· A burgeoning body of research exists about factors associated with in-hospital mortality among COVID-19 patients; however, focus on intensive care settings remains limited.
· The most frequent predictors of critical care mortality integrate age, physiologic markers and laboratory parameters in the most parsimonious models or prognostic scoring systems.
· Commonly used prognostic scoring systems such as MEWS, APACHE and SOFA provide crude mortality risk prediction that may be improved with machine learning algorithms that potentially offer more clinically relevant windows and opportunities for mortality risk prediction prior to deterioration.
Rapid Review Report: CC120402 RR (Version 1: December 9, 2020 14:51) 2
· Between centre variation is potentially an important determinant of critical care mortality that needs to be explored.
Williams-Roberts, H; Valiani, S; McLean, M; Miller, L; Howell-Spooner, B. What are the predictors of mortality in hospitalized COVID-19 patients? 2020 Dec 9; Document no.: CC120402 RR. In: COVID-19 Rapid Evidence Reviews [Internet]. SK: SK COVID Evidence Support Team, c2020. 33 p. (CEST rapid review report)