Objective: To analyze the rate of infection of de novo, guidewire exchanged, and new site replacement catheters in a cohort of patients in whom catheters were changed on the basis of the clinical discretion of the attending physicians.
Design: We conducted an observational cohort study in catheterized patients in the intensive-care unit (ICU).
Material and methods: ICU patients admitted between Jan. 1, 1991, and Dec.31, 1992, were eligible for enrollment in the study. Catheter care, replacement, and duration were prospectively documented. Catheter-related infection was prospectively evaluated. Rates of catheter-related infection were determined for de novo, guidewire exchanged, and new site replacement catheters and analyzed relative to the duration of placement of individual catheters and the total duration of central venous catheterization for a specific patient.
Results: Fifty catheter-related infections developed in 2,470 patients. When the rate of catheter-associated infection was determined for each type of catheterization, de novo catheters had a lower observed rate of infection than either replacement type (P < or = 0.0001). After controlling for the effect of time, we found that the rate of catheter-related infections associated with a de novo catheter was less than the rate in guidewire exchanged catheters (P = 0.035). Rates of infection were similar between guidewire exchanged catheters and catheters replaced to a new site.
Conclusion: In a population of ICU patients in whom catheter change was governed by clinical judgement, no differences were noted between the observed rates of infection of new site replacement catheters and guidewire exchange catheters.