No abstract

From 1990 to 1993, an outbreak of respiratory Acinetobacter baumannii infection occurred in five intensive care units (ICUs) of a tertiary care center. A. baumannii was subsequently isolated from disinfected temperature probes and ventilator circuits. Pulsed-field gel electrophoresis suggested that a single strain accounted for 93% of patient isolates and 88% of environmental isolates. Univariate risk factors for A. baumannii acquisition were tracheostomy (P < .01), ventilation >3 days (P < .01), dialysis (P = .03), Stenotrophomonas maltophilia respiratory colonization (P = .02), parenteral nutrition (P = .05), and enteric feeding (P < .01). Logistic regression analysis showed duration of ventilation and enteric feeding to be independent risk factors. The outbreak strain was relatively antibiotic-susceptible, but the mortality attributable to respiratory A. baumannii acquisition was 23%. Only the APACHE II score was independently associated with increased mortality. Multifaceted control measures, including gas sterilization of temperature probes, terminated the outbreak.