Background. Therapeutic drug monitoring (TDM) of aminoglycosides and vancomycin is used to prevent oto- and nephrotoxicity in neonates. Analytical and non-analytical factors potentially influence dosing recommendations.
Aim. To determine the impact of analytical variation (imprecision and bias) and non-analytical factors (accuracy of drug administration time, use of non-trough concentrations, biological variation and dosing errors) on neonatal antimicrobial dosing recommendations.
Methods. Published population pharmacokinetic models and the Australasian Neonatal Medicines Formulary were used to simulate antimicrobial concentration-time profiles in a virtual neonate population. Laboratory quality assurance data were used to quantify analytical variation in antimicrobial measurement methods used in clinical practice. Guideline-informed dosing recommendations based on drug concentrations were applied to compare the impact of analytical variation and non-analytical factors on antimicrobial dosing.
Results. Analytical variation caused differences in subsequent guideline-informed dosing recommendations in 9.3−12.1% (amikacin), 16.2−19.0% (tobramycin), 12.2−45.8% (gentamicin), and 9.6−19.5% (vancomycin) of neonates (Figure 1). For vancomycin, inaccuracies in drug administration time (45.6%), use of non-trough concentrations (44.7%), within-subject biological variation (38.2%) and dosing errors (27.5%) were predicted to result in more dosing discrepancies than analytical variation (12.5%). Using current analytical performance specifications, tolerated dosing discrepancies would be up to 14.8% (aminoglycosides) and 23.7% (vancomycin).
Conclusion. Although analytical variation can influence neonatal antimicrobial dosing recommendations, non-analytical factors are more influential. These result in substantial variation in subsequent dosing of antimicrobials, risking inadvertent under- or over- exposure. Harmonisation of measurement methods and improved patient management systems may reduce the impact of analytical and non-analytical factors on neonatal antimicrobial dosing.
Figure 1. Discrepancies between antimicrobial dosing recommendations when using true (without analytical variation) compared with measured (with analytical variation) concentrations for 2 measurement methods per antimicrobial. The 8 measurement methods A to H are commonly used in Australia and have substantial analytical variation. Data are expressed as the mean proportions of neonates with a discrepancy in dosing recommendations.