Patterns of antimicrobial resistance in a surgical intensive care unit of a university hospital in Turkey

Background Several studies have reported higher rates of antimicrobial resistance among isolates from intensive care units than among isolates from general patient-care areas. The aims of this study were to review the pathogens associated with nosocomial infections in a surgical intensive care unit of a university hospital in Turkey and to summarize rates of antimicrobial resistance in the most common pathogens. The survey was conducted over a period of twelve months in a tertiary-care teaching hospital located in the south-eastern part of Turkey, Gaziantep. A total of 871 clinical specimens from 615 adult patients were collected. From 871 clinical specimens 771 bacterial and fungal isolates were identified. Results Most commonly isolated microorganisms were: Pseudomonas aeruginosa (20.3%), Candida species (15%) and Staphylococcus aureus (12.9%). Among the Gram-negative microorganisms P. aeruginosa were mostly resistant to third-generation cephalosporins (71.3–98.1%), while Acinetobacter baumannii were resistant in all cases to piperacillin, ceftazidime and ceftriaxone. Isolates of S. aureus were mostly resistant to penicillin, ampicillin, and methicillin (82–95%), whereas coagulase-negative staphylococci were 98.6% resistant to methicillin and in all cases resistant to ampicillin and tetracycline. Conclusion In order to reduce the emergence and spread of antimicrobial-resistant pathogens in ICUs, monitoring and optimization of antimicrobial use in hospitals are strictly recommended. Therefore local resistance surveillance programs are of most value in developing appropriate therapeutic guidelines for specific infections and patient types.


Background
Antimicrobial resistance among intensive care unit (ICU) pathogens is generally increasing, but variations do exist among different countries, probably due to individual antimicrobial use patterns. When new medical practices and alternative antimicrobials are introduced, changes in the dominant microbial etiologies may emerge prompting novel empiric selections. Appropriate therapy of ICU infections directed by local resistance data can have signif-icant consequences for both patient and the healthcare system.
Data from National Nosocomial Infections Surveillance (NNIS) shows that from 1988 to 1995 the number of intensive care unit (ICU) beds at the hospitals has increased 17%, whereas total hospital bed capacity decreased slightly [1]. Patients receiving care in ICUs are at high risk for nosocomial infections. The emergence of antimicrobial-resistant pathogens in ICUs has made treating these infections very difficult and, in some cases, impossible. Intensive care unit patients are particularly susceptible to nosocomial infections due to underlying illnesses, suppressed immune systems and frequent use of invasive devices.
This study reviews the pathogens associated with nosocomial infections among ICU patients, summarizes rates of antimicrobial resistance in the most common pathogens and provides an overview of strategies to prevent the proliferation of antimicrobial-resistant microorganisms.

Results
During the study period a total of 871 clinical specimens from 615 adult patients were collected. None of the patients were receiving immunosuppressive therapy. In 771 (88.5%) of cases the isolate recovered by culture was confirmed to be the etiologic agent of a nosocomial infection. The distribution of infections was examined by the major site of infection. Four major infection sites represented 91.8% of all reported infections; lower respiratory tract infections were most frequent (31.5%), followed by urinary tract infections (27.8%), bloodstream infections (23.1%) and surgical site infections (9.3%).
Each of the pathogens listed on Table 1 has demonstrated antimicrobial resistance to at least one, if not several, of the antimicrobial agents commonly used to treat infections caused by these pathogens. In general, Gram-positive microorganisms such as S. aureus and coagulasenegative staphylococci were commonly associated with bloodstream or surgical site infections. Examination of the rates of antimicrobial resistance among these pathogens showed that rates of methicillin-resistant S. aureus (MRSA) and methicillin-resistant coagulase-negative staphylococci were 82% and 98.6%, respectively ( Table 2). Antimicrobial resistance profiles of MRSA and MSSA (methicillin-sensitive S. aureus) isolated from infections in ICU were shown in Table 3.

