Continuous infusion of vancomycin in septic patients receiving continuous renal replacement therapy.

Vancomycin is frequently administered as a continuous infusion to treat severe infections caused by Gram-positive bacteria. Previous studies have suggested a loading dose of 15 mg/kg followed by continuous infusion of 30 mg/kg in patients with normal renal function; however, there are no dosing recommendations in patients with renal failure undergoing continuous renal replacement therapy (CRRT). Data from all adult septic patients admitted to a Department of Intensive Care over a 3-year period in whom vancomycin was given as a continuous infusion were reviewed. Patients were included if they received vancomycin for ≥48h during CRRT. Vancomycin levels were obtained daily. During the study period, 85 patients (56 male; mean age 65±15 years; weight 85±24kg) met the inclusion criteria. Median (interquartile range) APACHE II and SOFA scores were 24 (20-29) and 11 (7-14), respectively, and the overall mortality rate was 59%. Mean vancomycin doses were 16.4±6.4 (loading dose), 23.5±8.1 (Day 1), 23.2±7.4 (Day 2) and 23.3±11.0 (Day 3) mg/kg, resulting in blood concentrations of 24.7±9.0 (Day 1), 26.0±8.1 (Day 2) and 27.7±9.3 (Day 3) µg/mL. On Day 1, 43 patients (51%) had adequate drug concentrations (20-30 µg/mL), 17 (20%) had levels >30 µg/mL and 25 (29%) had levels <20 µg/mL. Most patients with adequate drug concentrations received a daily dose of 16-35 mg/kg. The intensity of CRRT directly influenced vancomycin concentrations on Day 1 of therapy.

Determinants of early inadequate vancomycin concentrations during continuous infusion in septic patients.

Vancomycin is frequently administered to critically ill patients by continuous infusion in order to optimise drug efficacy; however, there are few data available on the efficacy of this strategy in septic patients. In this retrospective analysis, 261 patients treated with continuous infusion of vancomycin in the Department of Intensive Care at Hôpital Erasme (Brussels, Belgium) were evaluated. Creatinine clearance (CL(Cr)) was calculated from 24-h urine collection and normalised to body surface area. During the study period, 139 patients (53%) had insufficient vancomycin concentrations (<20 µg/mL) on Day 1 and 87 patients (33%) on Day 2. Patients who had insufficient drug concentrations on Day 1 of therapy were more likely to be men, to have a higher CL(Cr) and to have received lower loading and daily vancomycin doses than other patients, who received greater vasopressor support and had higher Sepsis-related Organ Failure Assessment scores. In multivariate regression analysis, high CL(Cr) and male sex independently predicted the presence of insufficient vancomycin concentrations on Days 1 and 2 of therapy. Receiver operating characteristic curve analysis for CL(Cr) showed an area under the concentration-time curve of 0.75 (95% confidence interval 0.69-0.81) to predict insufficient drug concentrations on Day 1 of therapy. A CL(Cr)>120 mL/min/1.73 m(2) had a sensitivity of 26%, a specificity of 94% and an 84% positive predictive value of 84% for vancomycin concentrations <20 µg/mL. In conclusion, approximately one-half of the septic Intensive Care Unit patients treated with continuous infusion of vancomycin at currently recommended doses had insufficient drug concentrations in the early phase of therapy. A high CL(Cr) was the variable most strongly associated with insufficient drug concentrations.