• John R Gowardman, Iain K Robertson, Scott Parkes, and Claire M Rickard.
• Department of Intensive Care Medicine, Royal Brisbane and Woman's Hospital, Herston, Brisbane, QLD 4029, Australia. john_gowardman@health.qld.gov.au
• Intensive Care Med. 2008 Jun 1;34(6):1038-45.

ObjectiveTo compare colonization and catheter-related bloodstream infection (CR-BSI) rates among three insertion sites (subclavian, internal jugular, femoral) used for central venous catheter (CVC) placement.DesignTwenty-four-month prospective study, with relative effects analyzed by Cox proportional hazards regression.SettingEight-bed intensive care unit.PatientsFour hundred and ten critically ill patients requiring CVC placement.Measurements And ResultsAll short-term multi-lumen CVCs, including antimicrobial-coated devices, were studied with management standardized. Six hundred and five CVCs (4,040 catheter days) were analyzed. Colonization and CR-BSI incidence were, respectively, 15.1 (95% CI 13.5-21.0) and 1.8 (95% CI 1.2-4.2) per 1,000 catheter-days. Colonization was higher at the internal jugular (HR 3.64; 95% CI 1.32-10.00; p=0.01) and femoral (HR 5.15; 95% CI 1.82-14.51; p=0.004) sites than at the subclavian site. The femoral site carried a greater risk of being colonized by non-S. epidermidis species than the subclavian and internal jugular sites combined (HR 4.15; 95% CI 1.79-9.61; p=0.001). CVCs inserted in the Department of Emergency Medicine were more colonized than those inserted in the ICU or operating room (HR 2.66; 95% CI 1.27-5.56; p=0.01), and CVCs were less colonized in females than in males (HR 0.49; 95% CI 0.26-0.89; p=0.02). No difference in CR-BSI rates was noted between the three sites.ConclusionsColonization was lowest at the subclavian site. Regional differences exist with respect to type of pathogen isolated. Colonization was influenced by insertion location and gender. The incidence of CR-BSI was not different.

21 keywords...

hide…