Influence of Age on Frequency of Vancomycin Dosing: Statistical Analysis
Sample size was determined on the basis of a predicted OR of 3, with a desired power of 80% and a of 0.05. It was estimated that 100 controls and 50 cases (2:1 ratio) would be required. However, because of difficulty in locating an adequate number of controls, the ratio of cases to controls was changed to 1:1, which resulted in a new estimated sample size of 64 patients per group.
The primary outcome was calculated using a logistic regression model, with age category (less than 40 years versus greater than or equal to 40 years) as the predictor variable. The influence of other potential variables, specifically use of concurrent nephrotoxic drugs (aminoglycosides, angiotensin- converting enzyme inhibitors, nonsteroidal anti-inflammatory drugs), diabetes mellitus, and sex, was also explored with this model. levitra plus
The Student t test was used to compare differences in mean age, mean baseline serum creatinine, mean baseline creatinine clearance (as calculated by the modified Cockcroft- Gault formula), and mean vancomycin dose per interval received by each group. Baseline serum creatinine was the value of serum creatinine measured closest to the time of initiation of vancomycin. The Mann-Whitney U test was used to assess differences between the cases and controls with respect to time until achievement of first therapeutic level. The \2 test was used to assess differences between the 2 groups in terms of increase in serum creatinine above 30% and regimen change secondary to increase in serum creatinine above 30% or accumulation of vancomycin. Descriptive statistics were used to report the proportion of patients for whom each regimen was prescribed empirically, the proportion of patients in each age category requiring each regimen, and the professional responsible for dosing change.