Since the discovery of streptomycin in 1944, aminoglycosides retain a remarkable bactericidal activity vis-à-vis including aerobic gram-negative bacilli. Thus, their synergistic effect with beta-lactams and their rapid bactericidal on many make unavoidable pathogens and make it a cornerstone of the treatment of patients with severe sepsis or state of septic shock.

This is antibiotics exclusively parenteral administration. Their effectiveness is concentration-dependent and are administered by 30-minute infusion. Tolerance of venous is usually excellent. Their potential nephrotoxicity or cochleovestibular toxicity requires accurate monitoring of antibiotic residuals.

Moreover the fact that the effectiveness of the aminoglycosides is concentration dependent, the rate at the peak is decisive. A first sub-therapeutic dose leads to adaptively resistant bacteria compared to the aminoglycoside and therefore an increase of Minimal Inhibitory Concentrations (MIC). Many studies have been conducted in patients hospitalized in intensive care, highlighting underdoses in aminoglycosides when the prescribed dosages consistent with those used in non reanimated patients. Dr Moore showed in 89 ICU patients with bacteremia gram-negative bacilli, the relationship between the clinical course and obtaining whether therapeutic levels during the first administration of aminoglycosides. Thus, mortality in patients whose antibiotic concentrations to peak were subtherapeutic, amounted to 20.9% against 2.4% when concentrations were within the therapeutic range. In the context or an initial peak in the PK / PD ( Pharmacokinetic / Pharmacodynamic) objectives namely Cmax / MIC ≥ 8-10 desirable, individualized therapeutic drug monitoring and identification of factors that may cause a concentration of antibiotic at sub-therapeutic peak seems necessary , in patients for the majority an increased volume of distribution.

In addition to the β-lactams and glycopeptides, due to the increased volume of distribution in critically ill patients in sepsis, evaluation of serum 24 hours after starting treatment to check that the PK / PD goals for these molecules is achieved.

Primary / secondary objective

Investigators propose to conduct a study with the goal:

• Evaluate the rate of patients for whom efficacy endpoint PK / PD Cmax / MIC ≥ 8-10 is reached at the first dose of the usual doses.

• Compare the 30-day mortality among patients with subclinical a first rate versus those who have reached the desired peak

• Evaluate the rate of patients for whom PK / PD efficiency target for related antibiotics is reached:

  • Serum residual> 4X MIC for the β-lactam rate and continuous function of β-lactam and the germ.
  • Serum between 25 and 35 mg / L vancomycin continuously.

• Assess the factors associated with obtaining a rate of aminoglycoside subtherapeutic peak in a population of critically ill patients.

• Assess the residual to 12h after injection to an anticipation of residual dosed at 24.

Methodology :

• Type Of study:
• Prospective, single-center type of professional practice evaluation
• Expected duration of the study six months
• Objective inclusion: 50 to 100 patients. Computing the number of patients to be included was done via the BioStaTGV software.
• Description Of the study (according to recommendations):
• Admission of patients in intensive care unit
• Levy peak aminoglycoside 30min after the end of the 30 minutes of infusion the first administration
• Collection of residual aminoglycoside to + 12h and between 23h-30 and 24 hours after injection.
• Extraction of the residual H + 24h, 30 min before injection to the β-lactam associated administered discontinuously or without time constraints if administration batchwise to the β-lactam and vancomycin.
• Routing unsettled and not frozen tubes in <2 hours of collection
• Reception and registration levy of the laboratory assay of anti-infective under his usual care as recommended service
• Determination of aminoglycoside and vancomycin immunoturbidimetric method (Indiko PLUS)
• Β-lactam assay by LC-MS
• Rendering the usual result (result server consultation for partial and report PDF available on the patient's medical record DXcare).
• Director-post of MIC of aminoglycoside and β-lactam vis-à-vis the offending germ for each bacterial strain isolated and documenting infection
• Destruction Levies: levies will not be retained after being used for the study. They will be destroyed in accordance with the rules of good practice research laboratories.

Data Monitoring and support:

Patient data will be collected on a computer CRF including clinical, biological and microbiological data of patients via the patient's medical record on DXcare: demographics, weight, height, BMI, hydration status, kidney function, liver function, shock, chronic heart failure, malnutrition, hypoalbuminemia burned, cirrhosis, chronic rheumatic disease, other antecedents, pregnancy, reason for hospitalization, etiology of sepsis motivating the introduction of aminoglycosides, β-lactam or glycopeptide and also other associated treatments. The data collected will help identify risk factors associated with a sub-therapeutic dosing aminoglycosides, β-lactam or glycopeptide.

Inclusion Criteria:

• Age w< 18 years old
• admitted in ICU unit for whom aminoglycoside treatment for severe infection was prescribed

Exclusion Criteria:

• Age less than 18 years
• Treatment with aminoglycoside off label
• Patient non hospitalized in intensive care

All adult patients ≥ 18 years admitted to the intensive care unit of GH Paris Saint-Joseph and who will prescribe the entrance or during their stay an aminoglycoside treatment for severe infection