Multi-Drug Resistant Bacteria as Aetiological Factors of Infections in a Tertiary Multidisciplinary Hospital in Poland

Global and local initiatives were recently undertaken to reduce the burden of antibiotic resistance. The aim of the study was to describe the incidence and the aetiology of bacterial infections among hospitalized patients with special attention paid to the multidrug resistant (MDR) bacteria. This retrospective study was based on prospectively collected data from 150,529 consecutive patients hospitalized in a tertiary multidisciplinary hospital in the years 2017–2019. All consecutive microbiological tests from any biological material performed in the analyzed period were included. Microbiological screening tests (n = 10,677) were excluded. The analysis was focused on aetiological factors of bacterial infections, especially the incidence of MDR bacteria and mechanisms of antibiotic resistance. There were 58,789 microbiological tests performed in the analyzed period. The highest testing rate was noticed for intensive care unit (mean of 3.1 tests per one patient), followed by neonatal intensive care unit (2.7), internal medicine (1.9), pediatrics (1.8), and urology (1.2). Among 58,789 tests, 7690 (13.1%) were positive. MDR bacteria were responsible for 1783 infections (23.2%). The most common antibiotic resistance mechanism reported was ESBL production by Klebsiella spp. or Escherichia coli or Enterobacter spp. isolates (47.3% of all MDR cases). ESBL cases were followed by MRSA (14.7%), VRE (14.2%) and MBL producing Klebsiella spp. (5.6%). Among all infections caused by MDR bacteria, 1175 (65.9%) were diagnosed after 72 h of hospitalization (hospital-acquired infections). Apart from AmpC and ESBL producing Escherichia coli, all MDR bacteria were significantly more common in hospital-acquired infection. MDR bacteria are aetiological factors of a significant portion of infections in hospitalized patients with no remarkable change in the incidence in recent years. Production of ESBL is the most common mechanism of antibiotic resistance and should be regarded as one of the most urgent problems in clinical microbiology.


Introduction
Bacterial infections belong the most common and disabling diseases [1]. Moreover, today they are closely related to the phenomenon of drug resistance, namely antibiotic resistance. According to data from EARS-Net, resistance to at least one antimicrobial therapeutic group in Europe exceeded 57% for Escherichia coli (E. coli) isolates (aminopenicillins 57.1%, fluoroquinolones 23.8%, third-generation cephalosporins 15.1%, and aminoglycosides 10.8%) and 36% for Klebsiella pneumoniae (K. pneumoniae) isolates (third-generation cephalosporins 31.3%, fluoroquinolones 31.2%, aminoglycosides 22.3%, carbapenems 7.9%; intrinsically resistant to aminopenicillins). In 2019, the percentage of combined resistance, measured as resistance to fluoroquinolones, third-generation cephalosporins, and aminoglycosides (multidrug resistant strains) was reported to be 5.9% for E. coli, 19.3% for K. pneumoniae, 29.7% for Acinetobacter spp., and 12.1% for Pseudomonas aeruginosa (P. aeruginosa). Simultaneously, the incidence of meticillin-resistant Staphylococcus aureus (MRSA) was 15.5%, and the incidence of vancomycin-resistant Enterococcus faecium (VRE) was 18.3%. Resistance to vancomycin among Enterococcus faecalis was low [2]. As reported by the World Health Organization, antibiotic resistance is associated with severe negative clinical and economic consequences [3]. At the same time, the incidence of bacteria resistant to antibiotics increases [4][5][6]. Today, there are regions in the world where multidrug resistant (MDR) bacteria are responsible for of >75% of infections in hospitalized patients [7][8][9].
Local and global initiatives have been initiated to reduce the burden of antibiotic resistance in last years. According to recent data from EARS-Net and ESAC-Net study groups, these actions resulted in reduction of antibiotic use and stabilization of the antibiotic resistance incidence [10]. However, such outcomes are not universal. In Poland, the rate of MDR bacteria is increasing along with the use of carbapenems [11]. Next to continuation of the efforts to save antibiotics, detailed and updated data on the incidence and resistance patterns in bacterial infections are needed.
The aim of the study was to describe the incidence and the aetiology of bacterial infections among hospitalized patients with special attention paid to the MDR bacteria.

