Introduction

Nosocomial infections (NIs) are one of the most significant concerns of medical centers worldwide [1,2,3]. About 80% of the NIs are in the form of nosocomial urinary tract infections (UTIs), nosocomial bloodstream infection (NBSIs), nosocomial surgical site infections (SSIs), and nosocomial pneumonia infection (NPNEU) [4]. The increasing rate of NIs causes more drugs usage, accordingly leading to economic burdens. In spite of global control efforts, NIs still remains a prevalent issue and a cause of antibiotic resistance. The rapid spread of this resistance and its dissemination and burden has become an important health problem throughout the world [5]. According to the Centers for Disease Control and Prevention (CDC) reports, more than 70% of the bacterial agents of NIs are resistant to at least one of the medications used [6].These infections occur 48–72 h after the patient’s admission to different wards of a hospital, particularly the intensive care unit (ICU) [1]. The presence of underlying diseases or dysfunction of organs, immune system disorders, and immunosuppressive drugs, as well as the need for respiratory, cardiac and renal support, makes the patients prone to such infections. Moreover, the use of long-term invasive methods, such as intravenous and arterial catheterization, urinary catheterization, tracheal intubation, and mechanical ventilation and also the use of injective antibiotics have made these infection unavoidable in the ICU [2, 7].

Antibiotic resistance has emerged as a main determinant of outcome in ICU patients, mostly due to the administration of inadequate antibiotic treatment. Gram-positive bacteria (e.g. Staphylococcus species) or mostly Gram-negative bacteria (e.g. Enterobacteriaceae, Pseudomonas, and Acinetobacter species) are the cause of 70% of NIs in the ICUs [8, 9]. Specific Gram-negative bacteria such as Enterobacteriaceae producing extended -spectrum β-lactamases (ESBLs), carbapenemases, multidrug-resistant (MDR) Pseudomonas aeruginosa, and carbapenem-resistant Acinetobacter baumannii are the main reasons for high levels of resistance to antibiotics [2]. Hence, the issue of increasing antibiotic resistance of Gram-negative bacteria in the ICU is greatly worrying and remains a threat to the public health [10, 11].

Enterobacteriaceae, the most prevalent Gram-negative bacteria in NIs, are associated with resistance to several groups of antibiotics, frequently complicating treatment options. Escherichia coli and Klebsiella pneumoniae are the most opportunistic Gram-negative opportunistic organisms and the most common pathogens in NIs [9, 12]. In the Enterobacteriaceae family, the largest producers of ESBLs are often found in E. coli and K. pneumoniae [13,14,15]. The ESBL production by these bacteria has led to resistance of multiple classes of antibiotics and limitation of treatment choices [16]. ESBLs as one of the most significant mechanisms of resistance among Gram-negative bacteria have become a major threat to human health [17].

Carbapenems are still used as a first-line treatment and a therapeutic alternative for severe infections caused by Enterobacteriaceae, especially MDR and ESBLs E. coli [18]. However, the widespread use of these antibiotic agents has led to the development of carbapenem-hydrolyzing enzyme, namely carbapenem-resistant Enterobacteriaceae (CRE). One of the main challenges for ICU physicians, in addition to the prevention of NIs, is to identify the cause of infection and to select an appropriate experimental antibiotic regimen [2, 9, 19].

The type and rate of resistance vary in different regions; therefore, identification of bacterial agents and their resistance act a key role in determining their strategy and decreased bacterial resistance (7,19). Today, NIs and bacterial resistance to antibiotics have dramatically increased, and treatment-resistant NIs are growing. By focusing on the alarming facts, mentioned above, we conducted the present study to evaluate the prevalence of NIs and their antimicrobial susceptibility pattern in the ICUs at Imam Khomeini Hospital, Ilam, Iran.

Materials and methods

Study design and setting

This descriptive cross-sectional study was conducted on patients admitted to the ICUs 1 and 2 at Imam Khomeini Hospital, Ilam, Iran, from July to January 2020. Inclusion criteria for NIs were defined according to the Centers for Disease Control and Prevention (CDC) [20]. Written informed consents were received from each patient.

Phenotypic characterization of the Isolates

Sampling was performed on patients who had been admitted to the ICU for 48–72 h. Clinical samples such as respiratory secretions, blood, middle urine, and other body fluids, as well as wounds were collected. Bacterial identification was carried out using phenotypic and biochemical examinations with standard methods [21]. In brief, clinical specimens were first transferred to the laboratory and then cultured on blood and MacConkey agars (both from Merck Co., Germany). Thereafter, differential tests were used to identify bacterial species and genus.

