Multicenter Analysis of Clinical Features and Prognosis of COVID-19 Patients with Hepatic Impairment

Background/Aims: Recent data indicate the presence of liver enzyme abnormalities in patients with coronavirus disease 2019 (COVID-19). We aimed to evaluate the clinical features and treatment outcomes of COVID-19 patients with abnormal liver enzymes. Methods: We performed a retrospective, multicenter study of 874 COVID-19 patients admitted to five tertiary hospitals from February 20 to April 14, 2020. Data on clinical features, laboratory parameters, medications, and treatment outcomes were collected until April 30, 2020, and compared between patients with normal and abnormal aminotransferases. Results: Abnormal aminotransferase levels were observed in 362 patients (41.1%), of which 94 out of 130 (72.3%) and 268 out of 744 (36.0%) belonged to the severe and non-severe COVID-19 categories, respectively. The odds ratios (95% confidence interval) for male patients, patients with a higher body mass index, patients with severe COVID-19 status, and patients with lower platelet counts were 1.500 (1.029 to 2.184, p=0.035), 1.097 (1.012 to 1.189, p=0.024), 2.377 (1.458 to 3.875, p=0.001), and 0.995 (0.993 to 0.998, p>0.001), respectively, indicating an independent association of these variables with elevated aminotransferase levels. Lopinavir/ ritonavir and antibiotic use increased the odds ratio of abnormal aminotransferase levels after admission (1.832 and 2.646, respectively, both p<0.05). The median time to release from quarantine was longer (22 days vs 26 days, p=0.001) and the mortality rate was higher (13.0% vs 2.9%, p<0.001) in patients with abnormal aminotransferase levels. Conclusions: Abnormal aminotransferase levels are common in COVID-19 patients and are associated with poor clinical outcomes. Multivariate analysis of patients with normal aminotransferase levels on admission showed that the use of lopinavir/ritonavir and antibiotics was associated with abnormal aminotransferase levels; thus, careful monitoring is needed.