1 Introduction
Burn infections are a serious hindrance to patient recovery. Infections have been estimated to account for 75% of burn patient deaths.[1,2] [3] [4]
By reviewing the variable history of the burn wound bacterial ecology, we have observed changes with time[5] [6] drug resistance varies in response to local or systemic medications.[7]
In this study, a retrospective analysis was conducted concerning the pathogen distribution and antimicrobial resistance of a total of 1891 isolates from 1449 patients with nosocomial infections in 4 burn wards in Fujian province (located in the southeast of China) from January 2013 to December 2015. This study could serve as a reference for the prevention or treatment of burn infections and the rational use of antimicrobials.
2 Materials and methods
2.1 Patients
Patients treated in the burn wards of Fujian Medical University Union Hospital, the 180th Hospital of PLA, the 92nd Hospital of PLA, and the First Hospital of Longyan City between January 2013 and December 2015 were included in this retrospective analysis. Patients with incomplete data will be excluded. Fujian Medical University Union Hospital is a Class A tertiary provincial public hospital. Its burn department, established in 1976, is the burn center for Fujian province. The center has 3 subdivisions, including burn treatment, plastic surgery, and rehabilitation. The center consists of an independent outpatient department for wound treatment, independent operating rooms, a burn intensive care unit (BICU), and 82 ward beds, which include 8 sickbeds in the BICU and 6 for emergencies. Between 2013 and 2015, a total of 2552 new burn patients were admitted to this center. The 180th Hospital of Chinese People's Liberation Army (PLA), a Class A tertiary military hospital with 79 beds in its burn department including 7 sickbeds in the BICU, treated 4591 burn patients between 2013 and 2015. The 92nd Hospital of PLA is also a Class A tertiary military hospital with a total of 44 beds in its burn department, which treated 956 burn patients during the same period. The First Hospital of Longyan City is a Class A tertiary municipal public hospital and the only center of its type in the western Fujian province. Its burn department is equipped with 20 sickbeds and treated 727 new burn patients in 2013 to 2015. The condition and treatment of burn patients at these 4 burn wards are satisfactorily representative of the entire Fujian province.
The inclusion criteria were as follows: burn data from the first admission were extracted using the International Classification of Diseases, Tenth Revision (ICD-10) codes in the X00 to X19 range; patients admitted to these 4 hospitals for >24 hours or who died after arrival at the hospital were studied. The following categories of patients were excluded: readmissions for scar contracture; outpatients; inpatients not diagnosed with a burn as the primary cause of admission; and incomplete clinical data.
A total of 1449 patients were enrolled in our study (843 cases, 160 cases, 160 cases, and 286 cases, respectively, from the aforementioned 4 burn departments). The patients’ ages ranged from 8 days to 94 years (median age, 32 years; interquartile range, 29 months to 48 years). The subjects included 996 men and 453 women. The total burn surface area (TBSA) ranged from 1% to 99% (median TBSA, 11%; interquartile range, 4%–24%). The burn depths were between II and III with 446 cases of flame burns, 732 cases of scalding, 108 cases of contact burns, 105 cases of electrical burns, 34 cases of chemical burns, and 24 cases of other burn etiologies. Mild burns accounted for 313 cases, whereas 709 cases were moderate burns, 216 cases were severe burns, and 211 cases were extremely severe burns.
In total, 1891 strains of pathogenic bacteria were cultured from 3835 clinical specimens obtained from and distributed as follows: wound secretions, 1992; blood samples, 979; respiratory secretions, 595; central venous catheter specimens, 133; and other sources (such as urine, stool, and tissue fluid), 136.
The ethics committee of Fujian Medical University and each collaborating institution reviewed and approved the study protocol. All collaborating institution provided written consent for their information to be collected and used for research, and they had the right to withdraw from the study at any time without prejudice.
