Statistical analysis
Minimum sample size required in each arm, calculated based on a previous
study, were 33 subjects.16 Each outcome was defined as
a febrile neutropenic event. All statistical analyses were performed
using STATA version 16 (College Station, TX). Chi-square test
(two-tailed) was used to compare proportion for categorical variables
while t-test was used to compare means for continuous variables.
Antibiotic utilization was reported as defined daily dose per 1000
patient-days. Trend of antibiotic utilization was analyzed by linear
regression and reported as coefficient and p-value. Univariate and
multivariate analysis of variables influencing on 30-day infectious
diseases-related mortality were performed. All comparisons were 2-sided
and a P value<0.05 was consider statistically
significant.