Attributable mortality of nosocomial Acinetobacter bacteremia.

OBJECTIVE: To determine the attributable mortality and outcome of nosocomial Acinetobacter bacteremia. DESIGN: Matched, retrospective cohort study. SETTING: Large, university-based, tertiary care center. PATIENTS: Of 219 patients with nosocomial Acinetobacter bacteremia identified by prospective surveillance during a 3-year period, 52 met the criteria for the study and were matched to a control patient by age, sex, primary and secondary diagnosis, operative procedures, and date of admission. RESULTS: A 100% success rate was achieved in the proportion of case patients and control patients matched for the compared criteria, except for major operative procedures (88%) and the presence of an important secondary underlying disease (54.5%). Twenty-nine (55.7%) of the case patients died, compared with 10 (19.2%) of the control patients (P<.001). The attributable mortality was 36.5% (95% CI, 27%-46%) and the risk ratio for death was 2.9 (95% CI, 1.58-5.32). In a multivariate survival analysis, older age, mechanical ventilation, renal failure, and Acinetobacter bacteremia (hazard ratio [HR], 4.41; 95% confidence interval [CI], 1.97-9.87; P<.001) were found to be independent predictors of mortality. There was a trend for a longer median duration of hospitalization among case patients, compared with control patients (11.5 vs. 6.5 days; P=.06). Three isolates were resistant to all but 1 antibiotic tested (colistin), and 45 isolates (86.5%) were resistant to 4 or more different antibiotic classes. CONCLUSIONS: When adjusted for risk-exposure time and severity of disease at admission, nosocomial Acinetobacter bacteremia is associated with mortality in excess of that caused by the underlying diseases alone.