Antimicrobial susceptibility testing

Kirby-Bauer disk diffusion method was performed on Müller-Hinton agar (MHA) (Merck Co.) in accordance with the Institute of Clinical and Laboratory Standards guidelines (CLSI 2020) [22]. Antimicrobial disks used for Gram-negative bacteria included Amikacin, trimethoprim/sulfamethoxazole, ciprofloxacin, imipenem, meropenem, aztreonam, cefotaxime, ceftazidime, ceftriaxone, and cefpodoxime, 30 μg for each (TAV, Turkey). Linezolid, clindamycin, tigecycline, penicillin, amikacin, gentamicin, cefoxitin, and trimethoprim/sulfamethoxazole (30 μg for each) were antimicrobial disks used for Gram-positive bacteria. To evaluate the quality control of the discs, we utilized E. coli ATCC 25,922 standard strain and S. aureus 29,213 for Gram-negative and Gram-positive bacteria, respectively. Following an overnight incubation at 37 °C, the size of inhibition zone was measured and interpreted as sensitive, intermediate and resistant.

Phenotypic screening of ESBLs, MBLs, and CRE in E. coli and K. pneumoniae

E. coli and K. pneumoniae isolates related to NI were used for phenotypic ESBLs, metallo-β-lactamases (MBLs) and CRE tests. The initial screening was performed with five antibiotics, including cefotaxime (30 μg), ceftazidime (30 μg), ceftriaxone (30 μg), cefpodoxime (10 μg), and aztreonam (30 μg) (TAV). In each antibiotic, the inhibition zone resistance was examined by confirmatory tests according to CLSI 2020 [22]. Briefly, ceftazidime (30 μg) and ceftazidime/clavulanate (30/10 μg; TAV) discs were used as the combination disk test. Sensitivity zone diameter of at least 5 mm in the composite disk compared to ceftazidime was considered as ESBLs. K. pneumoniae ATCC 700,603 was employed as the control strain. To determine MBLs, the two discs of imipenem (10 μg) with a distance of 15 mm were placed on MHA plate that was inoculated by 0.5-McFarland standard. One imipenem disc was impregnated with 10 μl of 0.5 M ethylenediamine tetraacetic acid (EDTA). The plates were incubated at 35 °C for 16–18 h. In the combination disk with EDTA, the sensitivity zone diameter of 7 mm and higher was considered as MBLs.

Phenotypic screening of CRE resistance in E. coli and K. pneumoniae

The modified carbapenem inactivation method was used to detect CREs resistance in E. coli and K. pneumoniae. After incubation of the meropenem disc in bacterial suspension for 2 h, the disk was replaced on MHA (Merck Co.) and inoculated with E. coli ATCC 25,922 as the standard strain. CRE activity was distinctive by no zone on MHA (Merck Co.) [22].

Molecular detection of ESBLs, MBLs, and CRE genes

Following the bacterial DNA extraction by boiling method [23], polymerase chain reaction (PCR) was performed for specific primers shown in the Table 1. Genes encoding ESBLs (blaCTX-M, blaOXA-11, and blaSHV), MBLs (blaNDM, blaVIM, and blaIMP), and CRE (blaKPC, blaOXA-48, and blaOXA-23,) were detected.

Table 1 Primers used in this study
Full size table

Statistical analysis

Statistical analysis was conducted using the “IBM SPSS statistics 22” software (IBM analytics; USA). Statistical significance of variables was determined by chi-square and Fisher's exact tests. The results were presented as descriptive statistics in terms of relative frequency. A p-value of ≤ 0.05 was considered statistically significant.

Results

Patient characteristics and prevalence of NIs

A total of 602 patients in the ICU 1 (n = 239; 70.39%) and ICU 2 (n = 363; 30.605) were recruited and analyzed for six months. The average length of hospital stay on the admission was 22 days. The lowest stay was two days, while the maximum stay was 217 days. According to the positive cultures, 71 patients, including 44 (61.97%) men and 27 (38.03%) women with the mean age of 49.53 years (± 23.06 standard deviation), showed NIs. The patients in the age range of 14–29 years had the most hospital infections. Overall, the prevalence of NIs in ICU1 was estimated as, 17.99% and 7.77% in ICU 2. The prevalence of NI was significantly different in both ICUs. However, the rate of this infection in ICU 1 was higher than that of ICU 2 (p < 0.001). The detailed information of the patients is represented in Table 2.

