Best evidence in critical care medicine: central venous catheterization: the impact of insertion site.

摘要

Background: Central venous catheterization is a common procedure in critical care medicine and anesthesia. Infection impacts patient outcome and, depending on definition, complicates 5-19% of central venous catheter (CVC) insertions. One of the factors hypothesized to affect infection rates is the site of venous cannulation.Design: A three-year prospective observational study comparing the site of non-antimicrobial impregnated venous catheter insertion, and incidence of both catheter related local infections (CRLI) and catheter related blood stream infections (CRBSI). CRLI was defined as any sign of local infection, induration, erythema, heat, pain or purulent drainage, as well as catheter tip colonization. Catheter related blood stream infections were defined as positive peripheral blood culture, catheter tip colonization with the same organism, and signs of systemic infection with no other apparent source.Patients: All patients admitted to a 24 bed intensive care unit during the three-year study period.Intervention: NonePrimary endpoint: The incidence of catheter tip infections (> 15 colony forming units on catheter tip), CRLI and CRBSI.Results: There were 2,595 CVCs in 2,018 patients, for a total of 18,999 catheter days. Catheter insertions were as follows: 917 subclavian CVCs (8,239 total days), 1,390 jugular CVCs (8,361 days) and 288 femoral CVCs (2,399 days). The incidence density (number of infections per 1,000 catheter days) of CRLI for each site was subclavian 1.57, jugular 7.65, and femoral 15.83 (mean 6.05). The incidence of CRBSI for each site was: subclavian 0.97, jugular2.99, and femoral 8.34 (mean 2.79). The femoral site had a significantly higher incidence density of CRLI and CRBSI than either the internal jugular or subclavian. The jugular site had a significantly higher incidence than the subclavian.Conclusion: To minimize the risk of CVC related infection, the preferred order of insertion, for non-antimicrobial impregnated catheters, should be subclavian first, followed by jugular and then femoral.