Results
There were 58,789 microbiological tests performed in the analyzed period. The number of tests did not differ significantly between years 2017, 2018, and 2019 (19,838 vs. 18,123 vs. 20,828 Among 58,789 tests, 7690 (13.1%) were positive. Table 2 presents aetiological factors of reported infections depending on the medical specialty of the hospital department. MDR bacteria were responsible for 1783 infections (23.2%). The most common antibiotic resistance mechanism reported was the production of extended-spectrum beta-lactamases (ESBL) by Klebsiella spp. or E. coli or Enterobacter spp. isolates (47.3% of all MDR cases). ESBL cases were followed by MRSA (14.7%), vancomycin-resistant Enterococcus (VRE) (14.2%), and Klebsiella spp. producing metallo-beta-lactamase (MBL) (5.6%). Among all infections caused by MDR bacteria, 1175 (65.9%) were diagnosed after 72 h of hospitalization (hospital-acquired infections). Table 3 presents a list of MDR bacteria as aetiological factors of infections together with data on the rate of hospitalacquired infections. Apart from AmpC and ESBL producing E. coli, all MDR bacteria were significantly more common in hospital-acquired infection. Table 4 presents a subanalysis of data for different hospital departments. One can see a trend for higher percentage of MDR strains in intensive care unit. Finally, we observed the increase in MRSA incidence and decrease in VRE incidence over time in the analyzed period, with no clinically significant difference in the rate of infections caused by other MDR bacteria (Table 5).

