3
3
2021
entrez:
17
1
2021
Statut:
ppublish
Résumé
Patients with coronavirus disease 2019 (COVID-19) requiring mechanical ventilation have high mortality and resource utilisation. The ability to predict which patients may require mechanical ventilation allows increased acuity of care and targeted interventions to potentially mitigate deterioration. We included hospitalised patients with COVID-19 in this single-centre retrospective observational study. Our primary outcome was mechanical ventilation or death within 24 h. As clinical decompensation is more recognisable, but less modifiable, as the prediction window shrinks, we also assessed 4, 8, and 48 h prediction windows. Model features included demographic information, laboratory results, comorbidities, medication administration, and vital signs. We created a Random Forest model, and assessed performance using 10-fold cross-validation. The model was compared with models derived from generalised estimating equations using discrimination. Ninety-three (23%) of 398 patients required mechanical ventilation or died within 14 days of admission. The Random Forest model predicted pending mechanical ventilation with good discrimination (C-statistic=0.858; 95% confidence interval, 0.841-0.874), which is comparable with the discrimination of the generalised estimating equation regression. Vitals sign data including SpO Machine learning techniques can be leveraged to improve the ability to predict which patients with COVID-19 are likely to require mechanical ventilation, identifying unrecognised bellwethers and providing insight into the constellation of accompanying signs of respiratory failure in COVID-19.
Sections du résumé
BACKGROUND
Patients with coronavirus disease 2019 (COVID-19) requiring mechanical ventilation have high mortality and resource utilisation. The ability to predict which patients may require mechanical ventilation allows increased acuity of care and targeted interventions to potentially mitigate deterioration.
METHODS
We included hospitalised patients with COVID-19 in this single-centre retrospective observational study. Our primary outcome was mechanical ventilation or death within 24 h. As clinical decompensation is more recognisable, but less modifiable, as the prediction window shrinks, we also assessed 4, 8, and 48 h prediction windows. Model features included demographic information, laboratory results, comorbidities, medication administration, and vital signs. We created a Random Forest model, and assessed performance using 10-fold cross-validation. The model was compared with models derived from generalised estimating equations using discrimination.
RESULTS
Ninety-three (23%) of 398 patients required mechanical ventilation or died within 14 days of admission. The Random Forest model predicted pending mechanical ventilation with good discrimination (C-statistic=0.858; 95% confidence interval, 0.841-0.874), which is comparable with the discrimination of the generalised estimating equation regression. Vitals sign data including SpO
CONCLUSION
Machine learning techniques can be leveraged to improve the ability to predict which patients with COVID-19 are likely to require mechanical ventilation, identifying unrecognised bellwethers and providing insight into the constellation of accompanying signs of respiratory failure in COVID-19.
Identifiants
pubmed: 33454051
pii: S0007-0912(20)30958-2
doi: 10.1016/j.bja.2020.11.034
pmc: PMC7833820
pii:
doi:
Types de publication
Journal Article
Observational Study
Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
578-589Subventions
Organisme : NHLBI NIH HHS
ID : K01 HL141701
Pays : United States
Informations de copyright
Copyright © 2020 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved.
Références
Chest. 2015 Dec;148(6):1477-1483
pubmed: 26271028
Thromb Haemost. 2020 Jun;120(6):998-1000
pubmed: 32316063
N Engl J Med. 2020 Feb 20;382(8):727-733
pubmed: 31978945
Respirology. 2014 May;19(4):538-43
pubmed: 24661379
Lancet. 2020 Jul 4;396(10243):27-38
pubmed: 32479829
J Med Internet Res. 2016 Dec 16;18(12):e323
pubmed: 27986644
BMJ. 2020 May 22;369:m1966
pubmed: 32444366
Clin Infect Dis. 2020 Sep 12;71(6):1393-1399
pubmed: 32271369
J Infect. 2020 Jun;80(6):639-645
pubmed: 32240670
Allergy. 2020 Jul;75(7):1730-1741
pubmed: 32077115
Chin Med J (Engl). 2020 Jun 5;133(11):1261-1267
pubmed: 32209890
Lancet Respir Med. 2020 May;8(5):475-481
pubmed: 32105632
BMC Anesthesiol. 2018 Feb 5;18(1):16
pubmed: 29402220
Ann Intern Med. 2020 Oct 20;173(8):623-631
pubmed: 32673060
JAMA Intern Med. 2020 Aug 1;180(8):1081-1089
pubmed: 32396163
JAMA. 2020 May 26;323(20):2052-2059
pubmed: 32320003
Lancet. 2020 Mar 28;395(10229):1054-1062
pubmed: 32171076
Clin Infect Dis. 2020 Jul 28;71(15):833-840
pubmed: 32296824
PLoS One. 2020 Aug 11;15(8):e0237419
pubmed: 32780765
Anesth Analg. 2020 Jan;130(1):165-175
pubmed: 31107262
Am J Emerg Med. 2020 Jul;38(7):1504-1507
pubmed: 32317203
Med Care. 2005 Nov;43(11):1130-9
pubmed: 16224307
Crit Care Med. 2009 Apr;37(4):1317-21
pubmed: 19242333
JAMA Neurol. 2020 Jun 1;77(6):683-690
pubmed: 32275288