Multidrug-Resistant Hospital Bacteria: Epidemiological Factors and Susceptibility Profile

Introduction

In a survey commissioned by the UK government, O'Neil ¹ estimated that by 2050, about 10 million people will die annually from infections caused by bacteria with multiple resistance genes. These mortality data are higher than those for cancer, with ∼8.2 million people per year. This prediction, however, has been questioned by De Kraker et al. ² who consider it overestimated.

Bacterial resistance to antibiotics has been particularly detected in bacterial pathogens, ESKAPEEc (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp., and Escherichia coli), which are the main cause of nosocomial infections.^3–7 Gram-positive cocci (GPC), such as E. faecium and S. aureus, have proved to be resistant to antimicrobial drugs, causing severe and even lethal infections, such as pneumonia, meningitis, and skin infections. ⁸ Multiresistant Gram-negative bacilli (GNB), such as K. pneumoniae, A. baumannii, P. aeruginosa, and Enterobacter spp, caused several infections of the urinary tract, blood (sepsis), pneumonia, and others, especially when catheters are used. ⁹

Whereas several studies have been carried out in the United States^10–12 on multiresistant drugs in hospitals, there are only a few publications on the subject in Brazil.^13–17 In southern Brazil, they seem to be limited surveys conducted in Santa Maria RS Brazil. ¹⁸ Since a controversy on estimates exists among researchers, it seems relevant to produce specific and localized studies on the state of bacterial resistance for further epidemiological information on the debate.

Current analysis discusses the susceptibility profile of multiresistant bacteria (MRB) from two hospitals in Pelotas RS Brazil, with regard to epidemiological factors (patient's data, sample origin, and site of infection).

Materials and Methods

The current study was approved by the Committee of Ethics on Research (CER) n. 2,961,379, 2,985,372, and by the National Committee for Ethics in Research, n. 2,880,831 (Plataforma Brasil).

The study was performed in two hospitals in Pelotas RS Brazil, between October 2018 and January 2019, comprising 286 antibiogram (N = 286) reports of patients hospitalized in Intensive Care Units (ICUs), Neo Natal Intensive Care Units (NICs), Wards (WARs), private and health insurance schemes (PHIs), and outpatients (OPTs), of both genders, divided into age groups: Newborns (NB) 0–2 years of age; 14–19 years of age; 20–49 years of age; 50–59 years of age; 60–69 years of age; and over 70 years of age.

The study was carried out in the municipality of Pelotas, state of Rio Grande do Sul, in the southern region of Brazil. According to the Brazilian Institute of Geography and Statistics (in Portuguese, “Instituto Brasileiro de Geografia e Estatística” (IBGE) ¹⁹ ), Pelotas's estimated population in 2018 was 341,648 inhabitants, with a density of 203.89 inhabitants per km².

The estimated sample number (n) was based on the formula n = [Np (1-p)]/[(d2/Z21)-α/2*(N-1)+p*(1-p)], ²⁰ considering an expected frequency (p) of 80 (d), a maximum error (d) of 5%, and a population considered infinite (N), with a minimum sample number of 246 people to represent the population of the city. However, 286 people were evaluated to increase the reliability of the information obtained.

Bacterial isolates, collected from lung, body fluids, skin, blood, and urine samples, proved resistant against at least three classes of antibiotics. The identification of bacterial isolates and their susceptibility profiles against standard antibiotics were performed by the Clinical Analysis Laboratories of each hospital under analysis, by automated VITEK^® 2 system (bioMérieux, France) and BD PHOENIX, following recommendations by the Clinical and Laboratory Standards Institute (CLSI). ²¹

Tested antibiotics were distributed into the following classes (16): Quinolones [Nalidixic acid (NAL), Levofloxacin (LEV), Moxifloxacin (MFX), Norfloxacin (NOR); Ciprofloxacin (CIP)]; Sulfonamides [Trimethoprim/sulfamethoxazole (SUT)]; Penicillins [Ampicillin (AMP), Penicillin (PEN), Oxacillin (OXA)]; Cephalosporins [Cephalin (CFL), Cefoxitin (CFO), Cefuroxime (CRX), Ceftriaxone (CRO), Cefepime (CPM), Cefotaxime (CTX), Cefazolin (CFZ), Ceftazidime (CAZ), Ceftaroline (CPT)]; Carbapenems: [Imipenem (IPM), Meropenem (MER), Ertapenem (ETP)]; β-lactamase Inhibitors [Ampicillin/sulbactam (ASB), Ampicillin/clavulanic acid (AMC), Piperacillin/tazobactam (PIT)]; Polypeptides [Colistin (CLS)]; Glycylcyclines [Tigecycline (TGC)]; Nitrofurans [Nitrofurantoin (NIT)]; Aminoglycosides [Amikacin (AMI), Streptomycin (ETR), Gentamicin (GEN)]; Glucopeptides [Teicoplanin (TEI), Vancomycin (NPV)]; Macrolides [Erythromycin (ERI)]; Oxazolidinones [Linezolid (LNZ)]; Lincosamides [Clindamycin (CLI)]; Rifampicin [Rifampicin (RIF)]; Tetracyclines [Minocycline (MNO)]; Fusidanes [Fusidic acid (AFD)]; and Lipopeptide [Daptomycin (DAP)].

