Arterial catheter-related infections in children. A 1-year cohort analysis.

To determine the incidence of infection secondary to arterial catheterization in children as well as the risk markers, we prospectively evaluated, during a 1-year period, all arterial catheters installed in children admitted to the pediatric intensive care unit. A total of 340 cannulas were placed in 310 children aged 80 +/- 4 months (mean +/- SEM) for a period of 64 +/- 4 hours. Most catheters were inserted percutaneously (99%) in the radial artery (86.5%). Ninety-two percent (313/340) of the catheters were sterile (group 1), 5% (17/340) were contaminated (less than 10 colony-forming units on semiquantitative culture) (group 2), and 3% (10/340) were considered either locally infected (ie, greater than or equal to 10 colony-forming units) (eight of 10) or associated with a possible catheter-related sepsis (two of 10) (group 3, or infected group). The incidence of local inflammation at the insertion site was higher in group 2 than in group 1 (18% vs 2.9%) but not statistically different between groups 3 and 1 (10% vs 2.9%). The duration of arterial catheterization was longer in group 3 than in group 1 (125 +/- 31 vs 61 +/- 4 hours). The risk of infection was nonexistent in the first 48 hours of catheterization. Thereafter it was calculated as being 6.2% (10/161), but it correlated poorly with the duration of arterial catheterization. These results confirm the very low incidence of infection related to arterial catheterization in children. Thus, routine catheter reinsertion is, in our opinion, unjustified.