Discussion
We conducted an analysis of the incidence and the aetiology of bacterial infections among representative sample of consecutive hospitalized patients. There are several messages coming from this study.
First, we found that Gram negative bacteria remain the main aetiology of infections, with E. coli and Klebsiella spp. being by far the most commonly isolated bacteria (48% of all infections). Apart from intensive care units, such aetiology is stably observed for a long time [12][13][14].
Second, regarding MDR bacteria, the most common mechanism noticed in this study was the production of ESBL by Klebsiella spp., E. coli or Enterobacter (47% of all MDR cases). This observation became standard and was also reported by Mataj et al. and Wang et al. in Asia [15,16], Acma et al. in America [17], and Arana et al. in Europe [18].
Third, among patients with hospital-acquired MDR infections, the absolute number of cases was still the highest for ESBL producing Klebsiella spp. and Enterobacter spp.; however, a high rate of VRE and MBL producing Klebsiella was also reported (79% of these infections were hospital-acquired). A high rate of ESBL bacterial strains results in increasing clinical pressure to use carbapenems, as well as increasing risk of persistence of carbapenemase producing Enterobacteriaceae (CPE) and transmission. AmpC beta-lactamases and ESBL producing E. coli were the only MDR bacteria that were more likely to be acquired before admission to hospital (71.4% and 59.6%, respectively). Unfortunately, other drug-resistant microorganisms were also found on admission to the hospital, especially ESBL producing Klebsiella spp. and MRSA, which were community-based in 39% and 27% of infections, respectively. This situation requires deep analysis of the risk factors for the multi-drugresistant infection, including a history of a hospitalization, previous antibiotic therapy, and others. Based on this analysis, the extension of screening tests for the early identification of patients infected/colonized with MDR bacteria would be of utmost importance to limit the spread of MDR bacteria in the hospital and to implement effective upfront antibiotic therapy. In face of the high incidence of MDR bacteria among patients with communitybased infections, there is a need to implement rapid diagnostic tests to early diagnose drug resistance. This should be considered in all patients, critically-ill patients, or at least in cases of positive blood cultures and cultured bacterial colonies. Examples of such tests are the carba NP test, identification of resistance mechanisms by mass spectrometry in positive blood samples or the most developing molecular methods such as real-time PCR (diagnosis of sepsis, respiratory tract infections), DNA microarrays, or a diagnostic method utilizing miniaturized magnetic resonance technology [19].
Finally, when comparing three consecutive years included into this analysis, we did notice clinically significant change in the rate of infections caused by MDR bacteria regarding MRSA (increase from 25% to 35%), E. faecium (decrease from 65% to 38%) and E. faecalis (decrease from 11% to 6%) strains. The decrease in rate of MDR-related infections was probably a result of the implementation of supervision over the hospital environment and the implemented strategy to combat Clostridioides difficile in the hospital. Interestingly, a peak of incidence of infections caused by New Delhi type carbapenemase (MBL) producing K. pneumoniae was noticed in the analysed period. However, after they accounted for 7.9% of infections in 2017, these percentage increased up to 14.8% in 2018 to finally decrease to 7.8% in 2019 (Table 5). This decrease is related to the implementation of a multi-module strategy to prevent the spread of CPE implemented by all municipal hospitals in Warsaw (described on the website of the National Antibiotic Protection Program) [20]. This multimodule strategy was implemented in ten municipal hospitals and the long-term care facilities in Warsaw. Moreover, basic rules to prevent and control CPE infections were adopted in all healthcare centers of the city, regardless of the presence of CPE. In brief, all centers were instructed about the indications for screening tests, methods of CPE infection prophylaxis, and management in case of CPE infection or colonization. Procedures were regularly controlled by internal and external audits.
However, it should be noted that the emergence of carbapenem-resistant strains in our hospital resulted in the increased colistin consumption. Since 2018, the occurrence of acquired resistance to colistin has been observed among Enterobacterales, in particular K. pneumoniae and E. coli. Resistance was detected by phenotypic methods, and each resistant strain was sent to a reference center (KORLD) for confirmation. Unfortunately, data on the molecular resistance mechanism is not available in this analysis and it remains to be determined whether it was a mcr gene associated resistance or chromosomal resistance. Due to the coexistence of colistin resistance with the production of ESBLs and/or carbapenemases, we were not able to distinguish a separate group of colistin-resistant strains as a part of microbiological monitoring. However, this phenomenon requires close monitoring as a part of the analysis of cumulative antibiograms, including the use of polymyxins in the hospital. Resistance to colistin in Enterobacterales has been widely reported worldwide, including Europe. The emerging outbreaks of resistant strains are the reason for the limited possibilities of empirical therapy in severe infections in hospital wards with CPE [21].
The significance of MDR bacterial infections is well known and can be considered at least in two aspects, namely clinical and economical. The MDR aetiology of infection, especially ESBL Enterobacterales, was shown to be associated with the prolonged hospitalization time and increased risk of death [22][23][24][25]. In the recent meta-analysis, the presence of cephalosporin-resistant E. coli was associated with over 2-fold higher 30-day mortality [26]. The inverse relation of the spectrum of resistance to antimicrobials and the survival was recently demonstrated by Santoro et al. [27].
From an economical point of view, treatment of MDR bacterial infections is 1.6fold more expensive [28]. Moreover, it is associated with patients' temporal disability. Cephalosporin-resistant E. coli and K. pneumoniae, mainly ESBL producing isolates, are associated with a significant number of disability-adjusted life-years (DALYs). Based on a population-level modeling analysis from Cassini et al., median number of DALYs per 10,000 population is estimated for 37 and 23 for cephalosporin-resistant E. coli and K. pneumoniae, respectively [29].
While in general the data presented in this study is in line with available European data, interesting is the local increase in the rate of MRSA isolation, which is contrary to general European observations reported by European Centre for Disease Prevention and Control [2]. Similar increasing trend was recently described by another Polish research team. Moreover, Kot et al. that the ratio of MDR MRSA isolates is increasing and exceed 92%. Authors concluded about rapid acquisition of new antimicrobial resistance determinants by MRSA isolates in the hospital environment [30]. This needs further concern among clinicians and appropriate action to stop the trend.
Every fifth bacterial infection diagnosed in the analyzed period was of MDR aetiology. There is a general increasing trend of MDR bacterial infections [4][5][6], while the rate of MDR infections can be as high as 90% in Asia or Africa [7,8]. Differences in the rate of MDR infections between countries and regions are well known. They were described by Dornbusch et al. for European regions in 1998 [31]. In Europe, lower rate of MDR bacteria in Nordic hospitals is still observed and was recently confirmed by Moller et al. [32]. Kock et al. indicated that these differences can result from differences in the density of in-patient care, density of hospital personnel, rate of ambulatory surgery, or average length of hospitalization. All these parameters can vary significantly even between neighboring countries [33]. However, significant differences between hospitals are described also within one country, as Kreidl et al. described for two hospitals in Austria [34]. These phenomena can be both a reason and a consequence of different antimicrobial consumption observed in different world regions [35].
Strengths of or study are the inclusion of high number of consecutive patients, prospective collection of data and separate analysis for different medical specialities and years. However, our study is not free from limitations. First, we did not analyze separately different clinical manifestations of infections. Second, data on the impact of MDR bacteria on clinical course was not reported. Finally, data on colistin resistance cannot be presented as, according to local policy, susceptibility to colistin is tested only in carbapenem-resistant strains.

