A recent article by Nestaas et al¹ deserves comment. This meta-analysis attempts to evaluate the relative efficacy and toxicity of extended interval dosing (EID) compared with traditional dosing (TD) of aminoglycosides administered to newborn infants. The conclusion of the analysis was that “extended interval dosing of aminoglycosides in neonates is safe and effective, with a reduced risk of serum drug concentrations outside the therapeutic range”.

This interpretation of the data presented can be questioned on several grounds. Even with meta-analysis, the number of reported treatment failures (two in the TD group, none in the EID group from a total of 555 patients) was too small to establish a statistically significant difference between the two regimens in relation to efficacy.¹

There were certainly more serum drug concentrations, both peak and trough, that were outside the defined therapeutic range using TD than using EID. However, this difference was only seen in trials that were aiming for a higher (usually 5–12 mg/l) peak serum drug concentration.¹ This observation is a self fulfilling prophecy. With EID, the drug is given at the same maintenance dose rate (mg/kg/day), but with higher single doses and longer dosage intervals than with TD. Basic pharmacokinetic principles indicate that during EID, peak serum drug concentrations will be higher and trough concentrations lower than when smaller drug doses are given at shorter dosage intervals (TD).²

The finding of increased numbers of serum drug concentrations outside the desired therapeutic range seems compelling, but is per se not necessarily a strong argument in favour of EID. The efficacy of aminoglycosides should not be based on this observation alone, even though there are theoretical and in vitro data to support the concept of “post antibiotic effect”.^1,3 Because treatment failures with TD are extremely rare,¹ the possible benefit in efficacy of EID becomes less important than establishing with confidence that this regimen has similar or less toxicity than TD.

Clinically significant renal impairment in neonates receiving aminoglycosides is extremely rare,^3,4 and is nearly always transient.⁴ Of greater significance is the potential for ototoxicity, which can occur with excessively high serum drug concentrations.³ In 1979, Finitzo-Hieber et al⁵ published a comprehensive evaluation of ototoxicity in newborns receiving TD aminoglycosides (gentamicin or kanamycin). More than 100 infants in each of three groups (gentamicin treated, kanamycin treated, and controls) were followed annually for up to four years. Detailed audiological testing was performed together with measures of vestibular function and psychometric assessment. There were no significant differences between the three groups in any of these measures.

In contrast, only four of the 16 studies included in this meta-analysis provided any data on ototoxicity. Brain stem evoked responses were only measured during, or shortly after, a course of an aminoglycoside (gentamicin or amikacin).^3,4,6,7 No adequate long term assessment of the incidence of hearing impairment has yet been reported in studies using EID. Until conclusive ototoxicity data are available in infants treated with EID, we will continue to use traditional dose aminoglycoside therapy.