Descriptive statistical analyses with SPSS statistical software 20.0 were carried out to determine the distribution of resistance to antibiotics, most affected age group, gender, sample origin and site of infection.

Results

The reports of 286 patients were analyzed. Patients were distributed into two hospitals (Hospital A with 128 patients and Hospital B with 158 patients), during a 90-day period, ranging between zero (NB) and 96 years of age, distributed into the following age group: 0–2 years of age (5.2%), 14–19 years of age (1.4%), 20–49 years of age (24.1%), 50–59 years of age (17.1%), 60–69 years of age (20.3%), and over 70 years of age (31.8%). Furthermore, (N = 156; 54.5%) were males and (N = 118; 41.3%) were females; (N = 12; 4.2%) were people with nonidentified gender.

Data on origin of sample by hospitalization site comprised WAR 55.6%; ICU 23.4%; OPT 10.2%; PHI 6.6%; and NIC 4.2%. Data on infection sites comprised urine (40.6%); lungs (30%); blood (18.2%); skin (10.5%); and body fluids (0.7%).

Isolated and identified MRB were (N = 53; 17.83%) K. pneumoniae; (N = 46; 16.08%) A. baumannii; (N = 22; 7.69%) S. aureus; (N = 20; 6.99%) P. aeruginosa; (N = 20; 6.99%) Enterobacter spp.; (N = 6; 2.10%) E. faecium; and (N = 44; 15.38%) E. coli, belonging to the ESKAPEEc group; (N = 16; 5.59%) Coagulase-negative staphylococci (CoNS), (N = 8; 2.80%) Citrobacter freundii, and (N = 8; 2.80%) Staphylococcus haemolyticus, (N = 7; 2.45%) Staphylococcus epidermidis, (N = 6; 2.10%) Staphylococcus hominis; Klebsiella oxytoca, Proteus mirabilis, Serratia marcescens, and Morganella morganii (N = 5; 1.75%) each; Providencia rettgeri and Stenotrophomonas maltophilia (N = 4; 1.40%) each; Aeromonas sobria, Shigella flexneri, Proteus vulgaris, and Streptococcus sp. (N = 1; 0.35%) each, not belonging to the group. Table 1 shows patients' epidemiological data, hospitalization site, and place of infection, with regard to bacteria reported.

When bacterial genera (GPC and GNB) were taken into account, the former showed resistance to at least three and, at the most, 10 antibiotics, within each class of antibiotics; the latter tested to at least four and, at the most, 18 antibiotics, within each class. Table 2 shows the number of resistant bacteria against the antibiotic classes tested.

Antibiotic classes with the greatest number of resistant GNB were penicillins (84.8%), quinolones (77.5%), and cephalosporins (75.7%). In the case of GPC, the most resistant were macrolides (95.4%), lincosamides (90.3%), and penicillin (77%). Among the GNBs, polypeptides (81.3%) and glycylcycline tested only for tigecycline (64%), and aminoglycosides tested for amikacin and gentamicin (61.9%), with greater sensitivity; among the GPCs, fusidanes, glycylcyclines, and lipopeptides tested 100% (Table 3).

Among GNB and GPC (286), K. pneumoniae (27.5%), A. baumannii (24.1%), E. coli (14.7%), and P. aeruginosa (14.5%) were the bacteria with the highest resistance rate to antibiotic classes. Table 4 shows the epidemiological profile of the highest bacterial resistance rate against antibiotic classes.

Discussion

Data revealed that males and groups over 60 years of age (55.1%) were predominant targets in infections by MRB, suggesting immunodepressed patients. Results are compatible with other Brazilian studies, such as research conducted in the state of Rio Grande do Sul by Lorenzoni et al. ¹⁸ who detected a higher occurrence in males (52.1%) and in the over-60-year age bracket (51.3%); by Souza et al. ¹⁷ who reported higher occurrence rates in people 80 years of age or older (62.2%), and by Seibert et al. ¹⁶ who registered higher rates in males (72.3%) and in people over 60 years of age (57.5%). International studies by Cek et al. ²² identified predominance of males (70.4%) and people averaging 59.9 ± 18.2 years of age.