2.2 Sample collection
Wound secretions were collected at patient admission and during hospitalization at least once; the required volume for each sample was >1 mL. Sputum secretions were collected from patients receiving preventive tracheotomy or ventilator support; the required sample volume for each was >2 mL. Central venous catheter samples were obtained at catheter replacement; the minimal required sample volume was 5 cm. Blood samples were obtained when the patient's temperature rose >38.5 °C or was <36 °C (sourced from 2 peripheral blood collection sites or 1 peripheral blood and 1 intraductal blood site using 2 sets of 4 bottles with aerobic and anaerobic cultures obtained for each sample; the required volume for each sample was 5–10 mL for adults and 2–5 mL for children). Urine samples were obtained when patients had irritative urinary tract symptoms (using 1 set of 2 bottles with aerobic and anaerobic cultures obtained for each sample; the required volume for each sample was 5–10 mL). Stool samples were obtained when patients had diarrhea (using 1 set of 2 bottles with aerobic and anaerobic cultures obtained for each sample; the required volume for each sample was at least 1 mL). For patients with suspected sepsis, all aforementioned sample cultures were obtained for 3 consecutive days. Patients were treated based on the revised guideline, Diagnostic Criteria and Treatment Guideline for Infection of Burns and Guideline for Diagnosis, Prevention and Treatment of Invasive Fungal Infection after Burn Injury.[8]
2.3 Species identification and antibiotic sensitivity
Species identification and antimicrobial sensitivities were assessed by the laboratory staff of the 4 hospitals using the Kirby-Bauer (K-B) disk diffusion method employing drug-containing test disks, culture medium and quality control strains (Staphylococcus aureus ATCC 25923, Pseudomonas aeruginosa ATCC 27853, Escherichia coli ATCC 25922, and Candida albicans ATCC 10231) (Oxoid Corporation, England). The American Clinical and Laboratory Standards Institute (CLSI) standard was applied to evaluate outcomes. All specimens were inoculated in the appropriate culture medium and incubated at 35 °C in accordance with their respective requirements for 18 to 20 hours. The VITEK AMs60 compact automatic microbial analyzer (BioMerieux, France) was employed to identify the strains. A 30-g cefoxitin disk was used to detect methicillin-resistant staphylococci. Methicillin resistance was detected in coagulase-positive staphylococci when the inhibition zone diameter was ≤21 mm. Methicillin resistance was excluded in coagulase-negative staphylococci when the inhibition zone diameter was ≤24 mm.
2.4 Statistical analysis
WHONET software, version 5.5, and SPSS software, version 19.0 (SPSS, Inc., Chicago, IL), were used for statistical analysis. Categorical data are presented as frequencies and percentages.
3 Results
3.1 Annual distribution of pathogens in clinical samples
A total of 1891 strains of pathogens were detected from 3835 clinical specimens collected from 1449 burn patients over a 3-year period. The demographics of the 1449 patients were shown in Table 1 .
Table 1: The demographics of 1449 patients.
The total detection rate was 49.3% (1891/3835). During these 3 years, the wound specimen strain detection rate showed a downward trend from 65.2% to 57.3%. The blood specimen strain detection rate was low with a range between 7.4% and 31.4% (Tables 2 and 3 ).
Table 2: Annual detection rates of clinical samples from wounds, blood, sputum, central venous catheters, and other sources.
Table 3: Annual detection rates of clinical samples from total samples.
3.2 Annual pathogen distributions and percentages
The most common pathogens isolated were gram-negative bacteria with 1089 strains (57.6%) followed by gram-positive bacteria with 689 strains (36.4%) and fungi with 113 strains (6.0%). The gram-negative bacteria detection rate was higher than that of gram-positive bacteria in every year. The 8 most common bacteria isolated were S aureus (19.0%), Acinetobacter baumannii (17.6%), P aeruginosa (16.7%), Klebsiella pneumoniae (7.4%), Enterococcus faecalis (4.5%), Staphylococcus epidermidis (3.6%), and E coli (3.2%). Over the study period, A baumannii and P aeruginosa exhibited a decreasing trend, while S aureus showed a rising trend (Table 4 ). Pathogen distributions and percentages among 1% to 30%, 31% to 60%, 61% to 100% TBSA patients were shown in Table 5 . Clinical samples such as wound secretions, blood, sputum, and central venous catheters all exhibited a predominance of gram-negative bacteria. Sputum samples had the highest detection rate for fungi (Table 6 ).
Table 4: Annual pathogen distributions and percentages.
Table 5: Pathogen distributions and percentages among 1% to 30%, 31% to 60%, 61% to 100% TBSA patients.
Table 6: Pathogen distributions among different clinical samples.
