Statistical analyses
Continuous data are presented as means ± standard deviations for normally distributed variables or as median and interquartile range for non-normally distributed variables. Categorical data are presented as absolute values and frequencies (%). Comparative analyses for the MBP and MMP groups were performed using Fisher’s exact or Chi-square test for categorical variables and Student’s t-test or Mann-Whitney U test for continuous variables, as appropriate. P values < 0.05 were considered statistically significant. All statistical analyses were performed using IBM SPSS 20 (IBM, USA). We used Cox proportional hazard models to analyze the competing risk data and Fine and Gray cumulative incidence function-based proportional hazard competing risk regression model to analyze survival data.[11]