Correlation Analysis Between Disease Severity and Inflammation-related Parameters in Patients with COVID-19 Pneumonia

Jing Gong; Hui Dong; Song Qing Xia; Yi Zhao Huang; Dingkun Wang; Yan Zhao; Wenhua Liu; Shenghao Tu; Mingmin Zhang; Qi Wang; Fuer Lu

For data of normal distribution (IL2R, ESR, ferroprotein, WBC and NC), comparisons among critical, severe and mild groups were analyzed by ANOVA analysis. With different homogeneity of variance, the pairwise comparison between groups was performed using the Bonferroni test or Dunnett's T3 test. For non-normal distribution (CRP and LC), the data were conversed using square root, followed by ANOVA analysis and pairwise comparison. With data below the detectable limit, including IL-1β, IL-6, IL-8, IL-10, TNFα, PCT, and EC, the data were ranked referring to the reference range and value rank (Table 1) ; then non-parametric Kruskal-Wallis test were performed. In correlation analysis, Spearman correlation coefficient was used for the variables of normal distribution, Pearson correlation coefficient for those of skewed distribution, and Kendall's tau-b correlation coefficient for ranked data.