Development and validation of COEWS (COVID-19 Early Warning Score) for hospitalized COVID-19 with laboratory features: A multicontinental retrospective study

• DOI: 10.7554/elife.85618
• Publication Date: 2023-08-24T11:40:13Z
• Authors: Riku Klén; Ivan A Huespe; Felipe Aníbal Gre...; Antonio Lalueza L...; Miguel Pedrera Ji...; Noelia Garcia Barrio; Pascual Ruben Valdez; Matias A Mirofsky; Bruno Boietti; Ricardo Gómez-Hue...; José Manuel Casas...; Juan Miguel Antón...; Javier Alberto Po...; David Gómez-Varela
• Keywords: 2019-20 coronavirus outbreak; Pandemic

The emergence of new SARS-CoV-2 variants with significant immune-evasiveness, the relaxation measures for reducing number infections, waning immune protection (particularly in high-risk population groups), and low uptake vaccine boosters, forecast waves hospitalizations admission to intensive care units. There is an urgent need easily implementable clinically effective Early Warning Scores (EWSs) that can predict risk complications within next 24-48 hr. Although EWSs have been used evaluation COVID-19 patients, there are several clinical limitations their use. Moreover, no models tested on geographically distinct populations or groups varying levels protection.We developed validated Score (COEWS), EWS automatically calculated solely from laboratory parameters widely available affordable. We benchmarked COEWS against NEWS2. also evaluated predictive performance vaccinated unvaccinated patients.The variables model were selected based coefficients wide availability these variables. final included complete blood count, glucose, oxygen saturation features. To make more actionable real situations, we transformed into individual scores each feature. global score serves as easy-to-calculate measure indicating a patient developing combined outcome mechanical ventilation death 48 hr.The discrimination external validation cohort was 0.743 (95% confidence interval [CI]: 0.703-0.784) performed 0.700 CI: 0.654-0.745) scores. area under receiver operating characteristic curve (AUROC) similar patients. Additionally, observed AUROC NEWS2 0.677 0.601-0.752) patients 0.648 0.608-0.689) predicts MV hr routine measurements. extensive validation, its high performance, ease use, positive benchmark comparison position reference tool assisting decisions improving upcoming pandemic waves.University Vienna.