Catheter-related infection in critically ill patients.

<bold>Objective: </bold>To describe the incidence of the catheter-related local infection (CRLI) and catheter-related bloodstream infection (CRBSI) of central venous catheters (CVCs) and arterial catheters (ACs).<bold>Design: </bold>Prospective, observational study.<bold>Setting: </bold>A 24-bed medical-surgical intensive care unit of a 650-bed university hospital.<bold>Patients: </bold>We included 988 consecutive patients admitted to the ICU during 18 months.<bold>Measurements: </bold>The incidence density of CRLI and CRBSI, per 1000 catheter-days, of CVC and AC.<bold>Results: </bold>Central venous catheters had a significantly higher incidence density of CRLI (4.74 vs 0.97/1,000 catheter-days; p<0.001) than ACs. Femoral venous access had a higher incidence density of CRLI than subclavian (13.15 vs 1.81/1,000 catheter-days, p=0.003) and than peripheral access (13.15 vs 2.30/1,000 catheter-days, p<0.001). Jugular venous access had a higher incidence density of CRLI (6.29 vs 1.81/1,000 catheter-days, p<0.001) than subclavian access. We found no significant differences in the incidence density of CRLI and CRBSI between the different AC accesses.<bold>Conclusions: </bold>In the CDC guidelines, catheter insertion at the subclavian site is recommended in preference to femoral and jugular accesses, and there is no recommendation about AC site insertion. Our data support these recommendations about CVCs. Because the AC infection rate was very low, our study suggests that the access site is probably not of major importance for this type of catheter. [ABSTRACT FROM AUTHOR]