2.7 Statistical analysis
Categorical variables are reported as counts and percentages and continuous variables as mean ±SD, or medians and interquartile ranges (IQRs). Differences between percentages were assessed by chi-square or Fisher exact tests. Student unpaired t -tests and analysis of variance were used for normally distributed continuous variables. Appropriate nonparametric tests (Mann-Whitney, Kruskal-Wallis and Spearman rank correlation tests) were used for all other variables.
Data were analyzed for the assessment of treatment effect on biomarkers performing a repeated measures MANOVA with one between subject factor (treatment group) and one within-subject factor (time at two or three levels). As covariates, we considered the possible random differences in age and comorbidities between the groups.
After dividing the population into groups, the cumulative incidence was estimated using a Kaplan–Meier product–limit estimator. Survival curves were formally compared using the log-rank test. Cox proportional hazards analysis was used to calculate the adjusted relative hazards of outcome events by each clinical variable.
For multivariate models, model selection was performed using forward stepwise regression on the basis of the Akaike information criterion.
Only p values <0.05 were considered statistically significant. All tests were 2-tailed, and analyses were performed using computer software packages (R version 2.15.2, R Development Core Team, Wien, Austria).