Methods: We retrospectively studied patients with gram-negative bacteremias from 2008-2013 to compare outcomes in those who received PK-PD optimized versus suboptimal cefepime doses after accounting for ARC (estimated GFR>130 ml/min/1.73 m2). The primary outcome was 30-day mortality or microbiologic failure. Secondary outcomes included bacteremia, hospital, and ICU duration. A multivariable logistic regression model was used to identify independent risk factors for microbiologic failure. Cohort assignment was based on cefepime MIC from Vitek-2 and results from a prospective PK study at our institution. Cefepime plasma concentrations from ARC and non-ARC patients were optimally fit to a population PK model and a 10,000-patient Monte-Carlo analysis was performed to estimate PTA of achieving 60% fT>MIC in ARC and non-ARC patients. Chi-square and Mann-Whitney U tests were used.
Results: Among 42 patients, 31.0% had burn injury, APACHE II 16±7, eGFR 106±87, and 35.7% had ARC. Klebsiella spp. (n=13), Pseudomonas (n=8), and E. coli (n=7) were most common. Population PK clearance in 4 ARC patients (eGFR 202±122) was 14.4 L/h and 8.8 L/h in 5 non-ARC patients (eGFR 99±20). In ARC, maximum cefepime doses resulted in a 51.5% PTA for MICs of 8 mg/L. Using these data and MICs, 32 were assigned to the PK-PD optimized and 10 to the suboptimal cohort. There were no differences in the primary outcome (15.6% versus 20%, p=0.74), bacteremia duration (4±6 versus 3±3 days, p=0.66), hospital duration (27±27 versus 51±56 days, p=0.19), and ICU duration (17±25 versus 11±18 days, p=0.36). Urinary source (OR 0.23, 95% CI 0.12-0.42), ARC (OR 2.91, 95% CI 0.55-15.27), coinfection (OR 6.25, 95% CI 1.04-37.7), and APACHE II >16 (OR 9.00, 95% CI 0.97-83.06) were not independent risk factors for microbiologic failure.
Conclusion: After accounting for ARC, no differences in clinical outcomes were noted in patients receiving PK-PD-optimized versus suboptimal cefepime doses for gram-negative bacteremias.
R. L. Arneson,
J. M. Cota, None