Background: Different pharmacokinetic dosage adjustment methods have been developed to individualise tobramycin treatment. This study aimed to compare several currently available methods to predict tobramycin exposure after once daily dosing in cystic fibrosis children.
Methods: Retrospective data from 173 cystic fibrosis patient’s treated at the Royal Children Hospital (Brisbane) were analysed. Each patient had at least one paired tobramycin plasma concentration-time measurement recorded after one or more dosing intervals. Tobramycin dosage and pharmacokinetic parameters were estimated using the following methods: (i) the Therapeutic Guidelines Aminoglycoside Dose Adjustment nomogram, (ii) a nomogram developed by Massie et al. (REF), (iii) linear regression analysis and Bayesian forecasting based on two software programs: (iv) TCIWorks and (v) DoseMe.
Results: The initial mean±SD tobramycin dose administered in the patient group was 10.11±2.42 mg/kg. Using the TG nomogram significantly greater doses for dose adjustment were recommended, 27.6 ± 22.3 mg/kg, compared to the other methods, (12.0- 14.4 mg/kg). Linear regression analysis was more biased and less precise in predicting observed tobramycin concentration compared to Bayesian forecasting methods.
Conclusions: TCIWorks and DoseMe are generally comparable for dose predictions in CF children. Linear regression analysis and the Massie nomogram can be considered as alternatives when Bayesian forecasting software is unavailable. The TG nomogram should not be used to aid dose adjustments of tobramycin in this population.