Hospitalization site was another relevant aspect in the current study. Ahn and Prince ²³ demonstrated that ICUs were environments with the highest rates for HAIs. In fact, ICUs are high-risk environments involving immunosuppressed patients, featuring invasive procedures, such as mechanical ventilation and catheters, prolonged hospitalization, the occurrence of opportunistic microorganisms capable of forming biofilms, and others. However, the predominant site of hospitalization in the current research was WAR with 55.6%, followed by adult ICU with 23.4%, and OPTs with 10.2%; PHI and NIC featured lower rates of 6.6% and 4.2%, respectively. Results differed from those in most Brazilian studies, which reported that a greater spread of MRB occurred in ICUs with 17.7% ¹⁸ and 25.6%.^16,17

The urinary tract is the predominant site of infection, with 40.6%, mainly due to E. coli with a higher frequency rate of 31 isolates, with 85.7% prevalence in urine, followed by K. pneumoniae with 30 isolates (56.6%). The current study demonstrates that urine and lungs are the most prevalent infection sites in the age group 20–49 years of age, but no comparative studies were found. Vincitorio et al. ²⁴ insisted that, due to catheters, urinary infections are a recurring cause of hospital infections worldwide. According to these authors, the main risk factors associated with these infections are elderly patients, abnormalities of urinary functions, immunosuppression, prolonged hospitalization, inadequate use of catheters, in many cases, even without adequate medical indication.

Other more frequent infection sites identified in the current study were lungs (30.0%), blood (18.2%), skin (10.5%), and body fluids (0.7%). Another Brazilian study by Lorenzoni et al. ¹⁸ found similar results, with the following rates: urinary (38.0%), pulmonary (21.6%), rectal (18.0%), and blood (5.3%) infections. However, Seibert et al. ¹⁶ forwarded different results and pinpointed the predominance (25.5%) of tracheal aspirates.

The current study identified the bacterial species K. pneumoniae (17.83%), A. baumannii (16.08%), and E. coli (15.0%) as predominant among GNBs. Seibert et al. ¹⁶ reported that GNB, which produces the β-lactamase enzyme called extended-spectrum β-lactamase (ESBL), are highly resistant to the β-lactam class due to the inactivation of the antibiotic by the break of the β-lactam ring. Our study showed 32.3% of ESBL-producing bacteria, with the highest occurrence rate for K. pneumoniae (15.0%) and E. coli (14.6%). Studies by Yayan et al. ¹¹ have shown that the prevalence of these isolates is constant in other countries. The authors reported that K. pneumoniae is one of the most frequent of the genus Klebsiella in HAIs and in community infections in patients with pneumonia in Germany. In a 10-year survey, Liu et al. ²⁵ identified E. coli (47.0%), Klebsiella spp. (26.3%), Salmonella spp. (10.4%), and Enterobacter spp. (9.2%) as prevalent bacteria. It has also been reported that ESBL was predominant in the same bacteria (E. coli and K. pneumoniae) with the following percentages for E. coli: (68.9%), from 2004 to 2005; (73.2%), from 2007 to 2008; (67.9%), from 2009 to 2010; (72.6%), from 2011 to 2012; and (58.4%) from 2013 to 2014; for K. pneumoniae with 75.9%, 50%, 41.4%, 40.2%, and 43.0%, respectively, at the same intervals.

Research by Rossi et al. ¹⁵ conducted in São Paulo, Brazil, during 4 years, indicated Enterobacteriaceae as the most frequent isolate, with the following percentages: (49%) for K. pneumoniae; (29%) for Pseudomonas spp.; and (22%) for Acinetobacter spp. Research by Souza et al. ¹⁷ in Londrina, Brazil, demonstrated that K. pneumoniae was the most prevalent MDRs at hospital discharge (19.0%) and at death (21.2%). A. baumannii (18.5%) ranked second in patients who died, whereas P. aeruginosa (11.3%) was highly frequent in hospital-discharged patients.

Carbapenems are last-resort broad-spectrum antibiotics for the treatment of HAIs in GNB, and for ESBL bacteria when they are resistant to multiple drugs and to other β-lactams (penicillins and cephalosporins). Although resistance to antibiotics by P. aeruginosa is a worldwide problem, Dantas et al. ¹³ concluded that the high incidence of carbapenem-resistant P. aeruginosa in Brazil is caused by the use of drugs resistant to glycopeptides. In the current study, A. baumannii (33.1%), K. pneumoniae (23.4%), and P. aeruginosa (19.9%) were greatly resistant to carbapenems, whereas E. coli demonstrated special sensitivity to them. These results corroborate those by Oliveira et al. ¹⁴ who conducted surveys in a hospital in Brazil between 1999 and 2008. The authors also detected an increase in the resistance of A. baumannii to antibiotics, with 7.4% at the start of the study and 57.5% at the end.

