Machine Learning Approach to Predicting COVID-19 Disease Severity Based on Clinical Blood Test Data: Statistical Analysis and Model Development.
COVID-19
blood
blood samples
data set
machine learning
morbidity
mortality
outcome
prediction
risk
severity
statistical analysis
testing
Journal
JMIR medical informatics
ISSN: 2291-9694
Titre abrégé: JMIR Med Inform
Pays: Canada
ID NLM: 101645109
Informations de publication
Date de publication:
13 Apr 2021
13 Apr 2021
Historique:
received:
20
11
2020
accepted:
21
03
2021
revised:
21
01
2021
pubmed:
30
3
2021
medline:
30
3
2021
entrez:
29
3
2021
Statut:
epublish
Résumé
Accurate prediction of the disease severity of patients with COVID-19 would greatly improve care delivery and resource allocation and thereby reduce mortality risks, especially in less developed countries. Many patient-related factors, such as pre-existing comorbidities, affect disease severity and can be used to aid this prediction. Because rapid automated profiling of peripheral blood samples is widely available, we aimed to investigate how data from the peripheral blood of patients with COVID-19 can be used to predict clinical outcomes. We investigated clinical data sets of patients with COVID-19 with known outcomes by combining statistical comparison and correlation methods with machine learning algorithms; the latter included decision tree, random forest, variants of gradient boosting machine, support vector machine, k-nearest neighbor, and deep learning methods. Our work revealed that several clinical parameters that are measurable in blood samples are factors that can discriminate between healthy people and COVID-19-positive patients, and we showed the value of these parameters in predicting later severity of COVID-19 symptoms. We developed a number of analytical methods that showed accuracy and precision scores >90% for disease severity prediction. We developed methodologies to analyze routine patient clinical data that enable more accurate prediction of COVID-19 patient outcomes. With this approach, data from standard hospital laboratory analyses of patient blood could be used to identify patients with COVID-19 who are at high risk of mortality, thus enabling optimization of hospital facilities for COVID-19 treatment.
Sections du résumé
BACKGROUND
BACKGROUND
Accurate prediction of the disease severity of patients with COVID-19 would greatly improve care delivery and resource allocation and thereby reduce mortality risks, especially in less developed countries. Many patient-related factors, such as pre-existing comorbidities, affect disease severity and can be used to aid this prediction.
OBJECTIVE
OBJECTIVE
Because rapid automated profiling of peripheral blood samples is widely available, we aimed to investigate how data from the peripheral blood of patients with COVID-19 can be used to predict clinical outcomes.
METHODS
METHODS
We investigated clinical data sets of patients with COVID-19 with known outcomes by combining statistical comparison and correlation methods with machine learning algorithms; the latter included decision tree, random forest, variants of gradient boosting machine, support vector machine, k-nearest neighbor, and deep learning methods.
RESULTS
RESULTS
Our work revealed that several clinical parameters that are measurable in blood samples are factors that can discriminate between healthy people and COVID-19-positive patients, and we showed the value of these parameters in predicting later severity of COVID-19 symptoms. We developed a number of analytical methods that showed accuracy and precision scores >90% for disease severity prediction.
CONCLUSIONS
CONCLUSIONS
We developed methodologies to analyze routine patient clinical data that enable more accurate prediction of COVID-19 patient outcomes. With this approach, data from standard hospital laboratory analyses of patient blood could be used to identify patients with COVID-19 who are at high risk of mortality, thus enabling optimization of hospital facilities for COVID-19 treatment.
Identifiants
pubmed: 33779565
pii: v9i4e25884
doi: 10.2196/25884
pmc: PMC8045777
doi:
Types de publication
Journal Article
Langues
eng
Pagination
e25884Subventions
Organisme : EPA
ID : EP-C-18-008
Pays : United States
Informations de copyright
©Sakifa Aktar, Md Martuza Ahamad, Md Rashed-Al-Mahfuz, AKM Azad, Shahadat Uddin, AHM Kamal, Salem A Alyami, Ping-I Lin, Sheikh Mohammed Shariful Islam, Julian MW Quinn, Valsamma Eapen, Mohammad Ali Moni. Originally published in JMIR Medical Informatics (http://medinform.jmir.org), 13.04.2021.
Références
Inform Med Unlocked. 2020;21:100449
pubmed: 33102686
Nat Commun. 2020 Oct 27;11(1):5411
pubmed: 33110070
Med (N Y). 2020 Dec 18;1(1):128-138.e3
pubmed: 32838352
Sci Rep. 2020 Mar 12;10(1):4605
pubmed: 32165685
Auton Neurosci. 2001 Jul 20;90(1-2):47-56
pubmed: 11485292
Curr Opin Crit Care. 2012 Dec;18(6):700-6
pubmed: 22954664
J Med Virol. 2020 Sep;92(9):1518-1524
pubmed: 32104917
Clin Chim Acta. 2020 Aug;507:174-180
pubmed: 32339487
Brief Bioinform. 2021 Mar 22;22(2):1254-1266
pubmed: 33024988
Brief Bioinform. 2021 Mar 22;22(2):1451-1465
pubmed: 33611340
Ann Lab Med. 2020 Mar 31;40(5):351-360
pubmed: 32237288
Nat Mach Intell. 2020 Jan;2(1):56-67
pubmed: 32607472
Z Gesundh Wiss. 2020 Apr 19;:1-9
pubmed: 32313806
J Clin Diagn Res. 2017 Sep;11(9):ZC36-ZC39
pubmed: 29207830
Clin Infect Dis. 2020 Jul 28;71(15):762-768
pubmed: 32161940
Heart Rhythm. 2020 Sep;17(9):1434-1438
pubmed: 32535142
Int J Infect Dis. 2020 May;94:128-132
pubmed: 32251805
Brief Bioinform. 2021 Mar 22;22(2):1387-1401
pubmed: 33458761
Brief Bioinform. 2021 Mar 22;22(2):1415-1429
pubmed: 33539530
Expert Syst Appl. 2020 Dec 1;160:113661
pubmed: 32834556
Brief Bioinform. 2021 Mar 22;22(2):1175-1196
pubmed: 32778874
J Clin Epidemiol. 2020 Oct;126:207-216
pubmed: 32712176
Clin Infect Dis. 2020 Mar 16;:
pubmed: 32173725
BMC Med Inform Decis Mak. 2019 Dec 21;19(1):281
pubmed: 31864346
Lancet. 2020 Feb 15;395(10223):507-513
pubmed: 32007143