Table 2 Characteristics of study patients
Full size table

Bacteria causing NIs

In total, 71 patients had positive culture. Of 103 bacterial isolates identified, 79 isolates (76.70%) were Gram-negative, and 24 (23.30%) were Gram-positive. The most prevalent Gram-negative bacteria were E. coli (n = 29; 28.16%), A. baumannii (n = 15; 14.56%), and K. pneumoniae (n = 13; 12.62%), whereas the most common bacterium was S. aureus (n = 12; 11.66%), as shown in Fig. 1. Regarding the type of clinical infection, respiratory system (52; 50.48%) was the most common organ involved, followed by urinary tract system (n = 26; 25.25%), bloodstream (n = 22; 21.36%), and wound (n = 3; 2.91%), as depicted in Table 3. A. baumannii and K. pneumoniae were the most common bacteria isolated from the respiratory tract system, and E. coli was the most prevalent infectious agent of urinary tract system.

Fig. 1
figure 1

Pathogens causing nosocomial infections in ICUs

Full size image
Table 3 Distribution of bacterial agents isolated and infections sites from patients with Nis
Full size table

Antimicrobial susceptibility pattern

Antimicrobial susceptibility pattern in Gram-negative bacteria showed the highest resistance to third-generation cephalosporins, including ceftazidime (n = 69; 87.34%), cefotaxime and cefpodoxime (n = 59; 74.69%), and ceftriaxone (n = 58; 73%). Moreover, the highest susceptibility was observed for amikacin (n = 25; 34.61%), followed by imipenem and meropenem (n = 27; 34.17%). Among Gram-negative bacteria, A. baumannii showed the highest resistance to most of antimicrobials. Gram-positive bacteria showed the highest resistance to penicillin (n = 22; 91.67%) and cefoxitin (n = 19; 79.17%), and the highest susceptibility to tigecycline (n = 24; 100%), followed by trimethoprim/sulfamethoxazole (n = 23; 84. 95%) and linezolid (n = 22; 91.67%), as illustrated in Table 4. Multidrug resistance is defined as resistance of a microorganism to at least one agent in three or more antimicrobial categories. Overall, the prevalence of MDR bacteria was 58.25% (n = 60). E. cloacae (n = 4; 100%), S. aureus (n = 11; 91.66%), and A. baumannii (n = 13; 86.66%) were the most commonly MDR bacteria isolated. The ratio of the number of MDR isolates to the total number of isolates was statistically significant (p < 0.001;

Table 4 Antimicrobial susceptibility pattern Gram negative (A) and Gram positive (B) Gram negative
Full size table

See supplementary file 1 for more details).).

Phenotypic detection of ESBLs and CREs in E. coli and K. pneumoniae isolates

A total of 42 isolates, including 29 E. coli and 13 K. pneumoniae isolates related to NI, were obtained in ICUs 1 and 2 and used for phenotypic tests. Of 42 isolates of E. coli and K. pneumoniae, 32 (76.19%) isolates, i.e. 22 (75.86%) E. coli and 10 (76.93%) K. pneumoniae, were ESBL producers. Also, 10 (23.81%) isolates were non-ESBLs isolates. Based on the phenotypic detection of CREs, 6 (14.28%) isolates, including 2 (6.89%) E. coli and 4 (30.77%) K. pneumoniae isolates, were identified as CRE producers (Table 5).

Table 5 Distribution of ESBL by phenotypic tests
Full size table

Molecular detection of ESBLs, MBLs, CRE genes and PCR product sequencing

The blaCTX-M gene encoding ESBLs was detected in 29 (90.62%) isolates, including 22 (100%) E. coli and 7 (70%) K. pneumoniae isolates. The blaSHV was identified in 7 (21.87%) isolates of E. coli (1; 4.54%) and K. pneumoniae (n = 6; 60%). The blaOXA-11 encoding ESBLs was found in 10 (31.25%) isolates of E. coli (n = 7; 31.81%) and K. pneumoniae (n = 3; 30%). The blaNDM gene was detected in 4 (66.66%) K. pneumoniae isolates, and the blaOXA-23 and blaOXA-48 genes were observed in 3 (50%) K. pneumoniae and only 1 (16.66%) E. coli isolate, respectively. However, the blaVIM, blaKPC, and blaIMP genes were not found in any isolates (Table 6). PCR sequencing products were performed for five blaNDM and one blaOXA-23 gene products, and 98% nucleotide similarity was observed in both genes. In addition, four blaNDM gene products were identified as blaNDM-1.

