This prospective study was performed from November 2012 to March 2013 in the toxicological ICU (454 patients) and general ICU) (272 patients) units at Loghman Hakim Hospital. The research ethics committee of Shahid Beheshti University, Tehran, Iran, approved the study protocol.

The mean age of patients was 50.47 ± 21.32 years, and 55% of the patients were male. Most (87.18%) were under mechanical ventilation. Sputum was detected in 75% of the cases. All patients with clinical signs including temperature > 38℃, leukocyte count > 10000/µl, development of a purulent respiratory secretion, rales or dullness on chest examination, and A. baumannii outbreaks were defined as clinically suspicious for this microorganism. Patients who developed infections in the wards before admittance to the ICU and patients who received antibiotics < 24 h prior to admission were excluded. Patients who developed infection within 48 h of admission to the ICU were also excluded because any infection they developed most likely was acquired outside the ICU. Inclusion criteria for patients under mechanical ventilation included age > 18 years and at least 48 h on mechanical ventilation with a clinical suspicion of pneumonia. Patients had no clinical evidence of nosocomial infections upon admission to the ICU. A total of 60 patients were determined to be clinically suspicious for Acinetobacter infections. Patient samples included tracheal aspirates for investigation of the lower respiratory tract, urine and blood samples, and wound catheter and nasal swab cultures. Repeatedly isolated species from the same patient were excluded.

All isolates were identified using standard bacteriologic and biochemical methods, which included Gram staining, oxidase tests, oxidation/fermentation (O/F) tests, catalase tests, motility, citrate tests, and growth at 37 and 44℃.

Soil sampling was performed by passing cotton-tipped transfer swabs over the soil within 30 m of the hospital using a standard protocol. Copan Venturi swabs (Copan Diagnostics) were passed over sample surfaces, labeled, and shipped at room temperature to laboratories in two centers including Loghman Hakim and Mofid hospitals. A total of 16 soil samples were evaluated.

Cotton-tipped swabs moistened with BBL (We used either a single culture swab [BBL culturette with liquid Amies; BD Diagnostics] to sample the skin in both the axilla and groin areas, or 2 swabs to sample the axilla and groin areas separately [Scott P 2007]). Trypticase soy broth (BD Diagnostics, Sparks, MD, USA) were placed in the soil samples and streaked across BBL trypticase with 5% sheep blood and BBL MacConkey agar, then incubated at 35℃ for 72 h. Organisms were identified as being a part of the A. baumannii-calcoaceticus complex (ABC) using standard automated biochemical testing methods as described previously [11].

Hospital environment samples were collected from the window of the ICU, equipment such as the mechanical ventilator machine, instrument storage area (The sterilized tools literally on the walls of the ICU), heaters, air conditioners, patient beds, wall corners, refrigerators, and the cardex (nursing treatment cards). Samples from the lips and hands of health care workers and patients were also examined.

Clinical isolates were collected for phenotypic and genotypic characterization.

The standard Mast Co (UK) disc diffusion method was used for antimicrobial susceptibility testing based on the Clinical and Laboratory Standards Institute (CLSI) guidelines for clinical samples. The antibiotics evaluated were as follows: meropenem (10 µg), imipenem (10 µg), gentamicin (10 µg), ceftazidime (10 µg), tetracycline (30 µg), ceftriaxone (30 µg), cefepime (30 µg), cefotaxime (30 µg), ciprofloxacin (5 µg), cotrimoxazole (25 µg), piperacillin (100 µg), piperacillin-tazobactam (100/10 µg), and polymyxin B (300 µg). Pseudomonas aeruginosa ATCC 27853 was used as the control strain [7].

Extraction of genomic DNA from A. baumannii isolates was performed using the iNtRON G-spin™ Total DNA Extraction kit. Determination of integron classes 1 and 2 was performed via multiplex PCR using the primers BioNeer (Table 1) [8,9,10]. PCR of this gene was performed by Int2-R, Int2-F, Int1-R, and Int1-F with the size ranging from 160-288 bp (Table 1). A PCR assay used to define the complete gene makeup of the class 1 integron was performed in 20 µl total volume of (according to the the iNtRON G-spin™ kit)the same concentration mixture previously mentioned.

Data are reported as means ± SD, frequency, and relative frequency. The Chi-square test or Fisher's exact test was used to compare proportions. P < 0.05 was considered statistically significant. The statistical analysis was performed with SPSS (version 16, Chicago, IL, USA).