Variability of antibiotic concentrations in critically ill patients receiving continuous renal replacement therapy

Abstract

Objectives: In critically ill patients receiving continuous renal replacement therapy, we aimed to assess the variability of antibiotic trough concentrations, the influence of effluent flow rates on such concentrations, and the incidence of suboptimal antibiotic dosage. Design: Prospective, observational, multicenter, pharmacokinetic study. Setting: Four tertiary intensive care units within the multicenter RENAL randomized controlled trial of continuous renal replacement therapy intensity. Patients: Twenty-four critically ill adult patients with acute kidney injury receiving ciprofloxacin, meropenem, piperacillin/tazobactam, or vancomycin during continuous renal replacement therapy. Interventions: We obtained trough blood samples and measured antibiotic concentrations. Measurements and Main Results: We obtained data from 40 dosing intervals and observed wide variability in trough concentrations (6.7-fold for meropenem, 3.8-fold for piperacillin, 10.5-fold for tazobactam, 1.9-fold for vancomycin, and 3.9-fold for ciprofloxacin). The median (interquartile range) trough concentrations (mg/L) for meropenem was 12.1 (7.8–18.4), 105.0 (74.4–204.0)/3.8 (3.4–21.8) for piperacillin/tazobactam, 12.0 (9.8–16.0) for vancomycin, and 3.7 (3.0–5.6) for ciprofloxacin. Overall, 15% of dosing intervals did not meet predetermined minimum therapeutic target concentrations, 40% did not achieve the higher target concentration, and, during 10% of dosing intervals, antibiotic concentrations were excessive. No difference, however, was found between patients on the basis of the intensity of continuous renal replacement therapy; this effect may have been obscured by differences in dosing regimens, time off the filter, or altered pharmacokinetics. Conclusions: There is significant variability in antibiotic trough concentrations in critically ill patients receiving continuous renal replacement therapy, which did not only appear to be influenced by effluent flow rate. Here, empirical dosing of antibiotics failed to achieve the target trough antibiotic concentration during 25% of the dosing intervals.