Piperacillin-tazobactam is an acylureido-penicillin-beta-lactamase inhibitor combination and is frequently used in the empirical treatment of hospital-acquired infections because of its antipseudomonal activity. Similar to other beta-lactam antibiotics, piperacillin-tazobactam exhibits time-dependent killing and the T > MIC appears to be the best outcome predictor. Because a majority of infections are treated empirically, it is necessary to achieve a T > MIC equal to 50% of the dosing interval (50% T > MIC) against the most likely pathogens, including those with only moderate susceptibility The aim of this study is to compare the same dose of piperacillin/tazobactam administered by an extended infusion versus a continuous infusion. A pharmacokinetic study will be performed in patients treated by extended (loading dose 4 G/30 min followed by 4 X 4 G /3h) and continuous infusion (loading dose 4 G/30 min followed by 16G /24h).

A population pharmacokinetic analysis with Monte Carlo simulations will be used to determine 95% probability of target attainment (PTA95) versus MIC

Inclusion Criteria:

• Adult patients (> 18 years) admitted on the intensive care unit (surgical and medical surgery).
• Starting a treatment with piperacillin/tazobactam
• Signed informed consent
• Hematocrit >= 21%
• Available arterial line

Exclusion Criteria:

• age <18 or >75 years
• patient's weight <50 or >100 kg
• renal insufficiency (estimated clearance < 50 ML /MIN)
• haemodialysis
• WBC < 1000 103 µl
• estimated survival <5 days
• meningitis or other proven infections of the CNS
• IgE-mediated allergy to penicillins
• pregnancy
• patients having participated in another study <30 days before inclusion in the present study
• retrospectively, marked deterioration of the renal function during the study period
• retrospectively, treatment < 96 h