Personalised dosing of vancomycin: A prospective and retrospective comparative quasi-experimental study.

Abstract

The 2019 update to the US consensus guideline for vancomycin therapeutic monitoring advocates using Bayesian-guided personalised dosing to maximise efficacy and minimise toxicity of vancomycin. We conducted an observational cohort study of the implementation of bed-side Bayesian-guided vancomycin dosing in vascular surgery patients. Over a 9-month prospective study period, vascular surgery patients were dosed vancomycin using Bayesian-guided dosing decision tool (DoseMeRx) and compared retrospectively with a control group admitted to the same ward in the 14 months prior to the study and dosed using a standard algorithmic approach. Primary endpoints were proportion of patients achieving mean area under the curve in 24 hours (AUC24 ) in the acceptable range 350-450 mg/L• h and percentage time in acceptable range (%TTR). Secondary endpoints focused on clinical outcomes including incidence of acute kidney injury. A significantly higher proportion of DoseMeRx patients achieved mean AUC24 values in the acceptable range compared to the control group; 71/104 (68.3%) vs 58/139 (41.7%), P < .005. The median %TTR was also greater in DoseMeRx patients compared to the control group (57.1 vs 30.0%, P < .00001). Patients in the DoseMeRx group missed an average of 0.23 doses per course compared to 1.04 doses in the control group (P < .00001). No difference was observed in secondary (clinical) outcomes between the 2 groups. Bedside Bayesian-guided personalised dosing of vancomycin increases the proportion of patients achieving target AUC24 and the %TTR.