3.3 The detection rate of methicillin-resistant staphylococci
Methicillin-resistant Staphylococcus aureus (MRSA) accounted for 74.1% (265/359) of S aureus isolates. The overall detection rate of MRSA was 14.0% with an increasing trend ranging between 12.2% and 16.7%. S epidermidis accounted for 40.6% (69/170) of coagulase-negative staphylococcal isolates, and 70.0% (48/69) of these were methicillin-resistant Staphylococcus epidermidis (MRSE). The overall detection rate of MRSE was 2.5% (48/1891), which changed little over the years (Table 7 ).
Table 7: Annual detection rates of MRSA, MSSA, MRSE, and MSSE.
3.4 Resistance rates of gram-positive bacteria to antimicrobials
The resistance rates of the isolated bacteria to commonly used antimicrobials were investigated (Table 8 ). The resistance rates of MRSA to penicillin and ampicillin was 100%, whereas the resistance rates to erythromycin, clindamycin, gentamicin, tetracycline, and ciprofloxacin were >75%. In general, the high-resistance antibiotics were penicillin, ampicillin, and erythromycin, and the low-resistance antibiotics were teicoplanin, tigecycline, vancomycin, and linezolid. E faecalis was 100% resistant to quinupristin/dalfopristin synercid, whereas Enterococcus faecium was 100% sensitive to quinupristin/dalfopristin synercid. More than 50% of E faecalis was resistant to tetracycline, erythromycin, and rifampicin. Both enterococcal species were completely sensitive to vancomycin, tigecycline, and linezolid. Additionally, the frequency of E faecalis isolates was higher than that of E faecium . However, the drug resistance of E faecalis to penicillin, ciprofloxacin, levofloxacin, and ampicillin was lower than that of E faecium .
Table 8: Resistance rates of gram-positive bacteria to antimicrobials.
3.5 Resistance rates of gram-negative bacteria to antimicrobials
The detection rates of extended-spectrum β-lactamases (ESBL) among K pneumoniae and E coli isolates were 44.6% (62/139) and 67.2% (41/61), respectively. For A baumannii , the most commonly isolated gram-negative bacterial strain, the resistance rate was 100% for ampicillin, amoxicillin, amoxicillin/clavulanate, and aztreonam, and >80% for third-generation cephalosporins such as cefotaxime, ceftazidime, ceftriaxone, and cefepime. The A baumannii resistance rates for carbapenems such as imipenem and meropenem were 58.5% and 83.3%, respectively.
In addition to resistance rates of K pneumoniae to ampicillin, amoxicillin, and cefazolin of 80%, these rates were 21% for meropenem and 15.4% for imipenem. More than 80% of E coli was resistant to ampicillin, piperacillin, cefazolin, amoxicillin, tetracycline, and sulfamethoxazole trimethoprim (Table 9 ).
Table 9: Resistance rates of gram-negative bacteria to antimicrobials.
3.6 Resistance rates of fungi to antimicrobial agents
The resistance rates of C albicans to amphotericin B, 5-fluorocytosine, and nystatin were 0%, and the resistance rate of Candida tropicalis to itraconazole was 50% (Table 10 ).
Table 10: Resistance rates of fungi to antimicrobials.
4 Discussion
Nosocomial infections (NIs) are more likely to occur among burn patients due to the immunocompromising effects of burns, the nature of burns themselves, the intensive diagnostic and therapeutic procedures, and prolonged hospital stays.[9] [10]
4.1 Pathogen detection rates and distribution characteristics
In total, 1891 strains of pathogenic bacteria or fungi were cultured from wound secretions, blood, central venous catheter, sputum, urine, stool, and tissue fluid. The overall detection rate was 49.3%. Specifically, the detection rates were 56.9% to 65.2% for wound secretions and 10.3% to 20.8% for blood, and these values are lower than the results obtained in the study conducted by Huang et al[11]
Among all the microbes detected, 36.4% (689/1891) were gram-positive bacteria, 57.6% (1089/1891) were gram-negative bacteria, and 6.0% (113/1891) were fungi. These results were similar to a study from Turkey,[12]
Furthermore, our results showed that gram-positive bacteria increased and gram-negative bacteria decreased over these 3 years. This likely because of the clinical application of broad-spectrum antibiotics and the increased use of indwelling devices and other invasive procedures.
