Statistical analysis
The study cohort was categorized into those who achieved a positive
treatment outcome and those who did not; these two groups were compared
using a Chi-square test or a Student’s t-test, as appropriate, with
regards to demographic and APE variables. Levene’s test was used to
determine if equality of variance was to be assumed for continuous
variables. A 2-sided α of 0.05 and/or a 95% confidence interval (CI)
was used in all analysis. Doses were converted from mg to mg/kg to
account for weight differences between patients. Antibiotic combinations
administered in concomitance to tobramycin were given their own category
if they were used in 5% or more cases, otherwise, they were included in
the « other » category. A multivariable logistic regression analysis was
performed for determination of risk factors for therapeutic failure. All
variables of clinical interest that differed between groups were forced
into the model to account for variability. Microorganism susceptibility
was also included. The Matouk Score was the only variable assessing the
severity of the APE included in the analysis. All statistical analyses
were performed on SPSS version 27.