RESUMEN
The results of central venous catheterization for total parenteral nutrition were prospectively evaluated in 200 consecutive patients. All catheters were fabricated of polyurethane tubing inserted by the Seldinger technique. Two hundred sixty-three lines were inserted in 200 patients for a total of 4103 days. Major complications occurred in 2.3% patients. Twenty-four per cent of catheters were associated with suspected sepsis; of these, 52% were removed directly and 48% were changed over a guidewire. The total catheter sepsis rate was 5.7%. The incidence of sepsis correlated with the number of attempts to insert the line and with positive skin cultures. These data indicate that: use of the Seldinger technique to insert nonthrombogenic flexible catheters results in lower technical morbidity; the incidence of established infection is much lower than the incidence of suspected sepsis; guidewire change may be performed without risk to the patient or interruption of therapy; sepsis rates can be decreased by reducing the number of attempts to catheterize the subclavian vein; and sepsis rates correlate with positive skin cultures at the insertion site.
