Background SARS-CoV-2 infection is novel and widespread globally, causing a range of illness severity. The impacts on the liver from SARS-CoV-2 are variable and appear unpredictable. Objective In the weeks following SARS-CoV-2 infection, liver-related blood tests are abnormal in patterns different than in matched controls without SARS-CoV-2 infection. Methods This was a retrospective study of health administrative data of individuals in Ontario, Canada between January 2020 and September 2022. Cases were individuals with their first positive COVID-19 test with at least one liver-related blood test (AST, ALT, albumin, INR, bilirubin) available in the following 28 days. Two control individuals (with a negative COVID-19 test) were age/sex matched to each case. Any patients with a diagnostic code indicting pre-existing liver disease, or a positive COVID-19 test in the preceding six months were excluded. Machine learning analysis approaches included logistic regression, random forest, extreme gradient boost (XGBoost), and neutral network models. Results 100,574 cases and 200,709 matched controls were included. 49% (49,299) had at least one abnormal liver test in 28 days following positive COVID-19 test. Of the controls, 41% (82,495) had at least one abnormal liver test in the 28 days following negative COVID-19 test. The optimal machine learning model used XGBoost for adult cases, with an area under the curve of 0.74. Conclusions Further studies could use this approach to create monitoring systems to use abnormalities like liver enzyme changes, at a population level, as a bellwether for not only COVID-19 outbreaks, but also other hepatotropic illness epidemics, and pandemics.
Dimitriev et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: