Author ORCID Identifier

Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation




Clinical and Translational Science

First Advisor

Dr. Peter E. Morris


Vancomycin is the most commonly prescribed antibiotic for hospitalized patients. Despite this fact and decades of clinical use, clinicians remain challenged to meet dosing targets of this narrow therapeutic index drug as well as minimize the risks of therapy, primarily nephrotoxicity. These concerns are magnified in critically ill patients given their severity. Accordingly, in a series of five clinical studies, we sought to identify optimal methods of vancomycin administration in critically ill patients to maximize efficacy and minimize nephrotoxicity via three techniques: use of continuous versus intermittent infusion, use of first-dose pharmacokinetic calculations to guide dosing, and use of loading doses. (1) To identify the landscape in which vancomycin is being used, we surveyed critical care pharmacists on self-reported vancomycin dosing practices. Ninety four percent (94.2%) of pharmacists reported rarely using continuous infusions and 89.2% rarely using first-dose pharmacokinetic evaluation. Loading doses were more commonly used, but rationale for not using included lack of evidence and concern for acute kidney injury (AKI). (2) Given this hesitation by clinicians, we performed a retrospective cohort study of 449 critically ill patients with confirmed methicillin-resistant Staphylococcus aureus (MRSA) pneumonia and/or bacteremia to test the association of a loading dose of vancomycin (≥ 20 mg/kg) with clinical failure. While we found no difference in clinical failure with use of a loading dose versus not, we also found no difference in AKI. (3) Given that few clinicians reported using first-dose pharmacokinetic evaluation to guide dosing, we performed a retrospective cohort study of 66 critically ill patients to test if first-dose pharmacokinetic evaluation was associated with greater area-under-the-curve (AUC) target attainment at steady state. Indeed, first-dose pharmacokinetic evaluation increased AUC target attainment to 58.6% compared to 32.4% (p=0.033) in those patients who received empiric dosing. (4) Method of infusion may also impact AKI risk in critically ill patients. We performed a systematic review and meta-analysis of vancomycin continuous versus intermittent infusion in critically ill patients. Eleven studies were identified which evaluated 2,123 patients. The risk of AKI was found to be significantly reduced in continuous compared to intermittent infusion: odds ratio 0.47 [95% confidence interval (CI) 0.34-0.65]. Additionally, continuous infusions were associated with 2.63 greater odds (95% CI 1.52-4.57) of pharmacokinetic target attainment compared to intermittent infusion. (5) In order to build from the theme that continuous infusions offer more precise dosing at a lower risk of AKI, we conducted a prospective observational study of 50 critically ill patients receiving continuous infusion vancomycin that consisted of 239 dosing events and 124 vancomycin concentrations. A population pharmacokinetic model was constructed to guide further precision dosing in future studies of continuous infusion vancomycin. These findings support further investigation of early pharmacokinetic evaluation and use of continuous infusions to maximize the precision of vancomycin delivery to critically ill patients and minimize the risk of AKI. Additionally, this work’s blueprint provides an approach for future study of precision dosing of antimicrobials in critically ill patients.

Digital Object Identifier (DOI)