Pereira-Maia et al. ²⁶ reported that tetracyclines and glycylcycline (analogous to tetracycline) are some of the few antibacterial drugs available in the treatment of infections caused by methicillin-resistant S. aureus (MRSA). In the case of the glycylcycline class, the current study revealed that GNBs had a 64.0% sensitivity to tigecycline. In fact, it was the only antibiotic of this class to be tested and the most effective (40.4%) in the control of infections caused by A. baumannii. All bacteria tested (100%) among the GPCs were sensitive to glycylcyclines, whereas only 28% had any sensitivity to tetracyclines. On the other hand, among the GPCs, S. aureus isolate showed sensitivity to tigecycline (18.0%) and minocycline (35.7%). Consequently, our conclusions are compatible with those by Lorenzoni et al. ¹⁸ who showed a 66.1% bactericidal sensitivity to tigecycline. Moreover, Seibert et al. ¹⁶ also reported almost similar rates (69.4%) for tigecycline sensitivity.

In the current study, A. baumannii (45.9%), K. pneumoniae (18.0%), and P. aeruginosa (16.4%) were bacteria with the highest sensitivity rate to aminoglycosides (gentamicin and amikacin) in GNB. On the other hand, S. aureus (61.1%) and CoNS bacteria (22.2%) showed the best antimicrobial activity results in this class, with 61.5% in GNB and 60.5% in GPC. Other Brazilian studies on the subject were developed by Seibert et al. ¹⁶ and Lorenzoni et al. ¹⁸ The former showed 91.5% sensitivity to amikacin and 57.4% to gentamicin, whereas the latter presented 78.6% and 64.9% sensitivity for amikacin and gentamicin, respectively. Seibert et al. ¹⁶ demonstrated that aminoglycosides are an alternative in the treatment of carbapenem-resistant Enterobacteriaceae and demonstrated that amikacin, gentamicin, and tigecycline were not recommended for severe systemic infections, and that monotherapy is not indicated due to increased microbial resistance. A study by Micek et al. ¹² forwarded the following sensitivity indexes to aminoglycosides, namely, United States (80.2%), France (39.6%), Italy (75.2%), Spain (58.9%), and Germany (58.3%). Results by Micek et al. ¹² are partially compatible with current ones, with approximate rates for Spain and Germany and discordant for United States, Italy, and France.

Although colistin was abandoned in the last decades of the 20th century due to its toxicity to the kidneys, its use was retaken as the last therapeutic option in the treatment of MDR. The reuse of colistin has produced an increased resistance of bacteria to therapeutic efficacy worldwide, due to genetic mutations by the mcr-1 gene, as pointed out by Wang et al. ²⁷ and Rossi et al. ¹⁵ The current study only tested the effects of colistin among the class of polypeptides, with higher sensitivity (81.3%) among all antibiotic classes. Whereas A. baumannii was the most sensitive bacterium (45.9%), K. pneumoniae and P. aeruginosa contrastingly had a 28.6% increased resistance rate to colistin. In Brazil, studies by Rossi et al. ¹⁵ showed that Enterobacteriaceae isolates had a high resistance to colistin, ranging between 6.6% (N = 111) in 2010 and 9.4% (N = 383) in 2014. The same authors observed that K. pneumoniae had a critical index, or rather, an 84.1% resistance to colistin. This fact did not occur among nonfermentative isolates, such as Acinetobacter spp. (1.4%) and Pseudomonas spp. (4.0%), both still very susceptible to colistin. An international study by Prim et al. ²⁸ detected a 0.67% resistance to colistin and underscored that the most resistant bacteria were Enterobacter cloacae (4.2%), E. coli (0.5%), and K. pneumoniae (0.4%).