Discussion
A number of factors contribute to the emergence of antimicrobial resistance in ICUs including the severity of patient illness, predisposition to nosocomial infections, cross-transmission of pathogens characteristic of critical care areas within the hospital, compromised membrane and skin barriers following the use of invasive devices, extended length of hospital stay, and the widespread use of prophylactic and therapeutic anti-infective agents [2,3].
The types of organisms that have emerged as most problematic for patients within the ICU include the members of the family Enterobacteriaceae, non-fermenters (P. aeruginosa and Acinetobacter spp.) oxacillin-resistant S. aureus and vancomycin-resistant enterococci. All of these resistance profiles have been documented in other regional and global ICU surveillance studies including Project Intensive Care Antimicrobial Resistance Epidemiology (ICARE, 1994(ICARE, -2000, the Meropenem Yearly Susceptibility Test Information Collection Program (MYSTIC, 1997(MYSTIC, -2000, the ICU Surveillance Study (ISS, 1990(ISS, -1993(ISS, , 1994(ISS, -2000 and the SENTRY Program (Europe, 1997-1998) [3][4][5][6][7][8]. Patients in the ICU are more likely than others to be colonized or infected with an antimicrobial-resistant pathogen, therefore the rates of resistance are significantly higher in patients cared for in the ICU than in non-ICU patients [9][10][11][12][13][14].  Examination of Gram-positive microorganisms, such as S. aureus and coagulase-negative staphylococci shows that rates of methicillin-resistant isolates in ICU have increased steadily over the past decade [13][14][15]. In our study 82% of S. aureus strains and 98.6% of coagulasenegative staphylococci were resistant to methicillin.
Gram-negative bacilli are frequently associated with nosocomial infections in ICUs. Data from a multicenter Intensive Care Unit Surveillance Study (ISS) in the United States demonstrated that resistance to antipseudomonal agents among ICU isolates of P. aeruginosa, especially fluoroquinolones, was increasing [16]. The isolates of P. aeruginosa in this study were resistant 71.4 to 98% to third generation cephalosporins, and 69.4 to 93% to antipseudomonal penicillins. The relatively high susceptibility of Gram-negative bacilli to fluoroquinolones in this study can be attributed to antimicrobial prescribing protocol used in our ICU, which is preferably directed to broadspectrum antibiotics.
In recent years Acinetobacter spp. have emerged as important pathogens of ICUs, most of them being resistant to ampicillin, carbenicillin, cefotaxime, chloramphenicol, and gentamicin [17,18]. In our study, all A. baumannii isolates were resistant to third generation cephalosporins, and most of them were resistant to aztreonam, ticarcillin, and gentamicin with 94.6%, 93.2%, and 85.1%, respectively.
Enterobacter species are resistant to first-generation cephalosporins and develop antibiotic resistance readily to second-and third-generation cephalosporins owing to an inducible chromosomally encoded cephalosporinase. Emergence of this form of resistance is seen frequently, when infection due to these organisms, particularly Enterobacter cloacae, are treated with broad-spectrum cephalosporins [19]. Pfaller et al. [20] reported that out of 230 Enterobacter isolates 35-50% were resistant to ceftazidime and piperacillin. In this study the resistance rates of Enterobacter strains to third generation cephalosporins were between 59.4-75%, whereas higher rates were recorded to aztreonam and gentamicin (both 81.2%).
Another type of commonly seen antimicrobial-resistant pathogen encountered among ICU patients is Klebsiella pneumoniae, which is producing extended-spectrum betalactamases (ESBLs). Our isolates of K. pneumoniae showed high resistance to broad-spectrum cephalosporins and aztreonam (72-88% and 88%, respectively).
Many studies have demonstrated that the development of antimicrobial resistance at hospital level is strongly correlated with the use of the relevant antimicrobial [21]. Data from project ICARE showed that use was significantly higher among ICU patients than non-ICU patients for third-generation cephalosporins combined, vancomycin, anti-pseudomonal penicillin, intravenous fluoroquinolones and imipenem [22]. This study supports that for each of the antimicrobial agent used at higher rates in ICU areas, there was a correspondingly higher rate of the respective resistant pathogen among ICU patients.

Conclusion
In order to prevent the emergence and spread of antimicrobial resistant pathogens in ICU, the pattern of antimicrobial use has to be determined. A multidisciplinary approach is required to succeed in combating the problem. Hospitals should monitor antimicrobial use to determine whether specific ICUs or the entire hospital is overusing antimicrobials. These data could be used, in conjunction with other related studies, to properly interpret significant resistance patterns and choose the most appropriate antimicrobial regimens for empirical therapy. Lower respiratory tract specimens included bronchoalveolar lavage, transtracheal aspiration, and pleural fluid. Urinary tract specimens included urine and aspiration from the urinary catheter. Other types of specimens obtained from the patients were; swabs from surgical wounds, cerebrospinal fluid, pericardial fluid, peritonal fluid, synovial fluid, biopsy material, oral swab, rectal and anal swabs. All specimens were collected at the bed site, transferred to the laboratory immediately and were inoculated on proper culture media within two hours. Clinical specimens were inoculated onto 5% sheep blood agar, eosinemethylene-blue agar, Sabouraud's dextrose agar, and except the urine specimens, onto chocolate agar (Oxoid, Basingstoke, United Kingdom). Plates were incubated at 37°C in 5% CO 2 for 18-24 hours. Methods used for confirmation of identification included examination of colonial morphology and haemolytic characteristics on appropriate agar media, Gram stain, rapid tests (catalase, oxidase, coagulase, bile solubility, spot indole, latex agglutination) and use of an automated identification system, VITEK (bioMérieux, Marcy-1'Etoile, France), which is a microbroth dilution method with two wells across predetermined breakpoints.
Penicillin G, ampicillin, amoxicillin/clavulanate, methicillin, vancomycin, clindamycin, erythromycin, trimethoprim/sulfamethoxazol, ciprofloxacin, and tetracycline were used for antimicrobnial susceptibility testing. Susceptibility tests were performed with a commercial broth microdilution method (VITEK; bioMérieux, Marcy-1'Etoile, France) according to the manufacturers' guideline recommendations and interpretative criteria. Colonies from 18-24 hours old culture medium were used to inoculate the microdilution plates. Quality control for testing by VITEK was performed before each new lot of cards was used. Antimicrobial susceptibilities of the control microorganisms (S. aureus ATCC 29213, P. aeruginosa ATCC 27853 and Enterococcus faecalis ATCC 29212) were performed according to the instructions of CLSI (previously NCCLS) [24].
All data were stored and analysed using Microsoft Access and Excel. Statistical analysis was performed using the chisquared test for independence and Fisher's exact test where appropriate. All statisticss were computed using Scientific Package for Social Sciences (SPSS) software (version 11.0; SPSS Inc., Chicago, IL).

Authors' contributions
AB carried out the collection of clinical samples from patients, cultured and identified pathogenic microorganisms from those specimens and perfomed antimicrobial susceptibility tests of the study strains. She also designed the study and performed acquisition, analysis and interpretation of data. IB conceived of the study, and participated in its design and coordination and helped to draft the manuscript. Both authors read and approved the final version of the manuscript.