Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Sep 16;11(1):18464.
doi: 10.1038/s41598-021-97990-1.

Early outcome detection for COVID-19 patients

Alina Sîrbu  1 , Greta Barbieri  2 , Francesco Faita  3 , Paolo Ferragina  4 , Luna Gargani  3 , Lorenzo Ghiadoni  2 , Corrado Priami  4
Affiliations

Early outcome detection for COVID-19 patients

Alina Sîrbu et al. Sci Rep. .

Abstract

With the outbreak of COVID-19 exerting a strong pressure on hospitals and health facilities, clinical decision support systems based on predictive models can help to effectively improve the management of the pandemic. We present a method for predicting mortality for COVID-19 patients. Starting from a large number of clinical variables, we select six of them with largest predictive power, using a feature selection method based on genetic algorithms and starting from a set of COVID-19 patients from the first wave. The algorithm is designed to reduce the impact of missing values in the set of variables measured, and consider only variables that show good accuracy on validation data. The final predictive model provides accuracy larger than 85% on test data, including a new patient cohort from the second COVID-19 wave, and on patients with imputed missing values. The selected clinical variables are confirmed to be relevant by recent literature on COVID-19.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
Selected clinical variables. Comparison of values over the two classes of patients.
Figure 2
Figure 2
Clinical data and missing values. The plots show the frequency histogram for the distribution of the number of clinical variables available per patient and patients per clinical variable.
Figure 3
Figure 3
Distribution of a selection of measured and imputed clinical variables for discharged and deceased patients. The plots compare the values measured at hospitalisation with those imputed during our analysis. The p values correspond to two-sample Kolmogorov–Smirnov tests, comparing discharged versus discharged imputed and deceased versus deceased imputed. The labels on the x-axis show in parentheses the number of patients in each group.
Figure 4
Figure 4
Decision tree trained on all 185 patients that have no missing values among the 6 selected clinical variables. Each node of the tree contains a condition on a clinical variable. We also include the number of patients, and the way they are divided into the two classes (discharged/deceased). The colour of the node shows whether patients are in the discharged class (orange) or deceased class (blue).
Figure 5
Figure 5
GA feature selection methodology. The result is a ranking of clinical variables, which is then validated on an independent set of patients.
See this image and copyright information in PMC

References

    1. World Health Organisation. Who coronavirus disease (covid-19) dashboard. Accessed 24th September 2020 (2020).
    1. Nishiga M, Wang DW, Han Y, Lewis DB, Wu JC. Covid-19 and cardiovascular disease: From basic mechanisms to clinical perspectives. Nat. Rev. Cardiol. 2020;17:1–16. doi: 10.1038/s41569-020-0413-9. - DOI - PMC - PubMed
    1. Gupta A, et al. Extrapulmonary manifestations of covid-19. Nat. Med. 2020;26(7):1017–1032. doi: 10.1038/s41591-020-0968-3. - DOI - PMC - PubMed
    1. Zheng Z, et al. Risk factors of critical & mortal covid-19 cases: A systematic literature review and meta-analysis. J. Infect. 2020;81:e16–e25. doi: 10.1016/j.jinf.2020.04.021. - DOI - PMC - PubMed
    1. Liu F, et al. Clinical characteristics and corticosteroids application of different clinical types in patients with corona virus disease 2019. Sci. Rep. 2020;10(1):1–9. doi: 10.1038/s41598-019-56847-4. - DOI - PMC - PubMed