Implementation of a Pharmacist-Driven Two-Level AUC-Based Vancomycin Dosing Strategy at a VA Medical Center

Background: This project sought to improve patient safety and outcomes by implementing a novel approach to monitor and adjust vancomycin dosages following release of the 2020 consensus vancomycin dosing guidelines. The implementation of vancomycin dosing according to vancomycin AUC concentrations has the potential to reduce associated acute kidney injury by about 50% from previous vancomycin dosing by trough concentrations. The primary outcome was the rate of vancomycin associated acute kidney injury. Secondary outcomes included length of stay, mortality, and desirability of outcome ranking, and pharmacologic outcomes included frequency of target AUC attainment and correlating vancomycin trough values. Methods: This project assessed vancomycin usage before and after implementation of an AUC-based therapeutic monitoring program. Post data comprised four months following implementation of a vancomycin AUC-based therapeutic monitoring program (November 1st, 2020 to February 28th, 2021) and compared to standard of care in the same time period of the previous year (November 1st, 2019 to February 28th, 2020). Inpatient clinical pharmacy specialists (CPS) were responsible for designing vancomycin regimens using a target 24-hour AUC goal of 400-600 mghr/L after the initial dose. Monitoring to estimate the vancomycin AUC, was performed by attaining vancomycin peak and trough levels. The CPS calculated a new vancomycin regimen which was predicted to attain a 24-hour vancomycin AUC level of 400-600 mghr/L according to a Microsoft® Excel®-based calculator using first-order pharmacokinetic equations. Data collected in this evaluation included eGFR and serum creatinine levels to stratify patients according the RIFLE and AKIN classifications of acute kidney injury, length of stay, mortality, desirability of outcome ranking, predicted and attained AUC values, and serum trough concentrations. Results: A total of 408 patients received at least one dose of vancomycin during the pre-implementation period and 602 patients received at least one dose of vancomycin in the post-implementation period. 85 patients in the pre-implementation and 40 patients in the post-implementation period were included in the final analysis. Acute kidney injury (AKI) defined by AKIN and RIFLE criteria occurred in 9-10 (10.5% – 11.6%) and 4-5 (10%-12.5%) of patients in the pre- and post-implementation groups respectively. Acceptable peak and trough vancomycin levels were drawn at steady state for 25 of 39 patients (64%) in the post-implementation group (steady state concentrations were not attained in one patient). Conclusion: A two-level Microsoft® Excel®-based vancomycin AUC-based calculator did not improve rates of AKI, length of stay, or DOOR outcomes compared to trough-only monitoring. While the AUC calculator is able to appropriately calculate vancomycin AUC, it may be unfeasible to conduct two-level vancomycin AUC monitoring at institutions without around-the-clock CPS coverage to facilitate appropriate peak and trough level draws. A one-level Bayesian approach may allow for more reliable vancomycin AUC-based monitoring in all qualifying patients.