Objectives Once-daily gentamicin therapy is becoming increasingly common in paediatric practice; however, little is known about pharmacokinetics in critical illness. Gentamicin exhibits concentration dependent killing; thus, peak serum concentrations at least eight times higher than minimum inhibition concentration of the target organism have been recommended. We wanted to derive pharmacokinetic parameters for gentamicin in critical illness and to evaluate whether a dose of 8 mg/kg provides an adequate peak serum concentration (>16 mg/l).

Methods Population-based pharmacokinetic analyses were undertaken using therapeutic drug monitoring data collected prospectively in an intensive care unit over 6 months (n=50 children). Monte Carlo simulations were used to estimate the probability of achieving (1) peak concentrations >16 mg/l; and (2) trough concentrations <2 mg/l at 24 and 36 h.

Results The optimal pharmacokinetic model was of two-compartment disposition with zero order input and additive residual error. Weight was associated nonlinearly with clearance and linearly with volume, and age was a significant covariate for clearance. An 8 mg/kg dose provided near 100% probability of achieving adequate peak concentrations at all ages. However this probability decreased rapidly at doses <7 mg/kg, with neonates being the most susceptible. Approximately 50% of non-premature neonates within the first week of life, 25% of infants, and 10% of children are likely to need a dose interval >24 h.

Conclusions A gentamicin dose of 8 mg/kg is highly likely to achieve peak concentrations >16 mg/l in critically ill children. A considerable proportion will require dose intervals >24 h; thus, therapeutic drug monitoring is essential.