Table 6 ESBLs and carbapenem resistance patterns genes among K. pneumoniae and E. coli isolates
Full size table

Coexistence of ESBL- and CRE-encoding genes in E. coli and K. pneumoniae isolates

PCR confirmed the coexistence of blaCTX-M and blaSHV in five isolates and blaCTX-M and blaOXA-11 genes in 10 isolates, and three genes (blaCTX-M, blaSHV, and blaOXA-11) encoding ESBLs were harbored by two K. pneumoniae isolates. In addition, the coexistence of two genes, blaNDM-1 and blaOXA-23, were observed in three isolates of K. pneumoniae. Also, three K. pneumoniae isolates carried four genes, including blaCTX-M, blaSHV, blaNDM, and blaOXA-23, and two K. pneumoniae isolates harbored five genes, including blaOXA-11, blaCTX-M, blaSHV, blaNDM -1 and blaOXA-23.

(See supplementary file 2 for more details).).

Discussion

NIs are one of the major global health problems involved almost all hospitalized patients at risk for these infections [10]. In is study, the prevalence of NIs in both ICUs was reported as 11.79%, which was 17.99% in ICU 1 and 7.77% in ICU 2. Previous studies conducted in Ahvaz and Gorgan cities of Iran have reported 10% and 13% NI prevalence in ICU, respectively [24, 25]. The rate of NI prevalence obtained in our study for ICU was lower than the estimates provided by studies in Zahedan (76.9%) and Birjand (27.2%) cities. Several reasons for differences in the prevalence of NIs by diverse studies in Iran might be variations in the study periods and type of units under investigation, as well as unreported real rate of infections. In the present study, NIs caused by fungal infections were not studied due to the limited number of samples compared to other investigations conducted in Iran [26, 27], Northern India (30.4%) [28], and China (26.07%) [29]. The length of patient stay is a risk factor and associated with the increased incidence of NIs. There was a difference in the number of hospitals and that of patients in prior studies [28, 29].

The most common infections in our study were respiratory tract infections, followed by urinary tract infection. In a previous study comparing NIs in the internal ICU with those of surgery ICU demonstrated that respiratory tract infection and UTI were the most common infections in the two ICUs, respectively. In other investigations in India and Germany, respiratory tract infection was the most common infected organ. Patients in the ICU require supportive treatment; therefore, mechanical ventilation and urinary catheterization are widely used, which can be risk factors for the infection of the patients in this unit [-31].

In the current study, Gram-negative bacteria were more identified than Gram-positive bacteria in positive cultures. Similar to our result, Nouri et al. [32] reported Gram-negative bacteria as the main causes of NIs. We also found that the most frequent Gram-negative bacteria were E. coli and A. baumannii, and Gram-positive bacterium was S. aureus. This outcome is in agreement with an earlier study in Hamadan showing E. coli and K. pneumoniae as the most prevalent Gram-negative bacteria and S. aureus as the most frequent Gram-positive bacterium [32].

Another study has also reported K. pneumoniae, Acinetobacter species, and P. aeruginosa as the most prevalent Gram-negative bacteria and S. aureus the most prevalent Gram-positive bacterium. Gram-negative bacteria have the greatest diversity, thus achieving high adaptability in the environment. These bacteria are flora in the human digestive system and skin and simply transfer from equipment and personnel to patients. S. aureus is a flora of the skin and nasal and can easily spread [7, 32, 33]. The prevalence of antibiotic-resistant bacteria has become a public health concern worldwide. Increased antimicrobial resistance among microorganisms causing NIs is associated with high mortality in hospitalized patients.

Antimicrobial susceptibility testing in Gram-negative bacteria showed the highest susceptibility to amikacin, followed by imipenem and meropenem. Also, the lowest susceptibility rates were related to ceftazidime, cefotaxime, and cefpodoxime, respectively. The result of our study showing that A. baumannii had the most resistance to all antimicrobial categories were consistent with other findings in Iran and Uganda [7, 34].

In Gram-positive bacteria, tigecycline, trimethoprim/sulfamethoxazole, and linezolid had the highest efficiency. Also, the highest resistance was related to penicillin, which supports Tolera et al.’ investigation in Ethiopia [10]. The prevalence of MRSA in our study was higher than a previous study conducted in Gauteng Academic Hospital in South, which seems worrying [35]. These differences in high level of antibiotic resistance could be due to regional and genetic variations of species and also genetic differences between individuals. However, it is thought that factors such as long-term hospitalization of patients, continuous use of broad-spectrum antibiotics during hospitalization, cultural diversity, irregular consumption of drug, and self-treatment are the most important factors affecting the high-level resistance of antibiotics [36, 37].

