Serum cholinesterase on admission as a predictor of COVID-19 pneumonia severity and mortality
Abstract Background Although some predictors of COVID-19 pneumonia severity and mortality have been identified, much of the pathophysiology of this emerging infectious disease remains unclear. We hypothesized that a patient’s cholinesterase level on admission could predict COVID-19 pneumonia severity and mortality. Methods We retrospectively collected data of 26 COVID-19 pneumonia patients from February–May 2020. Outcomes were aggravation of symptoms and in-hospital mortality. We compared receiver operating curves of cholinesterase, C-reactive protein, lymphocytes, albumin, D-dimer, and PaO2/FiO2 ratio and examined prediction accuracy. Regarding the interaction between cholinesterase and other variables, each independent variable was divided into two groups using cutoff values, and interaction terms were created. Results Cholinesterase levels on admission were significantly lower in the severe group than in the mild-to-moderate group (326 vs. 218 IU/L, p = 0.006; area under the curve: 0.81; 95% confidence interval 0.61–0.94). When comparing the area under the curve, cholinesterase was comparable to C-reactive protein, albumin, lymphocytes, and PaO2/FiO2 ratio other than D-dimer in the prediction accuracy of severe cases and mortality. Cholinesterase levels on admission were significantly lower in the death group than in the survival group (274 vs. 187.5 IU/L, p = 0.028; area under the curve: 0.79; 95% interval 0.58–0.93). Regarding the interaction between cholinesterase and established predictors, the prediction accuracy of both severity and death was higher when cholinesterase was combined with each predictor than when cholinesterase was used alone. Conclusions Cholinesterase may reflect the disease state of COVID-19 pneumonia, suggesting that a patient’s cholinesterase level on admission may be useful as a predictor of severity and prognosis.