The purpose of this study is to determine whether a catheter dressing every 7th day is not inferior to a catheter dressing every 3 days and if Chlorhexidine impregnated sponges are effective in preventing catheter-related infections in ICUs.

Central venous catheters (CVCs) are often required for the care of patient admitted to the intensive care unit (ICU), and are now indispensable in modern-day medical practice. In the United States, it is estimated that 15 million CVC days occur each year in the ICU , and that approximately 80,000 CVC-associated bloodstream infection (BSIs) occur each year [2]. Data from the NNIS system indicate that approximately 40% of the BSIs are associated with a CVC in the ICU. This definition, however, include CVC-related BSIs (CRBSIs) and primary BSIs. In other multicenter surveys, primary BSIs are the leading cause of BSIs (30-35%), followed by CRBSIs (20-30%), and BSIs originating from pneumonia (20%) .

The attributable mortality of CRBSIs remains debated. It ranges from no increase in mortality in some studies, up to an attributable mortality of 35% in others. In studies adjusting for severity of illness, attributable mortality ranged between 0 and 11.5%. The excess ICU length of stay is estimated 9-12 days.

The cost of CRBSIs is therefore substantial, and efforts are required to reduce the incidence of theses infections. Several publications suggested that multiple strategies should be implemented concomitantly. Besides the critical importance of staff education, technology brings new materials that could decrease the risk for CRBSI. Several studies have demonstrated that antimicrobial- or antiseptic-impregnated CVCs can decrease CRBSIs in the ICU setting. Furthermore, cost-benefit analysis have suggested that the use of impregnated CVCs was beneficial

The recent CDC Guidelines for the prevention of intravascular catheter-related infections recommend the use of antimicrobial- or antiseptic-impregnated CVCs in patients whose CVC is expected to remain in place for more than 5 days, and in ICUs where CRBSI rate remains above the benchmark rates, despite implementing a comprehensive strategy. This restricted recommended use is explained by the concern for emergence of resistance, the risk of adverse effects and the costs of these materials.

CRBSI rates in France could be lower than those observed in the United States. Data from two surveillance networks indicate that the rates of CRBSI range between 1 and 2 CRBSI per 1000 CVC days . Given these low rates, it is not clear that antimicrobial- or antiseptic-impregnated CVCs would be cost-effective.

Since most organisms responsible for CRBSI originate from insertion site in short-term CVC, there was a rationale to try to decrease bacterial colonization at cutaneous insertion site. Among the other new materials under development, a chlorhexidine-impregnated sponge (Biopatch TM), to be placed over the site of catheter insertion, has been proposed. In a prospective, controlled, bicenter, randomized, non blinded study, dressing changes every other day (control group) was compared to dressing changes every 7 days with Biopatch (Biopatch group) (Maki and al., ICAAC 2000). 1,401 lines (either CVCs, peripheral arterial catheters or pulmonary artery catheters) were included in 589 patients. Both groups of patients were comparable. Using proportional hazard models, both CVC colonization and CRBSI were significantly reduced in the Biopatch group, from 29% to 16% (HR, 0.62) for catheter colonization, and from 3.3% to 1.2% (HR, 0.38) for CRBSI.

This study demonstrated a significant reduction of CRBSI using Biopatch. Given the results presented at the ICAAC sessions, there is some concern, however, about the validity of the protective effect of the Biopatch.

Firstly, the intervention group associated Biopatch and the extension of the time between dressing changes, from 2 to 7 days. Preliminary data from cancer patients suggest that time between dressing changes could be extended. In a randomized study, Benhamou et al have recently compared a 4-day to a 15-day catheter-dressing change frequency in children undergoing chemotherapy. They have shown that skin cultures (27 vs 23%) and bloodstream infections (11 vs 13%) rates are not different between the 4-day and the 15-day groups. It is therefore unclear that the reduction of CRBSI observed in the Biopatch group was only due to the Biopatch.

Secondly, the control group in the Maki's study did not use a "placebo", i.e. a sponge not impregnated with chlorhexidine. The study was therefore not blinded for the ICU staff. It is strongly recommended to examine the catheter insertion site daily for local inflammatory signs. Biopatch impede to monitor the insertion site, with a potential for underestimation of local infections signs in these patients. It is possible that daily examination of the insertion site in the control group would conduct to remove the CVC more frequently in these patients, with a potential for higher rate of colonization. In addition, if a study is not blinded, it is useful for the validity of the results that a group of investigators, blinded to the randomized group, review the medical chart to classify catheter infection.

Thirdly, the rate if CRBSI was rather high in the control group (4.45 per 1000 line days). It is not certain that the benefit of Biopatch will be the same in ICUs with lower rates of CRBSI.

The aim of this study is therefore to evaluate the impact of Biopatch, and the impact of dressing changes (every 3 or 7 days) on the reduction of CVC infection

Inclusion Criteria:

• patients older than 18 years
• with at least a central venous catheter or an arterial catheter
• whatever the first or subsequent CVC in a same patient
• in any site of insertion (sub-clavian, jugular or femoral)
• whatever le CVC is tunnelled or not
• CVC inserted in the study ICU or immediately before by the intensisvist in the emergency unit or in the operative room,
• CVC inserted under maximal barrier precautions

Exclusion Criteria:

• pulmonary artery catheter, haemodialysis/haemodiafiltration CVCs
• known allergy to chlorhexidine
• CVC not inserted under maximal barrier precautions
• Expected duration of CVC for less than 48 hours
• CVC inserted under emergency conditions