2.4 Statistical analysis
Categorical variables were expressed as frequencies and percentages (%), and the frequencies of non-survived, survived severe, and non-severe patient groups (total patient number = 289) were compared by partition of chi-square test. Continuous variables were described as median and interquartile ranges (IQR) values, and one-way ANOVA test and Kruskal-Wallis test were used, as appropriate, to compare the data from three groups. A two-sided α of 0.0167 (after adjustment) was considered statistically significant for partition ofchi-square test when compare differences between categorical variables among 3 groups. On the other hand, a two-sided α of 0.05 was considered statistically significant for one-way ANOVA test andKruskal-Wallis test when compare differences between continuous variables among 3 groups.
Principal Component Analysis (PCA) was used for dimensionality reduction and visualization of the patients after imputing missing values using the implementation ”ppca” in pca Methods package (cite: https://academic.oup.com/bioinformatics/article/23/9/1164/272597). Euclidean distance and complete linkage were used for the heatmap between different laboratory parameters in three groups. Statistical analyses and figures were generated and plotted by GraphPad Prism version 7.00 software (GraphPad Software Inc), SPSS software (version 26.0, IBM) and R software (version 3.4.3, supported by R Foundation for Statistical Computing).