Determination of the insertion technique
The insertion technique (with anatomical landmarks or ultrasound-guided) will be determined by randomization. To that aim, we will use a software available on an intranet server and available from all ICUs involved in the study.
All central venous catheters included in the study will be inserted by experienced physicians trained to the use of ultrasound for the placement of access (minimum 2 previous CVC installations with ultrasound).
The central catheters used in the study (Arrow International, Inc.) are single, double or triple lumen (16 and 20 cm in length). The choice will be made by the clinician in patient management. Catheters impregnated with antibiotics or antiseptics are excluded.
The ultrasound used will be a Site\~Rite ® V (C.R. Bard Inc) type equipped with a multifrequency probe of 5-10 MHz.
Acts and examinations performed in the protocol:
* Systematic culture of the CVC and achievement of blood cultures in cases of hyperthermia
* Systematic ultrasound examination at D4 to look for thrombosis.
The examinations and the necessary samples required to the strict follow-up of the patients will be made in the ICUs participating in the study and sent to the usual laboratories for analysis. Ultrasounds will be performed by physicians of the departments participating in the study.
The primary endpoint is the rate of complications related to a composite criterion incorporating mechanical, thrombotic and infectious complications.
The secondary endpoints are:
* Rate of failure / success when placing catheters (failure is defined by three skin biopsies which did not lead to the laying of the venous access).
* The time required for the installation of CVC (time between the first puncture and the introduction of the catheter)
* The number of punctures before final laying
* Type of complications according to the method and site of installation
* The overall mortality according to the technique used
* The duration of stay in ICU / hospital
We define complications as follows:
* Mechanical complications:
* Number of failed insertion,
* Number of bleeding which required a compression for more than 30 minutes,
* Number of arterial wound,
* Number of misposition visualized on the radiological picture,
* Rare complications (pneumothorax, hematoma, air embolism).
* Infectious complications (according to the recommendations of the Twelfth Conference of Society Consensus resuscitation French updated in 2002):
* Colonization of the catheter: Positive culture (≥ 1 000 CFU / ml) according to the quantitative method described by Brun-Buisson of the catheter's distal end (4-5 cm) aseptically removed. This technique can diagnose colonization induced not only by subcutaneous infection, but also by intra vascular pathway. However, it does not differentiate the mode of colonization.
* Catheter-related infection:
* In the absence of bacteremia, diagnosis of catheter-related infection will be based on:
* a positive culture (\> 1 000 CFU / ml) and total or partial regression of signs of infection within 48 hours after catheter removal or
* purulence of the inlet of the catheter or
* tunnelite.
* Bacteremic infection related to the C.V.C. will be defined by:
* the association of one positive bacteremia occurring within 48 hours before or after the withdrawal of CVC (two in case of Staphylococcus epidermidis) and a positive culture of the insertion site by the same organism or a culture of CVC \> 1 000 CFU / ml of the same organism or
* a quantitative central / peripheral blood cultures ratio \> 5 or
* a differential blood culture positivity delay \> 2 hours.
* The infection is NOT related to CVC if :
* the CVC is finally sterile or
* the CVC culture is positive, but the strain is different from that isolated in the blood and / or other focus of infection present at the removal of the CVC and sepsis does not regress to the removal of the CVC or
* the CVC culture is positive and the strain is identical with that found in other infected site found at least 48 hours before removal of the CVC whether or not responsible for bacteremia and sepsis does not regress after the removal of the CVC.
* Thrombotic complications:
A venous Echo-Doppler exploration is performed within four days following the catheter ablation to look for a loss of compressibility of a venous segment under the pressure of the probe associated to an intravascular image measuring more than 3 mm in the anteroposterior diameter (maximum diameter).
