EVALUATION OF STRATEGIES FOR CENTRAL VENOUS CATHETER REPLACEMENT

Objectives: To assess the consequences of leaving a bacterially colonized central venous catheter in place and to compare the effects of three catheter replacement strategies for catheter repair control in an animal model. Design: Prospective study. Setting: Laboratory and animal facility of a large university. Subjects: Eighteen healthy, female, adult sheep. Interventions: Radiopaque-siliconized elastomer central venous catheters were inserted into the jugular veins and colonized with either Escherichia coli or Staphylococcus epidermidis. After 7 days of infection, the catheters were either: a) exchanged using a guidewire; b) removed and replaced with a new catheter in a new jugular vein site after a 48-hr interval; or c) exchanged using a guidewire and antibiotics (tobramycin, cephaloridine) injected into the catheter. Animals were euthanized 7 days after insertion of the new catheter. Quantitative microbiology was performed on blood samples collected daily from the catheters and a peripheral vein, as well as from catheters and tissue recovered from the sheep at the time of autopsy. Measurements and Main Result: When catheters were changed using a guidewire, they became colonized by bacteria within 48 hrs, and the sheep had embolic pneumonia and vegetative endocarditis at autopsy. Similar consequences were observed when antibiotics were administered into the catheter lumen. If colonized catheters were removed and a new catheter was inserted after a 48-hr interval, recolonization, pneumonia, and endocarditis were not observed. Conclusions: Replacement of a biofilm-colonized central venous catheter over a guide-wire is associated with rapid colonization of the replacement catheter and production of detached, slime-enclosed, antibiotic-resistant aggregates that colonize other catheters or initiate endocarditis or pneumonia by dissemination in the bloodstream.