GNB of the Enterobacteriaceae family have become resistant to penicillin because they produce β-lactamase enzymes that degrade the β-lactam ring. The development of penicillinase enzymes in Gram-positive (GP) and Gram-negative (GN) bacteria led to the production of new β-lactams, namely, cephalosporins (Yayan et al. ¹¹ ). In this investigation, MDR showed resistance to the class of penicillins in GN (84.8%) and in GP (77.4%), as Table 3 demonstrates. Zhang et al. ²⁹ revealed that between 2012 and 2016, there was a high bacterial resistance to penicillins with lower susceptibility in S. aureus, reported worldwide at 14.7% for penicillin. The following indexes were described for Africa with 2.3%; Asia with 6.0%; Europe with 15.9%; North America with 16.6%, and Latin America with 8.1%. The susceptibility of Enterococcus faecalis and E. faecium was 12.9% and 11.9% for penicillin and ampicillin, respectively. The overall rate for penicillin-resistant S. pneumoniae isolates was 9.0% worldwide, or rather, Asia with 25.8%; Africa with 13.0%; Latin America with 11.8%; North America with 9.6%, and Europe with 7.5%.

Current results on the cephalosporin class demonstrated K. pneumoniae (24.6%), A. baumannii (22.7%), and E. coli (18.7%) as the bacteria with the highest resistance rates. On the other hand, GNB demonstrated a 75.7% resistance to cephalosporin, whereas GPC showed 54.2% sensitivity to this class. Ali et al. ³⁰ pointed out resistance rates between 54% and 80.4% to all generations of cephalosporins. On the other hand, studies by Cek et al. ²¹ showed a resistance to cephalosporins between 35% and 50%.

Studies by Dalhoff ³¹ and Arana et al. ¹⁰ reported that quinolone resistance was reported for the first time in S. aureus, particularly MRSA and P. aeruginosa. Resistance to fluoroquinolones, such as ciprofloxacin and levofloxacin, was initially detected in GP and GN bacteria, in hospitals. Since the 1960s, fluoroquinolones have been shown to be less effective in combating urinary, respiratory, gastrointestinal, urogenital, intraabdominal and skin infections due to increased bacterial resistance by ESBL production in Enterobacteriaceae. Consequently, E. coli, frequently treated with fluoroquinolones, has been indicated as being especially relevant in urinary and gastrointestinal infections.

GNB resistance has been reported in 77.5%, with high prevalence rates in K. pneumoniae (28.2%), followed by A. baumannii (26.4%), to the quinolone class. A 74.5% general resistance of isolates has been shown in GP, with greater resistance in CoNS (48.1%) to quinolones. A study by Ali et al. ³⁰ (N = 91; 61%), showed resistance rates to ciprofloxacin (60%), sparfloxacin (59%), levofloxacin (58%), and norfloxacin (57%). In their 2003–2010 study, Cek et al. ²¹ reported lower rates of resistance to fluoroquinolones (26.6%), with high resistance rates to ciprofloxacin (>50%), while Lorenzoni et al. ¹⁸ showed increased resistance of K. pneumoniae to ciprofloxacin (68%).

Conclusion

The epidemiological profile of the study conducted in two hospitals in Pelotas, Brazil, between October 2018 and January 2019, represented statistically the total population of the municipality by the sample size calculation of 286 bacterial isolates.

The results showed that bacterial resistance was prevalent among male patients (54.5%), more than 60 years of age (52.1%), hospitalized in wards (55.0%), and urinary tract (40.6%) as the predominant site of infection.

Most GNB causing HAIs were K. pneumoniae, A. baumannii, E. coli, and P. aeruginosa, particularly resistant to the following classes of antibiotics: penicillins, quinolones, and cephalosporins. In the GPC group, S. aureus (N = 22; 7.69%). Coagulase-negative staphylococci (N = 16; 5.59%) and S. haemolyticus (N = 8; 2.80%) showed the highest resistance rates to macrolides (95.4%), lincosamides (90.3%), and penicillins (77.4%).

The urinary tract was the predominant site of infection (40.6%), and the main bacteria found in urine samples were E. coli (70.5%) in male patients (63.6%), within the age range of 20–49 years (31.8%), admitted to wards (68.2%), and K. pneumoniae (56.6%) in male patients (52.8%), over 70 years of age (35.8%), admitted to wards (58.5%). The lung ranked second as the infection site, with (30.0%) of the total samples of patients with HAIs and with prevalence of bacteria A. baumannii (66.0%) in male patients (53.2%), over 70 years of age (36.2%), admitted to ICU (51.1%), and P. aeruginosa (45.0%) in male patients (70.0%), over 70 years of age (50.0%), admitted to wards (80.0%).

The classes of penicillins, quinolones, and cephalosporins were ineffective in the treatment of GNB-caused infections, especially in the urinary tract and among patients over 70 years of age (male patients admitted to nursing homes). In the case of penicillin, lincosamide, and macrolide classes, the highest resistance rate of GPC occurred predominantly in blood infections, among males over 70 years of age and hospitalized in wards.

The study indicates that HAIs constitute a serious public health problem, suggesting the need for further research and the development of prevention and combat strategies at the local level.