The clinical application of blood routine and liver and kidney function test in COVID-19 disease prediction
Y. Liang,Y. Ma,K. Huang,H. Luo,A. M. Wang,Z. Xiong,Y. Luo,B. Yi
DOI: https://doi.org/10.3760/cma.j.cn112150-20200629-00947
2021-01-01
Abstract: To explore the clinical application value of routine indicators such as blood routine and liver and kidney function in auxiliary diagnosis and prognosis of COVID-19 patients. SNK-q and other methods were used to retrospectively analyzed the differences of blood routine test, liver and kidney function and other inflammatory indexes of 30 patients with covid-19, 29 patients with other viral pneumonia, 35 patients with influenza A/B and 25 healthy persons from January 28 to February 14, 2020 in Xiangya Hospital of Central South University. The neutrophils count increased gradually in COVID-19 group, influenza A/B group and other types of viral pneumonia group, and the difference between COVID-19 group and other viral pneumonia groups was statistically significant(=-19.064,0.05); The lymphocyte count decreased gradually in the control group, influenza A/B group, other viral pneumonia group and COVID-19 group. In addition, DB, UA and GLU were also different among groups. Subgroup analysis showed that there were statistically significant differences in N(9.581,-0.152,0.05), N%(5.723,-0.600, 0.05), NLR(4.773, -1.161, 0.05), PCT(17.464, -1.477, 0.05)and CRP(7.656, -1.973, 0.05) between patients with lung involvement +-++ and patients with lung involvement +++-++++. There were statistically significant differences in NLR(63.931, -2.815, 0.01), AST(15.704, -1.930, 0.01), ALT(35.551, -2.199, 0.01), LDH(7.715, -2.703, 0.05) and GLU(6.306, -5.116, 0.05) between the light+common subgroup and the heavy+critical subgroup of COVID-19 clinical classification. Correlation analysis showed that clinical stage and imaging credit period were significantly correlated with NLR (=0.406=0.026; =0.397=0.030), ALT (=0.403=0.049; =0.418=0.047), LDH (=0.543<0.01; =0.643<0.01) and GLU(0.750<0.01; =0.471=0.042). A total of 5 principal components were extracted from all the included indicators, and the comprehensive information extraction rate was 82.86%. Indicators of a large load included Ur, PCT and CRP in PC1; ALT, AST and GLU in PC2; N%, L%, L and NLR in PC3. It indicated that the indicators of acute infection, liver function and blood routine had certein warning effect on disease surveillance. The results of ROC curve analysis showed that the combined detection of N+TB+Urea was the best practice to distinguish COVID-19 and other viral pneumonia, while the combined detection of N+L+UA was the most effective solution to make a distinction between COVID-19 and influenza A/B patients. In the aspect of disease evaluation, NL+LDH+GLU+ALT combined detection represent the best diagnostic performance to distinguish the clinical stage of light+common type and heavy+critical type, achieving the AUC (ROC) to 0.904, with the sensitivity 75% and the specificity 100% at the cut-off value of 0.477. In addition to etiology and imaging examination, doctors can also improve the routine laboratory tests such as blood routine test, liver and kidney function to assist diagnosis and disease prediction of patients with respiratory tract infection.