This review compared extended interval dosing (EID) with aminoglycosides with traditional interval dosing in neonates. The authors concluded that EID is safe and effective and reduces the risk of serum drug concentrations outside the therapeutic range. The evidence was largely based on pharmacokinetics with limited evidence about clinical efficacy, thus a more cautious conclusion about clinical effects may be appropriate.

To assess the efficacy and toxicity of extended interval dosing (EID) with aminoglycosides in neonates compared with traditional interval dosing (TID).

BIOSIS Previews, the Cochrane CENTRAL Register, EMBASE and MEDLINE were searched from inception to October 2004 for reports published in any language; the search terms were reported. The reference lists of reviews and identified studies were also screened. Ongoing and unpublished studies were sought in the UK National Research Register and the Current Science Register of controlled trials. Manufacturers of aminoglycosides were not contacted for unpublished studies.

Study designs of evaluations included in the review

Controlled clinical trials (CCTs) were eligible for inclusion. Studies with relevant subgroups were included, provided that data were reported separately for the subgroups of interest.

Specific interventions included in the review

Studies that compared equal daily doses of the same aminoglycoside (maximum 20% difference in daily dose) given in EID (dosage interval at least 24 hours and at least 50% longer than TID interval) with TID were eligible for inclusion. Most of the included studies used gentamicin (daily dose 2.5 to 5 mg/kg); other studies used amikacin (daily dose 15 mg/kg). The dosing intervals were predominantly 24 hours for EID and 12 hours for TID.

Participants included in the review

Studies of neonates aged less than 30 days at the start of treatment were eligible for inclusion. The included studies often excluded neonates with asphyxia, risk of deafness, kidney malformations and co-treatment with other potentially nephrotoxic or ototoxic drugs.

Outcomes assessed in the review

The included studies were required to report the clinical cure rate, toxicity, or prevalence of peak and trough serum drug concentrations (SDCs) inside and outside the therapeutic range. The main outcome measure of the review was the risk ratio (RR) for an SDC outside the therapeutic range. Other outcomes reported were treatment failure and toxicity (ototoxicity and nephrotoxicity). The studies used different definitions for nephrotoxicity. The reviewers accepted the definitions used in the primary studies.

How were decisions on the relevance of primary studies made?

One reviewer conducted the search.

Studies were assessed for study design and the following potential sources of bias: selection bias (biased allocation to treatment); performance bias (unequal care apart from treatment of interest); detection bias (biased assessment of outcome); and attrition bias (biased occurrence and handling of protocol deviations and losses to follow-up). Two blinded reviewers independently assessed validity and resolved any disagreements by consensus.

Two blinded reviewers independently extracted the data using a form and resolved any disagreements by consensus. The reviewers generally extracted SDCs at 48 to 72 hours after the start of treatment. The authors of studies that did not report clinical cure rate completely were contacted for additional information.

How were the studies combined?

Summary RRs with 95% confidence intervals (CIs) were calculated using a random-effects model, weighted by the inverse variance of the natural logarithm of the RR. For studies with zero events, 0.5 was added to each cell in the meta-analyses. Publication bias was assessed using funnel plots.

How were differences between studies investigated?

A subgroup analysis was used to assess the influence of therapeutic peak SDC range upon the RR. Sensitivity analyses were conducted for study design (randomised versus other designs) and study quality (bias versus no bias). The statistical significance of differences between the subgroups was tested using the chi-squared statistic.

Sixteen CCTs (n=823) were included.: 11 randomised controlled trials (RCTs), 4 CCTs with a historical control group and one cohort study. Data for 5 trials were derived from subgroups.

The studies seldom completely reported study design and methodology. The most common sources of bias were different daily doses and different times of assessing SDCs. All but 6 studies had potential sources of bias.

SDCs (12 CCTs, n=698).

EDI was associated with a significantly lower risk of peak (RR 0.50, 95% CI: 0.26, 0.94; P=0.033) and trough (RR 0.36, 95% CI: 0.25, 0.56; P<0.001) SDC outside the therapeutic range than TID.

There was no clear evidence of publication bias for peak or trough RR (P>0.10).

There were statistically significant differences between studies accepting lower versus higher peak SDC (P=0.013), for RCTs versus non-randomised CCTs (P<0.01) but not for trials with bias versus trials without bias (P>0.1).

For the 4 studies accepting lower peak SDC with TID, there was no significant difference between EID and TID. For the 8 studies using higher peak SDC, EID was associated with a significantly lower risk of peak SDC outside the therapeutic range (RR 0.38, 95% CI: 0.24, 0.61, P<0.001). There was no significant difference in the risk of peak SDCs outside the therapeutic range between EID and TID for all RCTs and for RCTs using higher therapeutic peak range (the results were reported); EID was associated with a significantly lower risk of troughs outside the therapeutic range than TID (RR 0.42, 95% CI: 0.26, 0.66, P<0.001). The 5 non-randomised trials showed a significantly lower risk of both peak and trough SDC outside the therapeutic range for EID in comparison with TID (the results were reported).

Efficacy and toxicity.

Treatment failure (9 CCTs, n=555): one trial reported treatment failure in two neonates, both in the TID arm. No deaths were reported.

Nephrotoxicity (12 CCTs, n=589): one trial found urinary alanine aminopeptidase in all participants, while another found raised creatinine concentration in 30 of the 40 participants (13 in the TID arm and 17 in the EID).

Ototoxicity (4 CCTs, n=210): one event was found in the EID group.

EID of aminoglycosides in neonates is safe and effective and reduces the risk of an SDC outside the therapeutic range.

The review addressed a clear question that was defined in terms of the participants, intervention, outcomes and study design. Several relevant sources were searched and attempts were made to minimise language and publication bias. Only one reviewer selected studies and this lack of duplication might have led to errors and bias. Two reviewers independently assessed validity and extracted the data, thus reducing the potential for bias and errors. Validity was assessed using specified established criteria, and results of the assessment were reported and appropriately applied to the analyses. There was no mention of the duration and completeness of follow-up. The pooling of randomised and non-randomised studies is of questionable value. However, a pre-specified sensitivity analysis was conducted to determine the impact on effect size of excluding non-randomised trials. The evidence was largely based on pharmacokinetics with limited evidence about clinical efficacy, and a more cautious conclusion about safety and efficacy may therefore have been more appropriate.

Practice: The authors stated that the SDC should be assessed when using EID.

Research: The authors stated that large studies would be required to assess clinical cure rates for EID compared with TID.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.