Statistical analysis
Continuous variables were expressed as medians and interquartile ranges (IQR), and categorical variables were presented as numbers and frequencies with percentages to describe the populations, as appropriate. The two-group comparison was done by the Mann-Whitney-U test, and Kruskal-Wallis H nonparametric test was used for multiple comparisons between different groups. A χ2 test or Fisher exact test was used to compare categorical variables, as appropriate. Actuarial survival rates were analyzed by the Kaplan-Meier method and survival was measured in days from diagnosis to death or the last review. The log-rank test was applied for comparisons between groups. Univariate and multivariate logistic regression was performed to explore the association of clinical characteristics and laboratory parameters and the risk for superinfection. A stepwise backward conditional method was used to select variables entering the multivariate model. Strongly correlated variables were excluded from multivariate analyses. Multi-collinearity was checked before determining the final model. We assessed goodness of fit with the Hosmer-Lemeshow (HL) test. Data analysis was performed with MedCalc and SPSS 24.0 for Windows software (SPSS Inc. Chicago). For all tests, a two-sided P < 0.05 was considered as statistically significant.