The Pivotal Role of Long‐Term Care Facilities in the Epidemiology ofAcinetobacter baumannii:Another Brick in the Wall

Acinetobacter species are ubiquitous in nature and are widely distributed in soil and water. Various Acinetobacter species differ significantly from each other in the ecological niches they occupy, in their likelihood to cause infections in human, in their virulence traits when causing an infection, and in their antibiotic susceptibilities. The clinicallymost important species of the Acinetobacter genus is Acinetobacter baumannii. Infections with Acinetobacter species were uncommon in the United States in the early 1990s, and Acinetobacter species have not been reported by the National Nosocomial Infections Surveillance system as important nosocomial pathogens [1]. During the 10-year period of 1987–1996, a mean of only 345 nosocomial Acinetobacter infections per year were reported to the National Nosocomial Infections Surveillance system, with a downward trend in incidence during this period [2]. However, by 2006–2007, the National Healthcare Safety Network had reported A. baumannii as the fifth most common gram-negative nosocomial pathogen, causing 8.4% of all episodes of ventilator-associated pneumonia [3]. Of the isolates causing infection, 29% were resistant to carbapenems. Similar trends of an increase in multidrug-resistant A. baumannii as an important nosocomial pathogen have been described in other parts of the world [4]. Infections with A. baumannii occur typically after >2 weeks of hospital stay. They affect severely ill patients with invasive devices who have received multiple antibiotics. The outcomes of the affected patients are often poor, partly because of their underlying diseases and partly because of the acquired infection. Variability in clinical outcome is evident in different geographical areas. This may be related to the type of patients affected, the clinical syndromes caused, or treatment factors, but probably the most important factor is the difference between the infecting strains. Clones of A. baumannii differ significantly from each other in their genetic composition. They vary in pathogenicity islands and in the genetic elements encoding for antibiotic resistance that they carry [5]. The epidemiology of Acinetobacter infections is not well understood. Many of the data were generated when classification of Acinetobacter at the species level was not performed or was performed using unreliable methods. The major differences between clones described above may lead to difficulties in generalizing the conclusions of studies from clone to another. The epidemiology of Acinetobacter species is intriguing; they appear to spread within and between institutions through outbreaks, and although some of these outbreaks are clonal, others are polyclonal and the mode of transmission is not well defined [6]. Current methodology is not sensitive enough to screen for carriage, and although the major site of carriage is believed to be the skin, in hospitalized patients it may be the respiratory tract [7]. The natural history and duration of carriage are not known. The impact of A. baumannii infections is controversial; despite multiple studies showing an association between infections with multidrugresistant A. baumannii and adverse outcomes, debate regarding the contribution of A. baumannii to these adverse outcomes continues, and some individuals even claim that, in most instances, A. baumannii should be considered a contaminant rather than a pathogen. Given this intricate and not well-understood epidemiology, it is difficult to devise strategies to control A. baumannii. To this complex picture, Sengstock et al add another important facet. In this issue of Clinical Infectious Diseases, Sengstock et al [8] describe the pivotal role of long-term care facilities in the epidemiology of multidrug- resistant A. baumannii in southeast Michigan. Residents of long-term care facilities, when hospitalized, were often affected by multidrug-resistant A. baumannii, more than half within 48 h after admission, which suggests that the patients were admitted with these organisms from the long-term care facilities. Although many patients with A. baumannii infection died during their hospital stay, more than two-thirds of those who survived were discharged to long-term care facilities. Thus, the import and export of A. baumannii between hospitals and longterm care facilities play an important role in the epidemiology ofmultidrug-resistant A. baumannii. Long-term care facilities may have a pivotal role in a global strategy for controlling the spread of multidrugresistant A. baumannii, as well as other resistant organisms [9]. They share with acute-care hospitals the same pool of patients, who are often transferred between institutions. These patients typically have multiple risk factors for carriage of multidrug- resistant organisms—for example, prolonged hospitalization, surgical procedures and invasive devices, multiple antibiotic exposures, and low functional status. Thus, long-term care facilities may serve as a reservoir for and amplification source of multidrug-resistant organisms. Any future infection control program in southeast Michigan, or in other areas with similar epidemiology, should be integrative across the spectrum of health care institutions or will likely fail. The study by Sengstock et al [8] is well designed to show the increase in A. baumannii and the contribution of the population of long-term care facilities to this increase. Its design also enables determination of the correlation between resistance and adverse outcomes; however, it is not designed to suggest causality, because it does not have enough individuallevel data to control for differences in underlying conditions, severity of illness, and other possible confounding variables. The study was not aimed to examine therapeutic interventions or the effect of delay in effective therapy or to identify predictors of A. baumannii infection in specific populations. All these are important clinical questions that need to be addressed in future studies. To withstand the tide of multidrug-resistant A. baumannii, we should devise our control strategies on the basis of current knowledge. However, we must realize that this knowledge is still very limited. Understanding of the epidemiology is required to build the wall to limit the spread of A. baumannii. The study by Sengstock et al [8] has provided another brick for this wall. However, new techniques for diagnosis, large studies examining the molecular epidemiology and modes of transmission both in acute-care and long-term care facilities, and an understanding of the mechanisms of resistance and selection pressures are urgently required to devise effective control measures. Potential conflicts of interest. T.d.M. and Y.C.: no conflicts.