In the present study, the phenotypical prevalence of ESBLs was observed in 32 (76.19%) E. coli and K. pneumoniae isolates, which in Hasani’s study, the prevalence rate of ESBLs was 87.8% [37]. However, different prevalence has been reported in Iran and other countries. In Bialvaei’s study, the rates of ESBLs production were 27.27% in E. coli and 25.9% in K. pneumoniae [38].

According to a systematic review and meta-analysis, the prevalence of ESBLs was reported as 43.2% in Iran, which is higher than the rate stated in developed countries such as France (1.5%) and Germany (3.3%) [39]. In our study, sample collection was carried out in ICUs. In this unit the use of broad-spectrum antibiotics is very high, which can be a reason for the high level of ESBLs. The investigation of ESBL genes showed high prevalence of the blaCTX-M and low prevalence of blaSHV. Similarly, previous studies have reported the prevalence of 92.5% and 100% for blaCTX-M in E. coli and K. pneumoniae isolates, respectively [37, 40].

The low prevalence of blaSHV was confirmed by a former study conducted in Khuzestan (3.5%) and Tehran cities (10%) of Iran [41, 42]. Phenotypical tests exhibited that 6 (14.28%) isolates of E. coli and K. pneumoniae harbored MBLs and CRE. This result is in line with a previous study showing that the overall rates of carbapenem resistance were 24% and 5% in K. pneumoniae and E. coli isolates, respectively [43].

In molecular detection, the most common carbapenemase genes included blaNDM-1 and blaOXA-23, which former gene was mostly prevalent in K. pneumoniae, and the latter gene was mainly detected in K. pneumonia isolates. Shahcheraghi et al., identified the first blaNDM-1 among K. pneumonia isolates in Iran [44]. Fazeli et al., [45] and Moghadampour et al., [46] have also found blaNDM-1 gene among K. pneumonia isolates, which is consistent with Huang’s finding conducted in China [41, 45,47,47]. In our study, the genes blaIMP, blaVIM, and blaKPC were not detected, while blaNDM, blaOXA-48, and blaKPC were the genes not identified in Kiaei et al.’s study in Keramn [48]. In this study, the prevalence of blaNDM and blaOXA-48 were similar or almost similar to Jalalvand’s study in Tehran city [49].

K. pneumoniae and E. coli are the most important pathogens of NIs and can be associated with increased mortality and mortality related to carbapenem resistance worldwide. The available antibiotics for carbapenem-resistant isolates are limited. Furthermore, isolates carrying the blaNDM-1 gene are MDR and resistant to almost all beta-lactam antibiotics, fluoroquinolones, and aminoglycosides. Therefore, the emergence and spread of these isolates is a great challenge and should be taken into consideration [50, 51]. Unsupervised and indiscriminate use of antibiotics has always been important factors in creating resistance and reducing treatment options. Thus, application of a specific guideline like antimicrobial stewardship program and establishment of surveillance systems to analyze antibiotic resistance pattern can decline trends of bacterial resistance and NIs in the ICU. This study had some limitations. The small sample size and single center study were among these limitations. Other limitations included lack of access to patient records, including the type of NIs, antibiotic prescription history, and risk factors for antibiotic-resistant bacteria. Moreover, the identification of bacterial with culture and biochemical tests and the time of the study may be potential sources of bias and possible confounder of this study, respectively. To generalize the results to the entire study area, it is suggested that further studies be conducted periodically with a larger sample and more centers.

Conclusion

In the present study, the most NIs was identified in the ICU 1. Moreover, the Gram-negative bacteria E. coli, A. baumannii, and K. pneumoniae with high levels of resistance were introduced as the most common bacteria causing NIs in the ICU. The interesting point of our study was that we, for the first time, detected blaOXA-11, blaOXA-23, and blaNDM-1 genes in E. coli and K. pneumoniae isolates in Ilam city of Iran. These genes showed high resistance to carbapenems, which is a significant issue that needs to be taken into account. Therefore, identifying endemic pthogenic bacteria, source of infection, determining antibiotic susceptibility pattern in ICU, finding and implementing an appropriate strategy for diagnosing, prescribing medication, and monitoring infection control could be effective in reducing the rate of such resistance, NIs, and also patients’ mortality.