Abstract 450
Poster Session III, Monday, 5/3 (poster 338)
Catheter-related sepsis (CRS) is the most common complication of percutaneously placed central venous catheters (PCVC). To minimize CRS rates in the newborn infants, a unique PCVC program was instituted at Michael Reese Hospital (MRH) since 1983. Its components were: a) a set of dedicated operators (four of them for all fifteen years) placing all PCVCs; b) catheters to be replaced every four weeks; c) the dressings to be changed only when there was a leak; d) blood cultures to be obtained at the time of insertion, removal of the catheter and at two week intervals in between; and e) ongoing monitoring of PCVC infection data. To assess the impact of such a program, we reviewed CRS rates during 1992-97 at MRH and compared them with: a) the 1995 CDC data; and b) a data set derived from a limited meta-analysis of sixteen publications (1990-95). We used the CDC definitions for CRS and catheter colonization rates. Results: 497 PCVCs were placed in 530 infants (mean weight at insertion 1127 grams, range 360-4082 grams, 88% of whom were < 1500 grams) for a total of 6993 catheter-days. The PCVCs were in place for a mean duration of 12.9 days (range 1-41 days). 122/497 (24.7%) PCVCs were colonized; the most frequent organism was Staph. epidermidis; however, only 11/497 (2.2%) of these resulted in CRS. 10 of the 11 episodes were caused be Staph. epidermidis and one due to Staph. aureus. The CRS rate, defined as sepsis per number of catheters inserted, was 2.2% (95% C.I., 3.5, 0.9) and defined as infections per 1000 catheter days, was 1.6 (95% C.I. 2.5-0.2). The median rate reported by the CDC in infants < 1500 grams was 14.6/1000 catheter days, a significantly higher rate than that of MRH. The mean CRS rate computed for infants of all birth weight from the meta-analysis was 6.5% (95% C.I., 5.4, 7.7%, range 0-29%) and 3.9 per 1000 catheter days (95% C.I. 3.17, 4.73, range 0-15.3 per 1000 catheter days). Conclusions: 1) The upper limit of the 95% C.I. for CRS rates per 1000 catheter days in our study was 420% lower than the CDC reported median rate, and 53% lower than those from meta-analysis. 2) We speculate that such low CRS rates were achieved by following a regimentally structured program and by limiting the number of operators handling PCVC related procedures. Such an approach should be more widely practiced and its impact on cost and quality of health care prospectively assessed.
