The effect of catheter type and site on infection rates in total parenteral nutrition patients.

Infections pose a major problem in patients receiving total parenteral nutrition. Controversy continues concerning the effect of catheter type (triple-, double-, single-lumen, or pulmonary artery), insertion site (subclavian, internal jugular, or femoral vein), and the incidence of catheter-related infections. We retrospectively studied multi-lumen catheter use for total parenteral nutrition over a 6-month period in 192 patients, a total of 3334 catheter days. Nonintensive care unit catheters were inserted by the Nutrition Support Service, and intensive care unit catheters were inserted by the intensive care unit staff. All catheters were cared for using Nutrition Support Service protocols, with multi-lumen catheters changed every 7 to 10 days and pulmonary artery catheters changed every 4 days. Infections were determined by semiquantitative cultures (> 15 colonies/plate). The incidence of infections for triple-lumen catheters was 5 (subclavian), 17 (internal jugular), and 36% (femoral) respectively; total infection rate for triple-lumen catheters was 10%. Infection rates for pulmonary artery catheters were 4 (subclavian), and 6% internal (jugular site), respectively, the overall infection rate was 5%. There were no differences in infection rates at any site based on catheter type; however, when triple-lumen catheter sites were compared, the differences were significant (p < .001 vs subclavian, chi 2). Catheter duration was 7.8 days (subclavian),, 7.3 days (internal jugular), and 4.6 (femoral) days. These data suggest that the use of multi-lumen catheters for total parenteral nutrition is safe, that there is a benefit associated with the subclavian route, and that the femoral site should be avoided.