Abstract
Background: Considering transaminase more than the upper limit of normal value as liver injury might overestimate the prevalence of liver involvement in COVID-19 patients. No meta-analysis has explored the impact of varied definitions of liver injury on the reported prevalence of liver injury. Moreover, few studies reported the extent of hypertransaminasemia stratified by COVID-19 disease severity. Methods: A literature search was conducted using PubMed and Embase. The pooled prevalence of liver injury and hypertransaminasemia was estimated. Results: In total, 60 studies were included. The overall prevalence of liver injury was 25%. Compared to subgroups with the non-strict definition of liver injury (33%) and subgroups without giving detailed definition (26%), the subgroup with a strict definition had a much lower prevalence of liver injury (9%). The overall prevalence of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) elevation was 19% and 22%. The prevalence of elevated ALT and AST were significantly higher in severe COVID-19 cases compare to non-severe cases (31% vs 16% and 44% vs 11%). In critically ill and fatal cases, no difference was found in the prevalence of elevated ALT (24% vs 30%) or AST (54% vs 49%). Sensitivity analyses indicated that the adjusted prevalence of ALT elevation, AST elevation, and liver injury decreased to 14%, 7%, and 12%. Conclusion: The overall prevalence of liver injury and hypertransaminasemia in COVID-19 patients might be overestimated. Only a small fraction of COVID-19 patients have clinically significant liver injury. The prevalence of hypertransaminasemia was significantly higher in severe COVID-19 cases compare to non-severe cases. Hence, in severe COVID-19 patients, more attention should be paid to liver function tests.