Sepsis is a significant complication of central venous catheter (CVC) use in children. Various techniques to decrease infection risk, such as site preparation with chlorhexidine, full-barrier precautions during insertion, specialized care teams, disinfection of ports before using the line for blood sampling, antiseptic-impregnated dressings, systemic antibiotic prophylaxis, and local application of antibiotic ointment, have been tried with varying degrees of success. Subcutaneous tunneling of short-term CVCs increases the distance from the insertion site to the vein opening, resulting in decreased infection rates among adults at the jugular and femoral sites but not at the subclavian site. Nahum and associates studied subcutaneous catheter tunneling among children in intensive care units who required CVC placement for 48 hours or more at the femoral vein site.

Using full-barrier precautions and site preparation with chlorhexidine, 98 participants were randomized to receive either tunneled or nontunneled catheters and completed the full protocol. In the tunneled technique, a cannula was used to create a 3- to 4-cm subcutaneous tunnel before femoral vein insertion. Catheters were removed when they were no longer needed or when a significant related complication occurred. All CVC tips were cultured for bacteria.

The colonization rate was significantly higher in the nontunneled catheters than in the tunneled catheters. The most common colonizing organisms were coagulase-negative staphylococci. Bloodstream and exit-site infections were not significantly different between the two groups. No cases of tunnel infection occurred. Although one participant developed a subcutaneous hematoma that spontaneously resolved, there were no early or late complications among children in the tunneled group.

The authors conclude that using a tunnel technique to place femoral CVCs in very ill children is safe and effectively reduces the catheter colonization rate but not the bloodstream infection rate. Presumably, if catheters had been left in longer than the mean of six days among participants in this study, a significant reduction in bloodstream infections would have been seen. Further studies are needed to confirm these findings.