Complications of central venous catheterization in critically ill children.

OBJECTIVE

Analysis of central venous catheterization complications in different access sites with the Seldinger technique. Patients and Methods: A prospective study (May 1992 through December 1996) of 308 central venous catheterizations in children of different ages in a pediatric intensive care unit. RESULTS: Access sites were the subclavian vein (76.3%), femoral vein (20.4%), and jugular vein (3.2%). The frequency of catheter placement-related complications was 22%, and the frequency of serious catheter placement-related complications was 2.9% (pneumothorax 1.9%, hemothorax 1%). Catheter placement-related complications were more common in the subclavian than in the femoral vein (chi-square, p =.02) for the larger bore catheters (chi-square, p =.01) and for the higher number of attempts (Student's t -test, p <.001). Catheter placement-related complications were not related to the age, weight, or whether the procedure was performed by the staff physician or resident. The overall complication rate for maintenance-related complications was 36%. Maintenance-related complications were more common in younger children (Student's t -test, p =.03). The most frequent maintenance-related complications were mechanical complications (catheter obstruction and central venous thrombosis), and these were higher for femoral access (chi-square, p <.01) and for catheters indwelling for a longer period of time. Infection was found in 5.8% of patients, mainly due to Staphylococcus epidermidis. Infection was not related to the site of venous access or to the length of time the catheter was left indwelling. CONCLUSIONS: Central venous catheterization can be performed readily in children of all ages with an acceptable degree of risk. The immediate complications were more frequent and severe for subclavian vein catheterization, and the highest risk factor was the number of attempts at catheter insertion. Although the most frequent late complications were mechanical, which were higher for the femoral access and long-indwelling catheters, femoral catheters can be left indwelling for longer periods if routine ultrasound follow-up is performed. Infectious complications were independent of the venous access site or the duration of catheterization.