Case Series; Cross-Sectional
Study population and setting
This study enrolled 417 patients with RT-PCR confirmed SARS-CoV-2 infection who were followed in the only referral hospital in Shenzhen, China from January 11 to February 21, 2020. The study evaluated the burden of abnormally elevated liver enzymes: aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (GGT), alkaline phosphatase (ALP) and total bilirubin among COVID-19 patients prospectively from admission through hospitalization. It also identified typologies of liver abnormalities and assessed their association with severe COVID-19 disease. Liver injury was defined as ALT and/or AST levels more than three times the upper limit of normal (ULN), GGT and/or ALP exceeding two times the ULN and a total bilirubin > two times the ULN.
Summary of Main Findings
The prevalence of abnormally elevated liver results increased from 50% on admission to 76% during hospitalization, while the burden of liver injury increased from 5% to 21% over the course of hospitalization. Overall, 43% of patients had mixed type i.e both higher levels of AST and/or ALT, and higher levels of ALP and/or GGT. Being male, having a high BMI, old age, having underlying liver disease and use of antibiotics, NSAIDs, interferon and herbal medicine were associated with abnormal liver results. In the multivariable logistic analysis, using peak values of hepatic results during hospitalization, the odds of severe disease were significantly higher for patients with abnormal hepatic results [adjusted OR: 2.48(95%CI:0.94-6.55] and liver injury [adjusted OR:9.04(95%CI:3.19-25.6)] compared with individuals with normal values. Except for Lopinavir/ritonavir use aOR: 4.4(95%CI:1.50-13.17); other medications including antibiotics, NSAIDs, ribavirin, and interferon were not associated with liver injury in the multivariable analyses.
The study collected salient clinical and laboratory variables. Repeated measurement of liver enzymes and total bilirubin enabled investigators to document the changes over time. The adjusted associations are independent of important prognostic factors including age, sex, and some underlying conditions.
The study was conducted in one referral hospital, which makes it subject to selection bias and its findings less generalizable to all hospitalized patients. The precision of estimates in multivariable analyses was low (confidence intervals are wide). Finally, it is possible that reported associations are biased by unmeasured or imperfectly measured differences between those with and without liver abnormalities, even after adjustment especially from medications, liver comorbidity and other unmeasured comorbidities.
This study is one of the most comprehensive to date on COVID-19 and liver function. It provides modest evidence of the high burden of hepatic injury among patients hospitalized for COVID-19 in this region, and its association with severe COVID-19 disease.