Machine learning analysis to identify the association between risk factors and onset of nosocomial diarrhea: a retrospective cohort study.
Antibiotics
Cerebral infarction
Machine learning
Nosocomial diarrhoea
Random forest
Tube feeding
Journal
PeerJ
ISSN: 2167-8359
Titre abrégé: PeerJ
Pays: United States
ID NLM: 101603425
Informations de publication
Date de publication:
2019
2019
Historique:
received:
24
06
2019
accepted:
01
10
2019
entrez:
6
11
2019
pubmed:
7
11
2019
medline:
7
11
2019
Statut:
epublish
Résumé
Although several risk factors for nosocomial diarrhea have been identified, the detail of association between these factors and onset of nosocomial diarrhea, such as degree of importance or temporal pattern of influence, remains unclear. We aimed to determine the association between risk factors and onset of nosocomial diarrhea using machine learning algorithms. We retrospectively collected data of patients with acute cerebral infarction. Seven variables, including age, sex, modified Rankin Scale (mRS) score, and number of days of antibiotics, tube feeding, proton pump inhibitors, and histamine 2-receptor antagonist use, were used in the analysis. We split the data into a training dataset and independant test dataset. Based on the training dataset, we developed a random forest, support vector machine (SVM), and radial basis function (RBF) network model. By calculating an area under the curve (AUC) of the receiver operating characteristic curve using 5-fold cross-validation, we performed feature selection and hyperparameter optimization in each model. According to their final performances, we selected the optimal model and also validated it in the independent test dataset. Based on the selected model, we visualized the variable importance and the association between each variable and the outcome using partial dependence plots. Two-hundred and eighteen patients were included. In the cross-validation within the training dataset, the random forest model achieved an AUC of 0.944, which was higher than in the SVM and RBF network models. The random forest model also achieved an AUC of 0.832 in the independent test dataset. Tube feeding use days, mRS score, antibiotic use days, age and sex were strongly associated with the onset of nosocomial diarrhea, in this order. Tube feeding use had an inverse We revealed the degree of importance and temporal pattern of the influence of several risk factors for nosocomial diarrhea, which could help clinicians manage nosocomial diarrhea.
Sections du résumé
BACKGROUND
BACKGROUND
Although several risk factors for nosocomial diarrhea have been identified, the detail of association between these factors and onset of nosocomial diarrhea, such as degree of importance or temporal pattern of influence, remains unclear. We aimed to determine the association between risk factors and onset of nosocomial diarrhea using machine learning algorithms.
METHODS
METHODS
We retrospectively collected data of patients with acute cerebral infarction. Seven variables, including age, sex, modified Rankin Scale (mRS) score, and number of days of antibiotics, tube feeding, proton pump inhibitors, and histamine 2-receptor antagonist use, were used in the analysis. We split the data into a training dataset and independant test dataset. Based on the training dataset, we developed a random forest, support vector machine (SVM), and radial basis function (RBF) network model. By calculating an area under the curve (AUC) of the receiver operating characteristic curve using 5-fold cross-validation, we performed feature selection and hyperparameter optimization in each model. According to their final performances, we selected the optimal model and also validated it in the independent test dataset. Based on the selected model, we visualized the variable importance and the association between each variable and the outcome using partial dependence plots.
RESULTS
RESULTS
Two-hundred and eighteen patients were included. In the cross-validation within the training dataset, the random forest model achieved an AUC of 0.944, which was higher than in the SVM and RBF network models. The random forest model also achieved an AUC of 0.832 in the independent test dataset. Tube feeding use days, mRS score, antibiotic use days, age and sex were strongly associated with the onset of nosocomial diarrhea, in this order. Tube feeding use had an inverse
CONCLUSION
CONCLUSIONS
We revealed the degree of importance and temporal pattern of the influence of several risk factors for nosocomial diarrhea, which could help clinicians manage nosocomial diarrhea.
Identifiants
pubmed: 31687281
doi: 10.7717/peerj.7969
pii: 7969
pmc: PMC6825409
doi:
Types de publication
Journal Article
Langues
eng
Pagination
e7969Informations de copyright
©2019 Kurisu et al.
Déclaration de conflit d'intérêts
The authors declare there are no competing interests.
Références
Comput Math Methods Med. 2015;2015:814104
pubmed: 26089977
mSphere. 2018 Nov 21;3(6):
pubmed: 30463925
JAMA. 2001 Jan 17;285(3):313-9
pubmed: 11176841
J Infect Chemother. 2018 Feb;24(2):88-91
pubmed: 28974364
Front Psychiatry. 2020 Feb 20;10:1029
pubmed: 32153432
JAMA Netw Open. 2018 Aug 3;1(4):e181405
pubmed: 30646122
Sci Rep. 2019 Mar 7;9(1):3791
pubmed: 30846783
BMC Public Health. 2014 May 13;14:451
pubmed: 24885464
J Infect Chemother. 2017 Oct;23(10):674-677
pubmed: 28751156
Med Biol Eng Comput. 2017 Jan;55(1):151-165
pubmed: 27106758
Infect Dis Ther. 2018 Mar;7(1):39-70
pubmed: 29441500
J Pain. 2013 Dec;14(12 Suppl):T102-15
pubmed: 24275218
Stroke. 1993 Jan;24(1):35-41
pubmed: 7678184
PeerJ. 2018 Nov 22;6:e5982
pubmed: 30498643
Anaerobe. 2019 Oct;59:126-130
pubmed: 31254655
Anal Biochem. 2018 Aug 15;555:33-41
pubmed: 29908156
Intern Med J. 2014 Dec;44(12a):1199-204
pubmed: 25228255
Lancet Infect Dis. 2017 Sep;17(9):990-1001
pubmed: 28629876
Infect Control Hosp Epidemiol. 2002 Nov;23(11):653-9
pubmed: 12452292
J Hosp Infect. 2018 Jun;99(2):133-138
pubmed: 29325870
J Infect Chemother. 2011 Dec;17(6):807-11
pubmed: 21725661
J Comput Chem. 2017 Sep 5;38(23):2000-2006
pubmed: 28643394
Anal Biochem. 2019 Apr 15;571:53-61
pubmed: 30822398
J Adv Nurs. 2008 May;62(3):354-64
pubmed: 18426460
Clin Infect Dis. 2018 Mar 19;66(7):987-994
pubmed: 29562266
Stroke. 2001 Dec 1;32(12):2735-40
pubmed: 11739965
Cochrane Database Syst Rev. 2017 Dec 19;12:CD006095
pubmed: 29257353
PeerJ. 2018 Oct 17;6:e5714
pubmed: 30357023
Clin Infect Dis. 2002 Feb 1;34(3):346-53
pubmed: 11774082
Scand J Gastroenterol. 1997 Sep;32(9):920-4
pubmed: 9299672
Am J Infect Control. 1995 Oct;23(5):295-305
pubmed: 8585641
Stroke. 1988 May;19(5):604-7
pubmed: 3363593
J Glob Health. 2017 Jun;7(1):010417
pubmed: 28607673
Pain Med. 2017 Jan 1;18(1):107-115
pubmed: 27252307
Front Neurol. 2019 Apr 24;10:401
pubmed: 31068892
Anaerobe. 2019 Dec;60:102011
pubmed: 30872073
J Cancer Res Ther. 2018 Apr-Jun;14(3):625-633
pubmed: 29893330
Exerc Sport Sci Rev. 2019 Apr;47(2):75-85
pubmed: 30883471
JAMA. 2012 May 9;307(18):1959-69
pubmed: 22570464
Anal Biochem. 2019 Jun 15;575:17-26
pubmed: 30930199