Objective: To identify mortality predictors of critically ill COVID-19 patients in ICU based on current available literatures.
Methods: Systematic literature search was conducted in open-access databases. Data extraction was conducted for publication date, methodology employed, sample size, and results of multivariate analysis. Eligibility criteria for analysis was observational analytic design, sample size of 100 or more, and availability of multivariate results. Primary measures assessed was risk ratio, presented as odds ratio or hazard ratio. Data was analyzed qualitatively for themes that emerged for mortality predictors.
Results: Several mortality predictors were identified, which included demographic, clinical history, laboratory results, and oxygenation profile at ICU admission. Several of the most consistently reported mortality predictors was older age, one or more comorbidities that constitute metabolic syndrome, chronic pulmonary disorder, low lymphocyte and platelet count, elevated d-dimer, and low PaO2/FiO2 ratio.
Conclusions: Mortality predictors identified in this review were similar to previously known mortality and severity predictors of COVID-19 patients in general. This consistency may point to the potential of developing a scoring system to predict COVID-19 severity and mortality for clinical practice use.