Introduction

Acinetobacter infections are becoming gradually more common in various settings and populations around the world. Specifically, Acinetobacter baumannii (previously known as Acinetobacter calcoaceticus var. anitratus) has become a common nosocomial pathogen, affecting especially patients receiving treatment in the intensive care unit (ICU) setting [1, 2]. Beyond this population, Acinetobacter affects war and trauma casualties, as it has been shown in soldiers in the Gulf War and the victims of trauma in various earthquakes, including that of Marmara, Turkey, in 1999 [1].

There are several unanswered questions regarding various aspects of Acinetobacter infections, including epidemiological and management issues. Among them is the continuing controversy regarding the mortality that is directly attributable to Acinetobacter baumannii infections in ICU patients, if any [2, 3]. However, there is accumulating evidence that this pathogen may cause serious infection and, thus, contributes substantially to the considerable mortality of this population [4].

Although the attention of Acinetobacter baumannii infections has been focused on hospitalized patients, there is another patient population that may be affected by this important pathogen, namely, patients in the community setting that have some form of morbidity, especially in the tropical and sub-tropical climates [5]. These patients may develop community-acquired Acinetobacter infections. Given the lack of synthesis of data on the issue, we sought to evaluate the available evidence regarding community-acquired Acinetobacter infections by performing a systematic review focused on this population.

Literature search

The relevant studies were retrieved through a search of the PubMed database (articles published after 1973 and archived by June 2006) by using the key-terms: “community AND Acinetobacter”. Two hundred and thirty two (232) studies were initially screened. In order for a study to be included in our review, it should have been written in English, French, German, or Italian. The study should have regarded adult immunocompetent patients. It should be also clear by the study’s authors that the person did not acquire the infection by his/her direct or indirect contact with the hospital environment and that it was with no doubt a community-acquired infection. For inclusion in our review, the study should have presented data regarding the clinical course of the infection and/or the rate of Acinetobacter isolates out of the total number of infective agents that were isolated from specimens taken from infected patients that presented the same type of infection and were being treated in the same clinic. Studies that regarded pediatric patients, hospitalized patients, immunocompromised patients, patients undergoing chemotherapy against malignancy, and patients with external medical devices that required medical attention on a regular basis were excluded from our review.

From the initially screened 232 studies, we excluded 23 that were written in languages other than English, French, German, or Italian. Out of the remaining 209 articles, we selected 59 studies that regarded adult immunocompetent patients with community-acquired Acinetobacter infections. Of those 59 studies, 16 were excluded (four evaluated the pathogen Acinetobacter from a microbiological scope, presenting its microbiological characteristics; three neither presented data regarding the infection’s clinical course nor evaluated the rates of isolates, and, thus, they could not be included in our review; three presented concentrated results that included both community- and hospital-acquired infections; two regarded pediatric patients; two regarded patients undergoing continuous ambulatory peritoneal dialysis; one regarded patients with tunneled central venous catheters undergoing prolonged outpatient intravenous therapy; and, finally, one regarded patients with head trauma, craniotomy, or traumatic hydrocephalus). Thus, we reviewed in detail 43 studies. The references of these studies were also checked in order to ensure that relevant studies were not missed. Two of the authors (EAK and IK) independently performed the literature search and, at the end of their efforts, compared their results.

It was our decision to include studies in which the authors clearly reported that the identified pathogen was Acinetobacter and, thus, studies in which the infective agent was identified as Mima or Herellea were excluded. Finally, we decided to include in our review all studies, regardless of the species of Acinetobacter, but, additionally, to present the species as they were reported by the authors of each study.

Regarding the characterization of the studies as prospective or retrospective, it was decided that, when available, the study’s authors’ characterization of their work would be presented. For the rest of the studies, an effort was made to identify in the studies’ methods section, whether the generation of the collected data preceded (retrospective study) or followed (prospective study) the hypothesis testing.

We extracted data regarding the underlying illnesses, clinical manifestations of the infection, antibiotic treatment, observed complications, clinical outcomes of the patients, and other characteristics of community-acquired Acinetobacter infections from relevant case series and case reports. We also collected data regarding the frequency of community-acquired Acinetobacter infections from other retrospective and prospective studies.

Reviewed studies

In Table 1, we present data from the reviewed case series of patients with community-acquired Acinetobacter infections. It is important to note that, in these studies, the authors evaluated groups of patients with the same type of infection and present their medical history, their infection’s characteristics, and the clinical course, as well as their response to treatment, in the form of rates. There were six studies presenting information regarding a total of 80 patients with community-acquired Acinetobacter baumannii (formerly known as A. calcoaceticus var. anitratus) infections. From those, 51 had pneumonia. Bacteremia occurred in the remaining 29 patients. Several forms of comorbidity were common in patients included in these case series: chronic obstructive pulmonary disease, renal disease, and diabetes mellitus. Heavy smoking and excess alcohol consumption were also commonly present in patients included in the reviewed case series. Of 80 patients, 45 (56%) died of the infection. Finally, the reviewed case series originated from China, Taiwan, and tropical Australia.