In our study, clinical samples such as wound secretions, blood, sputum, and central venous catheters all exhibited a predominance of gram-negative bacteria. This trend differed from an earlier report by Cen et al,[13] [13]
4.2 Analysis of bacterial resistance rates to antimicrobials
S aureus was the most prevalent bacterium observed. This result was similar to a previous study of 3615 microbial isolates from 114 patients with severe burns at the burn center of Shanghai Hospital (Shanghai, P.R. China) between 1998 and 2009, which found that S aureus accounted for 38.2% of all cases. In our study, MRSA represented 74.1% of S aureus isolates, approximating the result (73.0%) of another study in China by Wei et al[14] S aureus cases in our study were methicillin-resistant, a preponderance much higher than reported elsewhere, for example, by Dokter et al[15] [16] [5] [12] S aureus isolates were MRSA in China. The resistance rates of MRSA to penicillin and ampicillin were 100% and 78.1% to 84% for other antibiotics such as erythromycin, tetracycline, and clindamycin. The resistance rates of MSSA and MSSE were relatively low to the aforementioned antibiotics except for penicillin and ampicillin. Vancomycin, teicoplanin, and tigecycline remained the most effective antibiotics against MRSA and MRSE with no strains resistant to these antibiotics. The policy named “search-and-destroy” has ensured a low prevalence of MRSA in health facilities and the population of the Netherlands,[17]
The coagulase-negative staphylococci detected included S epidermidis and Staphylococcus saprophyticus , both of which typically constitute part of the normal flora of the skin and mucous membranes.[17] [18–20] S epidermis isolates, representing a relatively high detection rate. This finding should alert clinicians to control skin and mucosa disinfection strictly to avoid nosocomial infection.
The Enterococcus genus is among the normal flora of humans and animals, and ectopic microflora will lead to infection. Recently, Kozuszko et al[21] [22] [23]
P aeruginosa was reported as a major pathogenic cause of infection after burn injuries in the United States.[2] P aeruginosa as the most common isolate with high prevalence in regions such as the southwest of China[10] [24] [25] [26] [27] [28] P aeruginosa in burn wards was higher than in other wards. However, this finding was inconsistent with our observations. In our study, P aeruginosa was the second most prevalent gram-negative bacterium, accounting for 16.7% of bacterial isolates, which was slightly higher than in the studies by Coetzee et al[29] [12] P aeruginosa was 100% resistant to ampicillin, ampicillin/sulbactam, amoxicillin, amoxicillin/clavulanic acid, cefuroxime, cefoxitin and cefotaxime, 97.5% resistant to sulfamethoxazole trimethoprim, and 8.2% to 12.9% resistant to cefepime, cefoperazone/sulbactam, piperacillin/sulbactam, and amikacin.
Recently, A baumannii has become one of the crucial causes of nosocomial infections.[30] A baumannii (CNSAb) infections is becoming a growing problem in hospitalized burn patients.[31] A baumannii was the most common bacterium detected in most studies, for example, by Bayram et al[12] [32] [33] [34] A baumannii was the prominent pathogen among gram-negative bacteria, which presented high resistance to many antibiotics. The resistance rate to imipenem, the most effective broad-spectrum agent against gram-negative bacilli, was 83.3%. Moreover, diffusion disk testing using cefoperazone/sulbactam-containing inhibitors also showed a relatively high resistance rate of 55.3%. The resistance rate of A baumannii to ampicillin, piperacillin, aztreonam, amoxicillin, and amoxicillin/clavulanic acid was 100% and was >80% to cephalosporins.
The emergence of ESBL-producing strains among Enterobacteriaceae (E coli , K pneumoniae ) has become a special concern.[35] K pneumoniae and E coli were 44.6% (62/139) and 67.2% (41/61), respectively.
Fungal infections have been found to occur after the widespread use of antibiotic therapies, which have the effect of killing the beneficial bacteria that normally suppress fungi. More recently, Candida species have emerged as an important cause of invasive infections among patients in intensive care units.[36] C albicans , accounting for 38.1% (43/113) of fungal isolates. The antibiotic resistance of C tropicalis was higher than that of C albicans . A useful reference for China might be “Three steps to prevent invasive fungal diseases,” as concluded by Pemán and Salavert.[37]
4.3 Limitations
This was a multicentre study to investigate the epidemiology of bacteria among hospitalized burned patients. However, because of limited resources, we only investigated 1 province in the southeast of China. Because it was conducted retrospectively, we were unable to include the following steps: to test the homology of the MRSA; to analyze the impact of prehospital wound treatment on bacterial detection; to analyze the effect of antimicrobial administration on bacterial detection; and to analyze the effect of long-term catheterization on these bacteriological profiles. Future efforts are required to perform multicenter, prospective studies to develop dynamic monitoring for bacterial resistance.
