CATHETER-RELATED INFECTIONS ASSOCIATED WITH HOME PARENTERAL-NUTRITION AND PREDICTIVE FACTORS FOR THE NEED FOR CATHETER REMOVAL IN THEIR TREATMENT

The inpatient and outpatient records were reviewed for 527 patients, including 138 children, who were discharged on home total parenteral nutrition and followed by the University of California, Los Angeles, home total parenteral nutrition service for a minimum of 1 week between April 1973 and October 1991. The total follow-up time was 1154 patient years; the median follow-up time was 206 days (range 7 to 6344 days). Thirty-six patients were followed for more than 10 years (median 12.7 years). Three hundred fifteen adults (but only four children) were never infected. A total of 427 catheter-related infections occurred, giving an overall infection rate of 0.37 per patient year (0.51 per patient year in children and 0.28 per patient year in patients followed for more than 10 years). Seventy percent of the infections overall were sepsis (0.26 per patient year); of those, 17% were exit site (0.06 per patient year), and 2% were tunnel (0.01 per patient year) infections. In children, 67% of the infections were sepsis (0.37 per patient year) and 24% were exit site (0.13 per patient year). Fifty-four different organisms were responsible, and 12% of the infections were polymicrobial in origin. Sixty percent of sepsis in all patients was caused by Gram-negative organisms (44% of the catheters were removed). Forty-three percent of sepsis in children was caused by Gram-negative organisms. Twenty-six percent of sepsis in all patients was caused by Gram-positive organisms (40% of the catheters were removed). Thirty-six percent of sepsis in children was caused by Gram-positive organisms. Fourteen percent of sepsis was fungal in origin in both adults and children (92% of the catheters were removed). Seventy-four percent of the exit-site infections were caused by Staphylococcus aureus (38% of the catheters were removed). Fifty-seven percent of the tunnel infections were caused by S aureus (100% of the catheters were removed). Chi2 analysis demonstrated that the presence of fever, leukocytosis, bandemia, left shift, organism, Gram status, polymicrobial infections, or organism species (except for fungi and mycobacteria) could not reliably predict the need for catheter removal in any type of catheter-related infection. Catheter infection is uncommon in home total parenteral nutrition patients, although it is more common in children than in their adult counterparts. This, as well as the overall incidence of catheter infection in home total parenteral nutrition patients, may relate to the protocol and technique of catheter care.