Metabolic dysfunction-associated fatty liver disease and liver fibrosis score in patients with COVID-19 as predictors of adverse clinical outcomes: an artificial intelligent application through machine learning

Introduction and aims. Patients with coronavirus disease-2019 (COVID-19) and metabolic-dysfunction associated fatty liver disease (MAFLD) appear to be at higher risk for severe manifestations like acute respiratory distress syndrome, especially in the youngest decades. Our aim was to examine whether patients with imaging-defined MAFLD and/or with increased non-invasive liver fibrosis scores (FIB-4) are at higher risk for severe illness from COVID-19, using a machine learning model.

Methods. In this retrospective cohort study, we included 672 patients admitted for SARS-CoV-2 pneumonia between February the 28th 2020 and May the 1st 2021. Hepatic steatosis was detected by ultrasound or computed tomography (CT), whereas FIB-4 score was used to define the risk of advanced liver fibrosis. We used a machine learning (ML) model to evaluate the risks of both in-hospital death and prolonged hospitalizations (>28 days), considering MAFLD, a set of blood tests (hepatic profile; HP), and the FIB-4 score, either separately and together.

Results. Three hundred-thirty-three (49.6% of total) had imaging-defined MAFLD. The accuracy in predicting in-hospital death in the whole sample was 0.709 for the HP alone, and 0.721 for HP+FIB-4 combined together; in the 55-to-75 age subgroup, the accuracies were respectively 0.842 and 0.855 for HP alone and HP+FIB-4 together. In the MAFLD subgroup, the accuracy in predicting death was 0.739 considering HP alone, and 0.772 when considering HP+FIB-4 together; whereas in the MAFLD 55-to-75 years cohort, the accuracies were respectively 0.825 for HP and 0.833 for HP+FIB-4. Similar results were obtained both in the entire cohort and in MAFLD patients when considering the accuracy in predicting prolonged hospitalization (>28 days).

Conclusions. In our cohort of COVID-19 patients, the presence of a worse HP and a higher FIB-4 correlated with a higher risk of death and prolonged hospitalization, regardless of the presence of MAFLD. These findings could improve the clinical risk stratification of patients diagnosed with SARS-CoV-2 pneumonia.