This study was conducted at the Chi Mei Medical Center, a 900-bed hospital located in southern Taiwan. Patients with positive blood cultures for Aeromonas species between January, 2009 and September, 2013, were identified from the hospital’s computerized database. The medical records of all patients with bacteremia due to Aeromonas species were retrospectively reviewed and the following information was collected: age, gender, underlying conditions (history of immunosuppressant drug use, diabetes mellitus, liver cirrhosis, end-stage renal disease, and active cancer), laboratory data, microbiological findings, antimicrobial susceptibility test results, and patient outcome. The data was collected on a routine basis and the analysis was carried out retrospectively. The records and information of patient were anonymized and de-identified prior to analysis. Therefore, no informed consent was required and it was specifically waived by Institution Review Board. Ethics approval was obtained from Institution Review Board of Chi Mei Medical Center.

Blood were inoculated into BACTEC culture bottles using the BACTEC 9240 system (Becton Dickinson, Cockeysville, MD, USA). Sputum culture specimens were processed on blood agar, chocolate agar, and eosin methylene blue agar media (Becton Dickinson). Gram-negative isolates that tested positive for cytochrome oxidase, glucose fermentation, citrate usage, indole production, and ornithine decarboxylase were classified as Aeromonas species. All strains were identified to the species level by conventional methods and were further verified by the API-20E System (bioMérieux Vitek Inc, Hazelwood, MO, USA), the ID 32 GN System (bioMérieux Vitek Inc), or the Vitek 2 ID-GNB identification card (bioMérieux Inc, Durham, NC, USA). Susceptibilities of these isolates to a battery of antimicrobial agents were determined using the disk diffusion method as described by the Clinical and Laboratory Standards Institute [21].

The diagnosis of infection focus of bacteremia was made based on clinical, bacteriological, and radiological investigations. Peritonitis was diagnosed in patients with clinical peritoneum inflammation plus positive culture of ascites. Biliary tract infections were diagnosed in patients with clinical hepatobiliary tract inflammation plus positive bile cultures. The bile specimens were collected by percutaneous transhepatic cholangiodrainage, percutaneous gallbladder drainage, or operation. Catheter-related bloodstream infection was defined as a positive semi-quantitative tip culture (≧15 colony-forming units), bacteremia, and/or high clinical suspicion; pneumonia was defined as a positive culture for Aeromonas spp. in purulent sputum samples and the presence of newly developed lung infiltrates; Genitourinary tract infection (UTI) was defined as positive urine culture with growth of ≥ 10⁵ CFU/ml and pyuria; and skin and soft tissue infection (SSTI) was defined as clinical soft tissue inflammation plus positive soft tissue or pus culture and bacteremia. If no primary focus could be identified, the bacteremia was classified as primary. Shock was diagnosed in patients with a systolic blood pressure < 90 mm Hg or in patients who required inotropic agents to maintain blood pressure. In-hospital mortality was defined as death due to any cause during hospitalization. Bacteremia was classified as healthcare-associated infections (HCAI) in patients who acquired the disease during the course of treatment for other conditions within a healthcare setting [22]. Otherwise, bacteremia was classified as community acquired. Polymicrobial infections were classified if the patients whom had other non-Aeromonas pathogens grew from blood samples. Inappropriate usage of antibiotics was defined as that the clinical isolates were in vitro resistant to used antimicrobial agents.

Continuous variables are expressed as means ± standard deviations. Comparisons between each variable/category were using the chi-square test or one-way analysis of variance, as appropriate. A multivariate stepwise logistic regression model was used to identify risk factors for mortality. All statistical analyses were conducted using the statistical package SPSS for Windows (Version 19.0, SPSS, Chicago, Il, USA), and a P value <.05 was considered to show statistical significance.

The clinical characteristics of the 91 patients with bacteremia caused by Aeromonas species are summarized in Table 1. The patients ranged in age from 10 to 98 years (mean, 64.3. years) and half of them were classified as elderly patient with age ≥ 65 years. In addition to 16 (17.6%) primary bacteremia, the most common source of secondary infection is peritonitis (n = 27, 29.7%), followed by biliary tract infection (n = 18, 19.8%), and SSTI (n = 12, 13.2%), pneumonia (n = 9, 9.9%), catheter-related bloodstream infection (n =  5, 5.5%), and genitourinary tract infection (n = 4, 4.4%). A. hydrophila (n = 35, 38.5%) was the most common pathogen, followed by A. veronii biovar sobria (n = 31, 34.1%), A. caviae (n = 14, 15.4%), and A. veronii biovar veronii (n = 9, 9.9%). Cancer (n =  40, 44.0%) was the most common underlying disease with lung cancer (n = 16) being the most common types of cancer. Liver cirrhosis (n =  36, 39.6%) was the second most common underlying disease. Fever, and shock was the initial presentation in 69 (75.8%) and 25 (27.5%) patients, respectively. Of the patients who had polymicrobial bacteremia, E. coli was the most common co-pathogen (n = 13), followed by Klebsiella pneumoniae (n = 8), enterococci (n = 3), Acinetobacter spp. (n = 2), Morganella morganii (n = 2), Citrobacter freundii (n = 1), Pseudomonas aeruginosa (n = 1), and Enterobacter species (n = 1).

In this study, 43 (47.3%) patients were classified as HCAI causes by Aeromonas species. Table 2 summarized the comparison between patients with healthcare-associated bacteremia and community-acquired bacteremia caused by Aeromonas species. We found patients with HCAI were more likely to have cancer, and to receive immunosuppressant than patients with CAI. Furthermore, the outcome including ICU admission, acute respiratory failure and in-hospital mortality were significant higher among patients with HCAI than CAI (all P < 0.05).

Clinical outcomes, including rate of ICU admission, acute respiratory failure, and mortality were 33.3%, 28.6%, and 23.1%, respectively. In-hospital mortality was significantly associated with nosocomial infection, cancer, receiving immunosuppressant drugs, and shock as the initial presentation (Table 3). In contrast, overall mortality was not associated with age, gender, or other underlying conditions, such as diabetes mellitus, end-stage renal disease, polymicrobial infection or initial inappropriate antibiotic therapy. Multivariate analysis showed that the in-hospital mortality was only significantly associated with underlying cancer (95% confidence interval (CI), 0.13–0.43, P <.001), and initial shock (95% confidence interval (CI), 0.28–0.63, P <.001).

The results of in vitro susceptibility testing of clinical Aeromonas isolates against various antimicrobial agents are shown in Table 4. More than 98% of clinical isolates were susceptible to amikacin and imipenem, and more than 90% of clinical isolates were susceptible to ceftriaxone, ceftazidime, cefepime, ciprofloxacin, and gentamicin. In contrast, more than 90% of the isolates were not susceptible to ampicillin or cefazolin.