Preprint Article Version 1 Preserved in Portico This version is not peer-reviewed

Development and Validation of a Prediction Model for 28-Day in-Hospital Mortality in Critically Ill Patients with COVID-19

Version 1 : Received: 29 January 2021 / Approved: 1 February 2021 / Online: 1 February 2021 (15:51:34 CET)

How to cite: Ferrando-Vivas, P.; Gould, D.; Doidge, J.; Thomas, K.; Mouncey, P.; Shankar-Hari, M.; Young, D.; Rowan, K.; Harrison, D. Development and Validation of a Prediction Model for 28-Day in-Hospital Mortality in Critically Ill Patients with COVID-19. Preprints 2021, 2021020059. https://doi.org/10.20944/preprints202102.0059.v1 Ferrando-Vivas, P.; Gould, D.; Doidge, J.; Thomas, K.; Mouncey, P.; Shankar-Hari, M.; Young, D.; Rowan, K.; Harrison, D. Development and Validation of a Prediction Model for 28-Day in-Hospital Mortality in Critically Ill Patients with COVID-19. Preprints 2021, 2021020059. https://doi.org/10.20944/preprints202102.0059.v1

Abstract

Objectives To develop and validate a prediction model for 28-day in-hospital mortality among adult patients critically ill with COVID-19 in the UK. Design Observational cohort study. Setting 287 adult critical care units in England, Wales and Northern Ireland, of which 260 admitted at least one eligible patient. Participants 10,933 patients with confirmed COVID-19 of whom 10,401 were eligible (excluding 532 patients with a duration of critical care less than 24 hours and 1 patient with unknown 28-day outcome): 8,666 development (March-April 2020) and 1,735 temporal validation (May-August 2020). Main outcome measures 28-day in-hospital mortality from start of critical care. Results Two models were developed using 14 patient level predictors selected from 30 candidate predictors, with and without adjustment for calendar time. In the temporal validation data, the model discrimination was maintained (c index 0.78) but calibration was poor, particularly for the model not adjusted for calendar time (ratio of observed to predicted mortality 0.74 versus 0.88 for the model adjusted for calendar time). Conclusions We developed and validated a prediction model for 28-day in-hospital mortality for patients critically ill with COVID-19. Although absolute predictions were inaccurate due to changing outcomes, the models will support risk-adjustment in analyses and monitoring changes in risk-adjusted outcome over time.

Keywords

COVID-19; Critical Illness; Hospital Mortality; Intensive Care Units; Outcome Assessment, Health Care; Prognosis; Risk Adjustment; Severe acute respiratory syndrome coronavirus 2

Subject

Medicine and Pharmacology, Immunology and Allergy

Comments (0)

We encourage comments and feedback from a broad range of readers. See criteria for comments and our Diversity statement.

Leave a public comment
Send a private comment to the author(s)
* All users must log in before leaving a comment
Views 0
Downloads 0
Comments 0
Metrics 0


×
Alerts
Notify me about updates to this article or when a peer-reviewed version is published.
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here.