Background: Meropenem plasma concentration above a pathogen's MIC over the whole dosing interval (100% ƒT .MIC ) is a determinant of outcome in severe infections. Significant variability of meropenem pharmacokinetics is reported in ICU patients.Objectives: To characterize meropenem pharmacokinetics in variable CL CR or renal replacement therapy and assess the appropriateness of recommended regimens for MIC coverage.Methods: A pharmacokinetic analysis (NONMEM) was conducted with external model validation. Patient characteristics were tested on meropenem clearance estimates, differentiated according to the presence/absence of continuous renal replacement therapy (CRRT, CL CRRT or CL no-CRRT ). Simulations evaluated the appropriateness of recommended dosing for achieving 100% fT .MIC in 90% of patients.Results: A total of 101 patients were studied: median 63 years (range 49-70), 56% male, . 32% had a CL CR .60 mL/min, 49% underwent CRRT and 32% presented severe sepsis or septic shock. A total of 127 pathogens were documented: 76% Gram-negatives, 24% Gram-positives (meropenem MIC 90 2 mg/L, corresponding to EUCAST susceptibility breakpoint). Three hundred and eighty plasma and 129 filtrate-dialysate meropenem concentrations were analysed: two-compartment modelling best described the data. Predicted meropenem CL no-CRRT was 59% lower in impaired (CL CR 30 mL/min) compared to normal (CL CR 100 mL/min) renal function. Simulations showed that recommended regimens appropriately cover MIC 90 in patients with CL CR ,60 mL/min. Patients with CL CR of 60 to ,90 mL/min need 6 g/day to achieve appropriate coverage. In patients with CL CR !90 mL/min, appropriate exposure is achieved with increased dose, frequency of administration and infusion duration, or continuous infusion.Conclusions: Recommended meropenem regimens are suboptimal in ICU patients with normal or augmented renal clearance. Modified dosing or infusion modalities achieve appropriate MIC coverage for optimized antibacterial efficacy in meropenem-susceptible life-threatening infections.

International audienceWeighted event graphs (in short WEG) are widely used to model industrial problems and embedded systems. In an optimization context, fast algorithms checking the liveness of a marked WEG must be developed. The purpose of this paper is to develop a sufficient condition of liveness of a WEG. We first show that any unitary WEG can be transformed into a graph in which the values of the arcs adjacent to any transition depend on the transition. Then, a simple sufficient condition of liveness can be expressed on this new graph and polynomially computed. This condition is shown to be necessary for a circuit with two transitions

No abstract