Search

Introduction: Optimized allocation of medical resources to patients with COVID-19 has been a critical concern since the onset of the pandemic., Methods: In this retrospective cohort study, the authors used data from a Brazilian tertiary university hospital to explore predictors of Intensive Care Unit (ICU) admission and hospital mortality in patients admitted for COVID-19. Our primary aim was to create and validate prediction scores for use in hospitals and emergency departments to aid clinical decisions and resource allocation., Results: The study cohort included 3,022 participants, of whom 2,485 were admitted to the ICU; 1968 survived, and 1054 died in the hospital. From the complete cohort, 1,496 patients were randomly assigned to the derivation sample and 1,526 to the validation sample. The final scores included age, comorbidities, and baseline laboratory data. The areas under the receiver operating characteristic curves were very similar for the derivation and validation samples. Scores for ICU admission had a 75% accuracy in the validation sample, whereas scores for death had a 77% accuracy in the validation sample. The authors found that including baseline flu-like symptoms in the scores added no significant benefit to their accuracy. Furthermore, our scores were more accurate than the previously published NEWS-2 and 4C Mortality Scores., Discussion and Conclusions: The authors developed and validated prognostic scores that use readily available clinical and laboratory information to predict ICU admission and mortality in COVID-19. These scores can become valuable tools to support clinical decisions and improve the allocation of limited health resources., Competing Interests: Conflicts of interest The authors declare no conflicts of interest., (Copyright © 2023 HCFMUSP. Published by Elsevier España, S.L.U. All rights reserved.)

Background: COVID-19 caused more than 622 thousand deaths in Brazil. The infection can be asymptomatic and cause mild symptoms, but it also can evolve into a severe disease and lead to death. It is difficult to predict which patients will develop severe disease. There are, in the literature, machine learning models capable of assisting diagnose and predicting outcomes for several diseases, but usually these models require laboratory tests and/or imaging., Methods: We conducted a observational cohort study that evaluated vital signs and measurements from patients who were admitted to Hospital das Clínicas (São Paulo, Brazil) between March 2020 and October 2021 due to COVID-19. The data was then represented as univariate and multivariate time series, that were used to train and test machine learning models capable of predicting a patient's outcome., Results: Time series-based machine learning models are capable of predicting a COVID-19 patient's outcome with up to 96% general accuracy and 81% accuracy considering only the first hospitalization day. The models can reach up to 99% sensitivity (discharge prediction) and up to 91% specificity (death prediction)., Conclusions: Results indicate that time series-based machine learning models combined with easily obtainable data can predict COVID-19 outcomes and support clinical decisions. With further research, these models can potentially help doctors diagnose other diseases., (© 2022. The Author(s).)