The FIB-4 scores in the emergency department to predict the outcomes of COVID-19 patients in taiwan

Abstract

Objective: We aimed to determine the reliability of using the Fibrosis-4 (FIB-4) index in COVID-19 patients without underlying liver illness.

Method: We employed multivariate logistic regression to identify variables that exhibited statistically significant influence on the ultimate outcome. Multilayer perceptron analysis was employed to develop a prediction model for the FIB-4 index concerning ICU admission and intubation rates. However, the scarcity of cases rendered the assessment of the mortality rate unfeasible. We plotted ROC curves to analyze the predictive strength of the FIB-4 index across various age groups.

Result: In univariate logistic regression, only the FIB-4 index and respiratory rate demonstrated statistical significance on all poor outcomes. The FIB-4 index for mortality prediction had an ROC and AUC of 0.863 (95% CI: 0.781-0.9444). It demonstrates predictive power across age groups, particularly for age ≥65 (AUC: 0.812, 95% CI: 0.6571-0.9673) and age <65 (AUC: 0.878, 95% CI: 0.8012-0.9558). Its sensitivity for intubation and ICU admission prediction is suboptimal.

Conclusion: FIB-4 index had promising power in prediction of mortality rate in all age groups.

Keywords: COVID-19; Critical care medicine; Respiratory distress.