ABSTRACT
Coronavirus disease 2019 (COVID-19) has caused more than 6 million deaths. Higher values of the FIB-4 index have been shown to be associated with disease severity. Although vaccination has helped to improve clinical outcomes and overall mortality, it remains important to identify clinical parameters that can predict a likely worse prognosis. Artificial intelligence and big data processing were used to retrieve data from patients with Covid-19 admitted during the period March 2020-January 2022 at the Fondazione Policlinico Gemelli IRCCS. Patients and clinical characteristics of patients with available FIB 4 data derived from the Gemelli Generator Real World Data (G2 RWD) were used to develop predictive models of mortality during the 4 waves of the Covid-19 pandemic. A logistic regression model was applied to the training and test set. The performance of the model was assessed by means of the ROC curve. After the pre-processing steps, 1143 patients and 35 variables were included in the final dataset. The FIB-4 discretization algorithm identified a cut-off of 2.54. After fitting the model for multiple mortality regression analysis: FIB-4>= 2.53 (OR=4.53, [CI: 2.83 - 7.25]), wave 3 (OR=0.34, [CI: 0.15 - 0.75], wave 4 (OR=0.40 [CI: 0.24 - 0.66]) and LDH (OR=1.001, [CI: 1.000 - 1.002]) were considered. The machine learning approach identified a cut-off of 2.53 for FIB-4 above which the risk of death increases significantly. These data may be useful in the clinical management of patients with Covid-19, as they can be calculated from the blood test after hospital admission.