Table 1 Case series of patients with community-acquired Acinetobacter infection
Full size table

In Table 2, we present the identified case reports of community-acquired Acinetobacter infections. There were 26 case reports presenting information regarding 43 patients (34 males and 9 females) with community-acquired Acinetobacter infections. In the case reports that presented more than one (up to five) case, the presentation focused separately on every one of the presented patients and reported the patient’s as well as the infection’s characteristics. From those 43 patients, 38 had pneumonia, two had meningitis, one had soft-tissue infection, one had ocular infection, and one had native valve endocarditis. The patient with the ocular infection grew A. junii in the relevant cultures [6]. The cultures of one patient with meningitis grew A. calcoaceticus var. Lwoffii [7]. In three studies (presenting a total of seven patients), the variety of A. calcoaceticus was not clarified [8–10], while in the remaining 21 studies (regarding 34 patients), it was specified that the infection’s causative bacterium was A. baumannii (Acinetobacter calcoaceticus var. anitratus).

Table 2 Case reports of community-acquired Acinetobacter infection
Full size table

Chronic obstructive pulmonary disease, renal disease, diabetes mellitus, alcoholism, and heavy smoking were also commonly recognized as underlying conditions among the patients included in the reviewed case reports that are presented in Table 2. For four patients [74 (two patients), 19, and 18 years old), it was clearly reported by the authors that no underlying condition (including smoking and alcoholism) was present [11, 12, 7, 9]. In 35 out of 38 patients, Acinetobacter pneumonia manifested with an acute course, while for the remaining three patients, the diagnosis of “chronic” pneumonia was given [13–15]. Of the 43 patients included in the reviewed case reports, 19 died of the infection, and, among the 19 deaths, 10 occurred in the first 48 h of hospitalization.

Lastly, in Table 3, we present data regarding the frequency of community-acquired infections that were caused by Acinetobacter in various patient populations and settings. One study contained useful data regarding both the clinical manifestations of community-acquired Acinetobacter infection and its frequency [9], and, thus, data from this study were included in both Tables 2 and 3, respectively. Seven of the 12 enlisted studies originated from the Far East (Taiwan, Malaysia, and Singapore). Two more studies were from Oceania (one from tropical Australia and the other from New Zealand). Finally, one was from N. Guinea, one from Palestine, and one from USA/Canada. The range of isolation of Acinetobacter from patients with community-acquired pneumonia in these studies was 1.3%–25.9%.

Table 3 Frequency of Acinetobacter infections among patients with community-acquired infections
Full size table

Evaluation of the available evidence

The most important finding of our systematic review is that there are several reports in the form of case reports, case series, and prospective and retrospective studies dealing with community-acquired Acinetobacter infections. This information suggests that this pathogen is a considerable cause of community-acquired infections, especially in tropical and subtropical climates.

The published evidence regarding community-acquired Acinetobacter infections mainly originated from countries with tropical or subtropical climate. It is noteworthy that, based on the available data, community-acquired Acinetobacter infections mainly occur in patients with some form of comorbidity, such as chronic obstructive pulmonary disease, renal disease, and diabetes mellitus, and/or are associated with heavy smoking and excess alcohol consumption. There was a scarcity of information in the publications we reviewed regarding the time of the year that the infections occurred.

It should be emphasized that the mortality of patients with community-acquired Acinetobacter pneumonia and/or bacteremia is considerable (56% in the case series included in this review). Although there is a possibility that publication bias is operating favoring the report of patients with severe community-acquired Acinetobacter infections, the reviewed data suggest that these infections are indeed associated with considerable mortality.

Our review has several limitations regarding the search strategy we used. It should be emphasized that we included in our review studies identified in the PubMed database that were written in English, French, German, or Italian. Subsequently, it is very possible that we missed other relevant articles on community-acquired Acinetobacter infections published in local journals. This is especially important for this review because of the increased incidence of community-acquired Acinetobacter infections in countries with tropical or subtropical climate.

Another limitation of our literature search for relevant studies on community-acquired Acinetobacter infections is related to the changes of the taxonomic classification of Acinetobacter. Specifically, the classification of oxidase-negative, non-nitrate-reducing, catalase-positive, Gram (−) bacteria in the genus Acinetobacter was introduced in 1968 [16, 17]. Prior to that, confusion over taxonomy existed among investigators, microbiologists, and clinicians who used classifications of Acinetobacter spp., in various genera, such as Mima and Herellea. Several different names were used in the literature before 1970 for bacteria of this genus, such as Herellea vaginicola, Mima polymorpha, Bacterium anitratum, Achromobacter anitratus, etc. The limited number of relevant papers in the older literature, as well as the uncertainty on whether each of those studies truly regarded Acinetobacter spp., led us to the decision not to expand our search to include articles published prior to 1973. Finally, it is important to note that it is likely that a publication bias exists in these reports.

Conclusion

Clinicians and microbiologists should be aware regarding the existence and severity of community-acquired Acinetobacter infections that mainly affect patients with some form of comorbidity or are associated with heavy smoking and excess alcohol consumption. It is noteworthy that, although community-acquired Acinetobacter infections were caused by isolates that did not have a very resistant pattern, the mortality associated with such infections was considerable; in fact, it did not appear to be substantially different from that reported in some studies of hospital-acquired Acinetobacter infections. A major limitation in interpreting the reviewed evidence regarding community-acquired Acinetobacter infections is that species identification has been inadequate, especially in early studies of Acinetobacter infections.