Breanne Mefford, PharmD,
Pharmacist
University of Kentucky Healthcare
Lexington, Kentucky
Disclosure information not submitted.
Chris Donaldson, PharmD
Pharmacist
University of Kentucky, United States
Disclosure information not submitted.
Brittany Bissell, BCCCP, PharmD, PhD
Clinical Pharmacist
University of Kentucky
Lexington, Kentucky
Disclosure information not submitted.
Katie Wallace, PharmD, BCPS-AQ ID
Pharmacist
University of Kentucky HealthCare, United States
Disclosure information not submitted.
Melissa Thompson Bastin, BCCCP, PharmD, PhD
University of Kentucky HealthCare
Lexington, Kentucky
Disclosure information not submitted.
Title: AKI with AUC vs trough-guided vancomycin monitoring with concomitant beta-lactams in the ICU
Introduction: Piperacillin-tazobactam (TZP), cefepime (FEP), or meropenem (MEM) and vancomycin (VAN) are commonly used in combination for sepsis. Studies have shown an increased risk of acute kidney injury (AKI) with TZP and VAN combinations compared to combination with FEP or MEM. VAN guidelines recommend AUC over trough monitoring to minimize risk of AKI. We sought to determine the association of AKI with optimal VAN monitoring when used in combination with beta-lactams.
Methods: Adult ICU patients from 2015-2019, receiving VAN > 72 hours were included. Patients with cystic fibrosis, kidney transplant, pregnancy, Chronic Kidney Disease Stage 3-5, AKI prior to or within 48 hours of VAN initiation or baseline CrCl of ≤ 30 mL/min were excluded. AKI was defined at any stage per Kidney Disease Improving Global Outcomes. Four cohorts were defined: group 1=FEP or MEM with trough-based VAN, group 2=FEP or MEM with AUC-based VAN, group 3=TZP with trough-based VAN, and group 4=TZP with AUC-based VAN. The Kruskal-Wallis equality-of-populations rank test was used to compare exposures and outcomes in the 4 groups, and multi variable logistic regression was used to model the primary outcome, development of AKI at any stage.
Results: Of 763 total patients, 261 were in group 1, 105 in group 2, 231 in group 3, and 166 in group 4. Median (SD) age was 54 (40-64) years, (p=0.185 for comparison), median (SD) baseline serum creatinine was 0.79 (0.6-1.04) (p=0.206), and median (SD) Charlson Comorbidity Index was 4 (2-6) (p=0.035). Sixty patients (23%) in group 1 had AKI, 25 (24%) in group 2, 83 (36%) in group 3, and 37 (22%) in group 4 (p=0.003). After controlling for Charlson Comorbidity Index, concomitant nephrotoxins, and beta-lactam days, patients in group 3 had a 2.8-fold increase in the odds of AKI of any stage as compared to those in group 1 (aOR 2.38 (95% CI 1.55, 3.64). There was no significant association with group 2 or group 4.
Conclusions: ICU patients receiving TZP with VAN guided by trough monitoring had the highest incidence of AKI among the 4 groups, and a higher likelihood of AKI as compared to patients receiving FEP or MEM with VAN guided by trough monitoring. Clinicians should be cognizant of this risk when evaluating empiric antibiotic choices. These results should be validated prospectively.