To THE EDITOR—We congratulate Fournier and Richet [1] on their insightful article on the hot topic of the epidemiology and control of Acinetobacter baumannii infection in health care facilities. We believe that the authors' statement that “potentially severe A. baumannii infection, such as bacteremia or pneumonia in patients in the intensive care unit [ICU] who are undergoing intubation, do not seem to be associated with a higher attributable mortality or increased length of hospital stay” [1, p. 692] needs some clarification. The authors referred to 2 matched cohort studies performed in critically ill patients by Blot et al. [2] and Garnacho et al. [3]. In both studies, the observed difference in mortality between patients with and without A. baumannii infection was not statistically significant. However, in the study by Blot et al. [2], an ICU stay in excess of 5 days was found in patients with A. baumannii bacteremia (the median length of ICU stay for patients with and without A. baumannii bacteremia was 25 days and 20 days, respectively). This finding was statistically significant (P = .04). García-Garmendia et al. [4] also performed a matched case-control study that compared the outcomes of patients hospitalized in the ICU with and without A. baumannii acquisition (defined as infection or colonization). We extracted the data from this study regarding the outcomes of the subsets of patients with A. baumannii infection (we excluded patients with A. baumannii colonization). The crude ICU mortality among patients with and without A. baumannii infection was 58% and 15%, respectively—a statistically significant difference (P < .001)—resulting in an attributable mortality for A. baumannii infection of 43%. In addition, the median length of ICU stay was 13 days longer for patients with A. baumannii infection. This finding was also statistically significant (P < .001). Furthermore, Weingarten et al. [5] performed a matched case-control study that compared patients, most of whom were hospitalized in ICUs, with and without A. baumannii acquisition. We also extracted the relevant data from this study regarding the subsets of patients with A. baumannii infection. Although data regarding the attributable mortality for A. baumannii infection were not available, a statistically significant longer ICU stay was found among patients with A. baumannii infection. We recently performed a systematic review of relevant matched cohort and case-control studies to evaluate the attributable mortality of A. baumannii infection among critically ill patients [6]. In the 6 matched cohort and case-control studies included in our review, we found that the attributable in-hospital mortality among patients with A. baumannii infection was 7.8%–23% and that the ICU mortality was 10%–43% [2–4, 7–9]. Although definitive statements regarding the attributable mortality of A. baumannii infection cannot be made from the available studies because of their methodological heterogeneity, we believe that the data from the relevant studies suggest that infection with A. baumannii may be associated with considerable attributable mortality and increased length of ICU stay. Potential conflicts of interest. All authors: no conflicts.
No takes yet. Share an insight, caveat, or question.
Falagas et al. (2006) studied this question.