A multipurpose machine learning approach to predict COVID-19 negative prognosis in São Paulo, Brazil

Página do item simplificado
dc.contributor.authorFernandes, Fernando Timoteo
dc.contributor.authorOliveira, Tiago Almeida de
dc.contributor.authorTeixeira, Cristiane Esteves
dc.contributor.authorANDRE FILIPE DE MORAES BATISTA
dc.contributor.authorCosta, Gabriel Dalla
dc.contributor.authorChiavegatto Filho, Alexandre Dias Porto
dc.creatorFernandes, Fernando Timoteo
dc.creatorOliveira, Tiago Almeida de
dc.creatorTeixeira, Cristiane Esteves
dc.creatorCosta, Gabriel Dalla
dc.creatorChiavegatto Filho, Alexandre Dias Porto
dc.date.accessioned2024-11-21T18:31:11Z
dc.date.available2024-11-21T18:31:11Z
dc.date.issued2021
dc.description.abstractThe new coronavirus disease (COVID-19) is a challenge for clinical decision-making and the effective allocation of healthcare resources. An accurate prognostic assessment is necessary to improve survival of patients, especially in developing countries. This study proposes to predict the risk of developing critical conditions in COVID-19 patients by training multipurpose algorithms. We followed a total of 1040 patients with a positive RT-PCR diagnosis for COVID-19 from a large hospital from São Paulo, Brazil, from March to June 2020, of which 288 (28%) presented a severe prognosis, i.e. Intensive Care Unit (ICU) admission, use of mechanical ventilation or death. We used routinely-collected laboratory, clinical and demographic data to train five machine learning algorithms (artificial neural networks, extra trees, random forests, catboost, and extreme gradient boosting). We used a random sample of 70% of patients to train the algorithms and 30% were left for performance assessment, simulating new unseen data. In order to assess if the algorithms could capture general severe prognostic patterns, each model was trained by combining two out of three outcomes to predict the other. All algorithms presented very high predictive performance (average AUROC of 0.92, sensitivity of 0.92, and specificity of 0.82). The three most important variables for the multipurpose algorithms were ratio of lymphocyte per C-reactive protein, C-reactive protein and Braden Scale. The results highlight the possibility that machine learning algorithms are able to predict unspecific negative COVID-19 outcomes from routinely-collected data.en
dc.formatDigital
dc.format.extent7 p.
dc.identifier.doi10.1038/s41598-021-82885-y
dc.identifier.urihttps://repositorio.insper.edu.br/handle/11224/7232
dc.language.isoInglês
dc.relation.ispartofScientific Reports
dc.titleA multipurpose machine learning approach to predict COVID-19 negative prognosis in São Paulo, Brazil
dc.typejournal article
dspace.entity.typePublication
local.identifier.sourceUrihttps://www.nature.com/articles/s41598-021-82885-y
local.publisher.countryNão Informado
local.subject.cnpqCIENCIAS DA SAUDE::MEDICINA
local.subject.cnpqCIENCIAS DA SAUDE::SAUDE COLETIVA::EPIDEMIOLOGIA
local.subject.cnpqCIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA
local.subject.cnpqCIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
local.subject.cnpqENGENHARIAS::ENGENHARIA BIOMEDICA
local.typeArtigo Científico
publicationvolume.volumeNumber11
relation.isAuthorOfPublicationb10d272e-98b2-4953-8e51-37aea3fde20c
relation.isAuthorOfPublication.latestForDiscoveryb10d272e-98b2-4953-8e51-37aea3fde20c
Arquivos
Pacote Original
Agora exibindo 1 - 2 de 2
Imagem de Miniatura
Nome:
Primeira_Pagina_Artigo_2021_A_multipurpose_machine_learning_approach_to_predict_COVID_19_negative_prognosis_in_Sao_Paulo_Brazil_TC.pdf
Tamanho:
303.43 KB
Formato:
Adobe Portable Document Format
Baixar
N/D
Nome:
ACESSO_RESTRITO_Artigo_2021_A_multipurpose_machine_learning_approach_to_predict_COVID_19_negative_prognosis_in_Sao_Paulo_Brazil_TC.pdf
Tamanho:
1.08 MB
Formato:
Adobe Portable Document Format
Baixar
Licença do Pacote
Agora exibindo 1 - 1 de 1
N/D
Nome:
license.txt
Tamanho:
236 B
Formato:
Item-specific license agreed upon to submission
Descrição:
Baixar
Coleções
Coleção de Artigos Acadêmicos