INTRODUCTION: Inappropriate use of vancomycin contributes to the development of resistant bacteria and jeopardizes the safety and effectiveness of treatment. The aim of this article was to design and validate an empirical dosing algorithm for vancomycin in premature neonates according to their population-based pharmacokinetic characteristics. PATIENTS AND METHODS: We performed a retrospective analysis of 129 serum samples from a cohort of 53 neonates. Homogeneous population groups were identified both from their individual pharmacokinetic parameters and from their biometric characteristics. The design of the dosing algorithm was based on simulation of the serum vancomycin concentration that would be reached with several different doses. The algorithm was validated in another cohort of 30 neonates and 108 serum samples. RESULTS: Introduction of the algorithm significantly increased the percentage initial values obtained with correct minimum and maximum concentrations in the first monitoring round (p<0.05). The mean number of serum samples obtained per patient for treatment monitoring was significantly reduced (3.6+/-2 vs. 4.9+/-3). CONCLUSIONS: The implantation of the dosing algorithm for vancomycin in premature neonates increased the efficiency of treatment, reduced monitoring requirements, and optimized serum vancomycin concentrations from the start of treatment.
INTRODUCTION: Inappropriate use of vancomycin contributes to the development of resistant bacteria and jeopardizes the safety and effectiveness of treatment. The aim of this article was to design and validate an empirical dosing algorithm for vancomycin in premature neonates according to their population-based pharmacokinetic characteristics. PATIENTS AND METHODS: We performed a retrospective analysis of 129 serum samples from a cohort of 53 neonates. Homogeneous population groups were identified both from their individual pharmacokinetic parameters and from their biometric characteristics. The design of the dosing algorithm was based on simulation of the serum vancomycin concentration that would be reached with several different doses. The algorithm was validated in another cohort of 30 neonates and 108 serum samples. RESULTS: Introduction of the algorithm significantly increased the percentage initial values obtained with correct minimum and maximum concentrations in the first monitoring round (p<0.05). The mean number of serum samples obtained per patient for treatment monitoring was significantly reduced (3.6+/-2 vs. 4.9+/-3). CONCLUSIONS: The implantation of the dosing algorithm for vancomycin in premature neonates increased the efficiency of treatment, reduced monitoring requirements, and optimized serum vancomycin concentrations from the start of treatment.