5 Conclusions
Gram-negative bacteria represented the majority of pathogens detected. S aureus , A baumannii , P aeruginosa , K pneumoniae , and E faecalis were the 5 most common bacteria detected in the 4 study burn units. The resistance of MRSA and A baumannii to antibiotics was relatively high in Fujian province. High-resistance antimicrobials included penicillin, ampicillin, amoxicillin, cefazolin, and cefotaxime. Low-resistance antimicrobials included teicoplanin, tigecycline, vancomycin, and linezolid. Dynamic bacterial monitoring should be established to restrain the development of bacterial resistance.
Acknowledgments
The authors thank Fujian Medical University Union Hospital, the 92nd Hospital of PLA, the 180th Hospital of PLA, and the First Hospital of Longyan City for their support.
Author contributions
Data curation: Lin Li, Jia-xi Dai.
Funding acquisition: Lin Li, Le Xu.
Investigation: Lin Li, Jia-xi Dai, Zhao-hong Chen, Xiao-dong Chen.
Methodology: Lin Li, Zhao-hong Chen.
Project administration: Le Xu.
Resources: Zhao-hong Chen.
Software: Xiao-dong Chen.
Supervision: Xiao-yi Li, Yu-qing Wen, Min Liu.
Validation: Xiao-yi Li, Xiao-dong Chen.
Writing–original draft: Jia-xi Dai.
Writing–review & editing: Jia-xi Dai.
References
[1]. Wang Y, Tang HT, Xia ZF, et al. Factors affecting survival in adult patients with massive burns. Burns 2010;36:57–64.
[2]. Hodle AE, Richter KP, Thompson RM. Infection control practices in U.S. burn units. J Burn Care Res 2006;27:142–51.
[3]. Magnotti LJ, Deitch EA. Burns, bacterial translocation, gut barrier function, and failure. J Burn Care Rehabil 2005;26:383–91.
[4]. Shelby J, Merrell SW. In vivo monitoring of postburnimmune response. J Trauma 1987;27:213–6.
[5]. Guggenheim M, Zbinden R, Handschin AE, et al. Changes in bacterial isolates from burn wounds and their antibiograms: a 20-year study (1986–2005). Burns 2009;35:553–60.
[6]. Li HM, Liang ZQ, Liu DE, et al. An investigation of bacterial ecology and analysis of bacterial resistance to antibiotics in a burn ward in Nanning district. Chin J Burns 2005;21:107–10.
[7]. Xu WS. Bacterial ecology on bum wound and antibacterial agent therapy. Chin J Burns 2008;24:334–6.
[8]. Peng YZ, Yuan ZQ. [Standardized definitions and diagnostic criteria for infection in burn patients]. Zhonghua Shao Shang Za Zhi 2007;23:404–5.
[9]. Peck MD, Weber J, McManus A, et al. Surveillance of burn wound infections: a proposal for definitions. J Burn Care Rehabil 1998;19:386–9.
[10]. Sun FJ, Zhang XB, Fang Y, et al. Spectrum and
drug resistance of pathogens from patients with burns. Burns 2012;38:1124–30.
[11]. Huang XZ, Hu ZZ, Lin PP, et al. The bacteria distribution and
drug resistance analysis of burn ward in five years. Chin J Burns 2012;28:227–9. [In Chinese].
[12]. Bayram Y, Parlak M, Aypak C, et al. Three-year review of bacteriological profile and antibiogram of burn wound isolates in Van, Turkey. Int J Med Sci 2013;10:19–23.
[13]. Cen H, Wu Z, Wang F, et al. Pathogen distribution and
drug resistance in a burn ward: a three-year retrospective analysis of a single center in China. Int J Clin Exp Med 2015;8:19188–99.
[14]. Wei LH, Zou FM, Liu G, et al. [
Drug resistance of methicillin resistant Staphylococcus strains in burn ward and relative analysis]. Zhonghua Shao Shang Za Zhi 2013;29:55–8.
[15]. Dokter J, Brusselaers N, Hendriks WD, et al. Bacteriological cultures on admission of the burn patient: to do or not to do, that's the question. Burns 2016;42:421–7.
[16]. Fransén J, Huss FRM, Nilsson LE, et al. Surveillance of antibiotic susceptibility in a Swedish Burn Center 1994–2012. Burns 2016;42:1295–303.
