395 - Impact of Pharmacist-Driven Vancomycin Dosing Utilizing Area Under the Curve (AUC) Software in a Neonatal Intensive Care Unit (NICU)

Publication Number: 3382.395

Irene R. Kelly, UVA Health Children's, Crozet, VA, United States; Irene Kelly, UVA Health Children’s, Crozet, VA, United States; David A. Kaufman, University of Virginia School of Medicine, Charlottesville, VA, United States

Background: Pharmacist-driven antimicrobial dosing and monitoring under a collaborative practice improves therapeutic outcomes while decreasing rates of supratherapeutic concentrations and acute kidney injury (AKI) in adult and pediatric patients, but data lacks in neonates. Additionally, pharmacokinetic studies show that an AUC/MIC ratio of 400 - 600 mg*h/L has less nephrotoxicity and similar effectiveness to trough-based dosing.

Objective: To compare the outcomes of pharmacist led dosing and monitoring of vancomycin utilizing AUC software (400-600 mg*h/L) (POST) vs clinician driven dosing using trough monitoring (8 to 15 mg/L) (PRE) with pharmacy assistance.

Design/Methods: This retrospective study evaluated neonates in a Level IV NICU treated with vancomycin 12 months (03/24 to 03/25) before (PRE) and 6 months (04/25 to 09/25) after (POST) implementation of a pharmacy driven consult service using Bayesian software. This was instituted with a pharmacy consult to dose and monitor therapeutic drug levels of vancomycin. We compared groups using descriptive statistics with significance at p< 0.05.

Results: Patient characteristics and clinical variables were similar for the 55 dosing episodes in the PRE versus 29 POST episodes (Table 1). Vancomycin was mostly initiated for Gram-positive infections. Verification time was < 15 minutes and doses were administered within < 60 minutes for most with the number of dose changes and levels obtained being similar between groups (Table 2). The initial and final dose was also similar, although the POST group was more likely to have the same initial and final dosing interval. While AKI occurred in some, creatinine levels decreased significantly during treatment in both groups (Table 3). The POST model detected supratherapeutic levels in 25% compared to 13% in the PRE period. There were 4 patients with a recurrent infection (PRE) vs 0 (POST). Survival of culture proven infections for all was ≥96%.

Conclusion(s): The pharmacist-driven AUC model was more likely to select the correct dosing interval and identify appropriate dose adjustments. Collaborative agreements of clinicians and pharmacists for vancomycin adds drug safety and improves outcomes. Some outcomes may not have changed due to active pharmacy involvement in the pre-implementation period. Further inclusion of neonates in this study evaluating risk factors that may affect dosing, nephrotoxicity and clinical outcomes will add to the neonatal data of pharmacy driven dosing and monitoring using AUC/MIC.

Table 1: Patient Characteristics and Infection Outcomes.
aOf 21 repeat cultures; bOf 10 repeat cultures; cOnly representative of those discharged at time of data collection
Median (range); PDA, patent ductus arteriosus. NEC, necrotizing enterocolitis. HTN, hypertension. CoNS, coagulase-negative staphylococcus, MRSA, methicillin resistant Staphylococcus aureus. MSSA, methicillin sensitive S aureus. GPC, gram positive cocci

Table 2: Vancomycin regimen and therapeutic monitoring
Therapeutic trough goal was 8-15 mcg/mL; therapeutic AUC level = 400-600 mg·h/L
TDM - therapeutic drug monitoring; median (range)

Table 3: Safety
aOf 39 patients who got ≥1 TDM; bOf 36 positive cultures; cOf 24 patients who got ≥1 TDM; dOf 12 positive cultures.
TDM=therapeutic drug monitoring; median (range)