An artificial intelligence approach for prognosis of COVID-19 course in hospitalized patients-Reference-Cited by-同舟云学术

An artificial intelligence approach for prognosis of COVID-19 course in hospitalized patients

Published:2023-10-13 Issue:3 Volume:15 Page:60-66
ISSN:2072-6732
Container-title:Journal Infectology
language:
Short-container-title:jour

Author:

Karonova T. L.¹,Korsakov I. N.¹,Mikhailova A. A.¹,Lagutina D. I.¹,Chernikova A. T.¹,Vashukova М. A.²,Smolnikova M. A.¹,Gusev D. A.²,Konradi A. O.¹,Shlyakhto E. V.¹

Affiliation:

1. National Medical Research Centre named after V.A. Almazov

2. National Medical Research Centre named after V.A. Almazov; Clinical Infectious Hospital named after S.P. Botkin

Abstract

Aim. To create algorithm and risk calculator for predicting the lethal outcome in patients with COVID-19.Materials and methods. Based on machine learning approach mortality risk calculator was developed in Almazov National Medical Research Centre using data of the hospitalised patients with an established diagnosis of COVID-19 (n=4071).Results. This mathematical model, which includes 11 significant features, has been proposed for estimation of fatal outcomes in the Clinical Infectious Hospital named after S.P. Botkin. Some key features were not assessed in most hospitals according to accepted standards of care for COVID-19. So systematic analysis of factors affecting the course of disease in patients (n=2876) were conducted and «urea» and «total protein» were replaced with «sex» and «BMI». Modified algorithm demonstrated high sensitivity and specificity. Conclusion. This calculator is able to predict hospitalisation outcome with high accuracy in patients infected with different strains of SARS-CoV-2. This decision support system may be used for risk stratification and following correct patients routing.

Publisher

SPRIDA

Subject

Infectious Diseases

Reference16 articles.

1. Stopkoronavirus: Oficial’naya informaciya o koronaviruse v Rossii: [Elektronnyj resurs]. URL: https://stopkoronavirus.rf/information. (Data obrashcheniya: 26.04.2023).

2. Smirnov A.YU. Analiz smertnosti ot koronavirusnoj infekcii v Rossii // Narodonaselenie. – 2021 – T. 24 – № 2 – C. 76-86.

3. Estestvennoe dvizhenie naseleniya v razreze sub”ektov Rossijskoj Federacii za aprel’ 2023 goda: [Elektronnyj resurs] // Rosstat. URL: https://rosstat.gov.ru/storage/mediabank/EDN_04-2023.html. (Data obrashcheniya: 26.04.2023).

4. Haldane V. et al. Health systems resilience in managing the COVID-19 pandemic: lessons from 28 countries //Nature Medicine. – 2021. – Т. 27. – №. 6. – С. 964-980.

5. Parra P.N.B, Atanasov V.A., Meurer J., et al. Data and Methodology Documentation for a COVID-19 Risk Calculator: Mortality Rates and Loss of Life Expectancy. SSRN; 2021; doi: 10.2139/ssrn.3795802.