Development and validation of an early warning score (EWAS) for predicting clinical deterioration in patients with coronavirus disease 2019

  1. Yabing Guo
  2. Yingxia Liu
  3. Jiatao Lu
  4. Rong Fan
  5. Fuchun Zhang
  6. Xueru Yin
  7. Zhihong Liu
  8. Qinglang Zeng
  9. Jing Yuan
  10. Shufang Hu
  11. Qiongya Wang
  12. Baolin Liao
  13. Mingxing Huang
  14. Sichun Yin
  15. Xilin Zhang
  16. Rui Xin
  17. Zhanzhou Lin
  18. Changzheng Hu
  19. Boliang Zhao
  20. Ridong He
  21. Minfeng Liang
  22. Zheng Zhang
  23. Li Liu
  24. Jian Sun
  25. Lu Tang
  26. Lisi Deng
  27. Jinyu Xia
  28. Xiaoping Tang
  29. Lei Liu
  30. Jinlin Hou

Reviewed by

Read the full article

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Background

Since the pandemic outbreak of coronavirus disease 2019 (COVID-19), the health system capacity in highly endemic areas has been overwhelmed. Approaches to efficient management are urgently needed. We aimed to develop and validate a score for early prediction of clinical deterioration of COVID-19 patients.

Methods

In this retrospective multicenter cohort study, we included 1138 mild to moderate COVID-19 patients admitted to 33 hospitals in Guangdong Province from December 27, 2019 to March 4, 2020 (N =818; training cohort), as well as two hospitals in Hubei Province from January 21 to February 22, 2020 (N =320; validation cohort) in the analysis.

Results

The 14-day cumulative incidences of clinical deterioration were 7.9% and 12.1% in the training and validation cohorts, respectively. An Early WArning Score (EWAS) (ranging from 0 to 4.5), comprising of age, underlying chronic disease, neutrophil to lymphocyte ratio, C-reactive protein, and D-dimer levels, was developed (AUROC: 0.857). By applying the EWAS, patients were categorized into low-, medium-, and high risk groups (cut-off values: two and three). The 14-day cumulative incidence of clinical deterioration in the low-risk group was 1.8%, which was significantly lower than the incidence rates in the medium-(14.4%) and high-risk (40.9%) groups (P <.001). The predictability of EWAS was similar in the validation cohort (AUROC =0.781), patients in the low-, medium-, and high-risk groups had 14-day cumulative incidences of 2.6%, 10.0%, and 25.7%, respectively (P <.001).

Conclusion

The EWAS, which is based on five common parameters, can predict COVID-19-related clinical deterioration and may be a useful tool for a rapid triage and establishing a COVID-19 hierarchical management system that will greatly focus clinical management and medical resources to reduce mortality in highly endemic areas.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2020.04.17.20064691: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Institutional Review Board StatementIRB: The study was conducted in accordance with the guidelines of the Declaration of Helsinki and the principles of International Committee of Harmonization (ICH) Good Clinical Practice (ICH - GCP).
    Randomizationnot detected.
    Blindingnot detected.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Statistical analysis: All data were entered into and analyzed using the Statistical Package for Social Science (SPSS version 20.0, Chicago, IL, USA) and R (Version 3.5.1).
    SPSS
    suggested: (SPSS, RRID:SCR_002865)

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Nonetheless, the study also has certain limitations. First, the patients in the training cohort were enrolled from more than 30 hospitals in Guangdong Province, and their disease status, exposure history, and treatment strategy were relatively heterogeneous; however, this heterogeneity strengthens the reliability of our scoring methodology, which shows similar predictive ability in different patient populations. Secondly, the EWAS was based on artificially defined categorical variables, which may have led to the loss of detailed continuous data. However, the categorical variables will be much simpler to apply and promote in highly endemic areas. Thirdly, the patients in the two cohorts were all Chinese. Whether the study results are applicable to patients in other Eastern or Western countries merits further investigation. In conclusion, the early-warning score, which is based on patients’ age, underlying chronic disease, NLR, CRP, and D-dimer, represents a reliable and simple scoring system for the prediction of clinical deterioration of COVID-19 within 14 days after admission. It may be a useful and convenient tool for a rapid triage and establishing a hierarchical management system of COVID-19 patients that will greatly focus clinical management and medical resources to reduce mortality in highly endemic areas.

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  2. Version published to 10.1101/2020.04.17.20064691 on medRxiv