[17]. Becker K, Heilmann C, Peters G. Coagulase-negative staphylococci. Clin Microbiol Rev 2014;27:870–926.
[18]. Danzmann L, Gastmeier P, Schwab F, et al. Health care workers causing large nosocomial outbreaks:a systematic review. BMC Infect Dis 2013;13:98.
[19]. Center for Disease Control and Prevention. Antibiotic Resistance Threats in the United States. Atlanta: Centers for Disease Control and Prevention; 2013.
[20]. National Nosocomial Infections Surveillance System. National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004;32:470–85.
[21]. Kozuszko S, Białucha A, Bogiel T, et al. [High level of aminoglycoside resistance among Enterococcus faecalis and Enterococcus faecium strains]. Med Dosw Mikrobiol 2011;63:105–13.
[22]. Faron ML, Ledeboer NA, Buchan BW. Resistance mechanisms, epidemiology, and approaches to screening for vancomycin-resistant enterococcus in the health care setting. J Clin Microbiol 2016;54:2436–47.
[23]. Emaneini M, Hosseinkhani F, Jabalameli F, et al. Prevalence of vancomycin-resistant enterococcus in Iran: a systematic review and meta-analysis. Eur J Clin Microbiol Infect Dis 2016;35:1387–92.
[24]. Rezaei E, Safari H, Naderinasab M, et al. Common pathogens in burn wound and changes in their drug sensitivity. Burns 2011;37:805–7.
[25]. Othman N, Babakir-Mina M, Noori CK, et al. Pseudomonas aeruginosa infection in burn patients in Sulaimaniyah, Iraq: risk factors and antibiotic resistance rates. J Infect Dev Ctries 2014;8:1498–502.
[26]. Elmanama AA, Laham NA, Tayh GA. Antimicrobial susceptibility of bacterial isolates from burn units in Gaza. Burns 2013;39:1612–8.
[27]. Rajput A, Saxena R, Singh KP, et al. Prevalence and antibiotic resistance pattern of metallo-beta-lactamase-producing Pseudomonas aeruginosa from burn patients–experience of an Indian tertiary care hospital. J Burn Care Res 2010;31:264–8.
[28]. Ullah W, Qasim M, Rahman H, et al. Multi drug resistant Pseudomonas aeruginosa: pathogen burden and associated antibiogram in a tertiary care hospital of Pakistan. Microb Pathog 2016;97:209–12.
[29]. Coetzee E, Rode H, Kahn D. Pseudomonas aeruginosa burn wound infection in a dedicated paediatric burns unit. S Afr J Surg 2013;51:50–3.
[30]. Owlia P, Azimi L, Gholami A, et al. ESBL- and MBL-mediated resistance in Acinetobacter baumannii: a global threat to burn patients. Infez Med 2012;20:182–7.
[31]. Mahdian S, Sadeghifard N, Pakzad I, et al. Acinetobacter baumannii clonal lineages I and II harboring different carbapenem-hydrolyzing-(-lactamase genes are widespread among hospitalized burn patients in Tehran. J Infect Public Health 2015;8:533–42.
[32]. Bahemia IA, Muganza A, Moore R, et al. Microbiology and antibiotic resistance in severe burns patients: a 5 year review in an adult burns unit. Burns 2015;41:1536–42.
[33]. Keen EF 3rd, Robinson BJ, Hospenthal DR, et al. Incidence and bacteriology of burn infections at a military burn center. Burns 2010;36:461–8.
[34]. McDonald LC, Banerjee SN, Jarvis WR. Seasonal variation of Acinetobacter infections: 1987–1996. Nosocomial Infections Surveillance System. Clin Infect Dis 1999;29:1133–7.
[35]. Giske CG, Monnet DL, Cars O, et al. Clinical and economic impact of common multidrug-resistant gram-negative bacilli. Antimicrob Agents Chemother 2008;52:813–21.
[36]. Sahiner F, Ergünay K, Ozyurt M, et al. [Phenotypic and genotypic identification of Candida strains isolated as nosocomial pathogens]. Mikrobiyol Bul 2011;45:478–88.
[37]. Pemán J, Salavert M. [Epidemiology and prevention of nosocomial invasive infections by filamentous fungi and yeasts]. Enferm Infecc Microbiol Clin 2013;31:328–41.
