OBJECTIVE: To investigate an outbreak of gram-negative bacteremia in an outpatient hemodialysis unit and to identify the source of contaminating bacteria and the route by which bacteria gained access to the bloodstream. DESIGN: A matched-pair, case-control study and a bacteriologic investigation of the hemodialysis unit and the implicated dialyzers. SETTING: A university outpatient hemodialysis unit. PATIENTS: Eleven patients receiving long-term hemodialysis who had a total of 12 episodes of primary gram-negative bacteremia and 12 matched controls. MEASUREMENTS: Clinical and demographic data were obtained for patients and controls. Dialysis unit procedures were observed for compliance with aseptic technique. Cultures of potential environmental sources of bacteria were obtained. Hemodialyzers from bacteremic and nonbacteremic patients were dismantled, and the component parts were cultured. Inoculation of O-rings (from Hemoflow F-80 dialyzer) with bacteria and simulated dialysis were done. RESULTS: During January to October 1988, 12 episodes of primary gram-negative bacteremia caused by Pseudomonas cepacia, Xanthomonas maltophilia, Citrobacter freundii, Acinetobacter calcoaceticus var. anitratus, or Enterobacter cloacae occurred in 11 patients. In 11 episodes, symptoms developed within 3 hours of starting hemodialysis. Intravenous antibiotics were administered for 11 episodes, 3 episodes resulted in hospitalization, and all patients recovered. Case patients were more likely to have received high-flux dialysis with Hemoflow F-80 dialyzers (odds ratio congruent to 11) than were controls. O-rings from dialyzers used by bacteremic patients were culture positive for the organism responsible for bacteremia. Three of the four dialyzers were disinfected using the standard automated method and were recultured 72 hours later; the O-rings of all three dialyzers remained culture positive. Simulated dialysis using dialyzers with contaminated O-rings caused blood pathway contamination despite intervening reprocessing. When the disinfection method for F-80 dialyzers included removal and complete disinfection of the O-rings, O-ring and blood pathway cultures were consistently negative. After this procedure was made routine, no episodes of primary gram-negative bacteremia occurred during the next 6 months. CONCLUSIONS: Because dialyzers with removable headers and O-rings are widely used in patients receiving long-term hemodialysis, disinfection procedures should include measures to ensure adequate disinfection of O-rings.
OBJECTIVE: To investigate an outbreak of gram-negative bacteremia in an outpatient hemodialysis unit and to identify the source of contaminating bacteria and the route by which bacteria gained access to the bloodstream. DESIGN: A matched-pair, case-control study and a bacteriologic investigation of the hemodialysis unit and the implicated dialyzers. SETTING: A university outpatient hemodialysis unit. PATIENTS: Eleven patients receiving long-term hemodialysis who had a total of 12 episodes of primary gram-negative bacteremia and 12 matched controls. MEASUREMENTS: Clinical and demographic data were obtained for patients and controls. Dialysis unit procedures were observed for compliance with aseptic technique. Cultures of potential environmental sources of bacteria were obtained. Hemodialyzers from bacteremic and nonbacteremic patients were dismantled, and the component parts were cultured. Inoculation of O-rings (from Hemoflow F-80 dialyzer) with bacteria and simulated dialysis were done. RESULTS: During January to October 1988, 12 episodes of primary gram-negative bacteremia caused by Pseudomonas cepacia, Xanthomonas maltophilia, Citrobacter freundii, Acinetobacter calcoaceticus var. anitratus, or Enterobacter cloacae occurred in 11 patients. In 11 episodes, symptoms developed within 3 hours of starting hemodialysis. Intravenous antibiotics were administered for 11 episodes, 3 episodes resulted in hospitalization, and all patients recovered. Case patients were more likely to have received high-flux dialysis with Hemoflow F-80 dialyzers (odds ratio congruent to 11) than were controls. O-rings from dialyzers used by bacteremic patients were culture positive for the organism responsible for bacteremia. Three of the four dialyzers were disinfected using the standard automated method and were recultured 72 hours later; the O-rings of all three dialyzers remained culture positive. Simulated dialysis using dialyzers with contaminated O-rings caused blood pathway contamination despite intervening reprocessing. When the disinfection method for F-80 dialyzers included removal and complete disinfection of the O-rings, O-ring and blood pathway cultures were consistently negative. After this procedure was made routine, no episodes of primary gram-negative bacteremia occurred during the next 6 months. CONCLUSIONS: Because dialyzers with removable headers and O-rings are widely used in patients receiving long-term hemodialysis, disinfection procedures should include measures to ensure adequate disinfection of O-rings.