Background/Objectives: Vancomycin dosing in neonates is challenging due to high pharmacokinetic variability and immature renal function. This study evaluated current therapeutic drug monitoring (TDM) practices, the association between vancomycin concentration and clinical outcomes, and the predictive performance of a locally developed population pharmacokinetic (popPK) model compared to the published models. Methods: This was a retrospective cohort study of neonates admitted to a tertiary neonatal intensive care unit (NICU). We assessed the persistent positive culture, infection recurrence, mortality and acute kidney injury (AKI) stratified by initial vancomycin trough concentrations (15 mg/L). The locally developed popPK model was externally validated and compared with 32 other published neonatal vancomycin popPK models (with a total of 33 models evaluated). Results: A total of 366 neonates were included (mean postmenstrual age of 28.9 ± 3.81 weeks; 191 received at least 5 days of vancomycin). Only 28% of neonates achieved initial vancomycin trough concentrations within the 10–15 mg/L using standard vancomycin dosing. Higher vancomycin trough concentrations (>15 mg/L) were not associated with improved efficacy but were significantly associated with a higher incidence of AKI. The locally derived popPK model demonstrated superior predictive accuracy, meeting all predefined performance criteria, whereas none of the 32 other published models met all the criteria. Conclusions: Current vancomycin dosing strategies often result in suboptimal exposure and increased nephrotoxicity without added efficacy. Model-informed precision dosing using the locally developed popPK model may offer a safer, more effective approach for neonatal vancomycin therapy.
Chung et al. (Wed,) studied this question.