Eight years' experience of an extended-interval dosing protocol for gentamicin in neonates

Dosing of gentamicin in neonates in Christchurch has been carried out since 2000 using a locally developed extended-interval dosing protocol. All dosing data have been recorded in a database. The aims of this study were to analyse the database to determine what percentage of neonates achieved target values for C(max), C(min) and AUC, and to use the pharmacokinetic values of gentamicin to simulate new dosing protocols. C(max), C(min) and AUC were compared with target values. Clearance (CL), volume of distribution (V) and half-life (t(1/2)) were estimated, and used to produce new predictive dosing protocols that were tested and compared with the results of the original protocol. Gentamicin concentrations from 1461 individual doses were recorded in the database. Four hundred and eight were excluded. Of the remaining 1053, 84% achieved the target C(max) (> 10 mg/L), 77% the target C(min) (< 1 mg/L) and 63% the target AUC (within 80% to 125%). The number achieving target C(max) and C(min) values was improved markedly by prolonging the dosing intervals, but not by altering the predictive equations. Since the majority of the neonates only received a single dose of gentamicin, a new V-based model was also tested, and performed well. CL (L/kg) increased, while V (L/kg) and t(1/2) (h) both decreased with respect to weight. Extending the dose interval improved the success in achieving target C(max) and C(min), while revision of the dosing equation did not. A V-based model provides an alternative approach to the first dose of gentamicin in neonates.