Material and Methods
In the years 2017-2019, 150,529 patients were hospitalized in a tertiary multidisciplinary hospital. This retrospective study is based on prospectively collected data from central microbiological laboratory. All consecutive microbiological tests from any biological material performed in the analyzed period were included. Duplicate isolates, namely further isolates of the same pathogen with indistinguishable antibiograms obtained from the same patient, were excluded. Also, microbiological screening tests (n = 10,677) were excluded. All strains were identified by mass spectrometry (MALDI-TOF Microflex LT, Bruker; Billerica, Massachusetts, USA). Drug susceptibility tests were carried out using the disc diffusion and microdilution method in accordance with the guidelines of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) [36] and the National Reference Centre for Susceptibility Testing in Poland (KORLD) [37]. The following reference strains were used to control the detection of drug resistance mechanisms: ESBL was detected by double disc synergy test (DDST) with 30 µg ceftazidime and 30 µg cefotaxime at a distance of 2 cm (between centers) from the 20/10 µg amoxicillin/clavulanic acid disc. In cases of doubt, discs containing 10 µg of cefpodoxime or 30 µg of aztreonam were added. The positive test was evidenced by the enlargement of the growth inhibition zone around the ceftazidime or cefotaxime (cefpodoxime, aztreonam) disc on the side of the clavulanic acid disc. In order to detect ESBLs in AmpC producers, the test was performed simultaneously on Mueller Hinton Agar with the addition of cloxacillin 0.25 g/L. The cloxacillin test also served as the basis for the detection of the production of AmpC cephalosporinases. The study was performed for epidemiological purposes because according to the EUCAST guidelines, the cephalosporin susceptibility result was based on routine susceptibility testing by microdilution or disc diffusion. This method has been described and is recommended by experts of the National Reference Centre for Susceptibility Testing in Poland [38]. Susceptibility to colistin was tested by broth microdilution using the commercial MIC Colistin Mikrolatest (Erba Lacheme).
The analysis was focused on aetiological factors of bacterial infections, especially the incidence of MDR bacteria and mechanisms of antibiotic resistance. Microbiological monitoring was conducted in the accordance to regulations by the Polish Minister of Health. The source document presents the list of alarming microorganisms and guidelines on the registers for hospital infections and alarming microorganisms and reports on the current epidemiological situation of the hospital (last update on 1/22/21) [39]. Special attention was paid to the following mechanisms: bacteria producing AmpC-type beta-lactamases (AmpC), bacteria producing extended-spectrum beta-lactamases (ESBL), bacteria producing metallo-beta-lactamase (MBL), meticillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococcus (VRE), multidrug resistant non-fermenting gram-negative bacteria Acinetobacter spp., and Pseudomonas aeruginosa (resistant to carbapenems or other two classes of drugs or to polymyxins). Bacteria that did not meet the definition of multidrug resistant organisms (resistant to at least three groups of drugs), such as Streptococcus pneumoniae and Clostridium perfringens, were excluded from the analysis. Secondary analyses were aimed at identification of differences in the rate or aetiology of MDR bacterial infections between years and hospital departments of different medical specialty. Hospital-acquired infections were defined as diagnosed after 72 h of hospitalization according to the National Healthcare Safety Network (NHSN) microbiological monitoring definition (Multidrug-Resistant Organism & Clostridioides difficile Infection (MDRO/CDI) Module) [40].
Results are presented as absolute numbers and percentages. Differences between years and hospital departments were considered only if a minimum of 30 isolates was available for each MDR bacteria.

Conclusions
MDR bacteria are aetiological factors of a significant portion of infections in hospitalized patients. Production of ESBL is the most common mechanism of antibiotic resistance and should be regarded as one of the most urgent problems in clinical microbiology. The problem of drug resistance in hospitals requires constant monitoring and active measures to reduce the risk of MDR bacteria transmission. Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki. Ethical review and approval were waived for this study, as retrospective and non-experimental analyses are not considered by review boards according to local law.

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author. The data are not publicly available due to legal and privacy issues.

Conflicts of Interest:
The authors declare no conflict of interest.