Trend of distribution and antimicrobial resistance in uropathogens in China from the CHINET antimicrobial resistance surveillance program, a 7-year retrospective study

Urinary tract infections (UTIs) are common urological diseases that easily relapse and have led to an increasing economic and health burdens. The China Antimicrobial Surveillance Network (CHINET) system is one of the most influential antimicrobial resistance surveillance networks in China. This study analyzed antimicrobial resistance and distribution trends of uropathogens from 2015 to 2021 using the CHINET system. A total of 261,893 non-duplicate strains were collected; Gram-positive bacteria accounted for 23.8% while Gram-negative bacteria accounted for 76.2%. Escherichia coli , Enterococcus faecium , Klebsiella pneumoniae


Introduction
Urinary tract infections (UTIs) are among the most common bacterial infections.Approximately 150 million people worldwide experience UTIs every year, causing severe health problems and socioeconomic burdens [1].Regular monitoring of the antimicrobial resistance in bacteria isolated from urine specimens is necessary to guide the rational use of antibiotics and reduce bacterial resistance.The China Antimicrobial Surveillance Network (CHINET) was founded in 2004; a total of 21 hospitals (including 19 general hospitals and 2 specialized children's hospitals) participated in the network as of 2015.As of 2021, there are 51 hospitals (including 45 general hospitals and 6 specialized children's hospitals) involved from 29 provinces, municipalities, and autonomous regions across the country.In this study, we analyzed the antimicrobial resistance of bacteria isolated from urine using data from CHINET system from 2015 to 2021.Bacterial species and antimicrobial resistance over seven years were summarized to guide antimicrobial therapy.
There were significant differences in the dominant bacteria among populations of various ages and genders.The proportions of E. coli, E. faecium, K. pneumoniae, P. mirabilis, and S. agalactis isolated from female patients were higher than those isolated from male patients, while the proportions of E. faecalis, P. aeruginosa, E. cloacae, A. baumannii, and S. aureus isolated from female patients were lower.Almost half of the isolates were isolated from people aged ≥ 60 years.Among 6910 strains of S. agalactiae, only 2.8% were isolated from individuals ≤ 18 years, which was significantly different from the distribution trend of other bacteria (Fig. 1).
The majority of the isolates were from non-ICU inpatients (74.6%), 18.4% were from outpatients, 2.4% were from emergency patients, and 4.6% were from ICU Compared with E. faecalis, E. faecium showed higher resistance rates to ampicillin, levofloxacin, high concentrations of gentamicin (GEH), and nitrofurantoin (P < 0.05) (Fig. 2).Less than 2.2% of E. faecalis and E. faecium were resistant to vancomycin, linezolid, and teicoplanin in our study.

β-hemolytic Streptococcus
A total of 7412 strains of β-hemolytic Streptococcus were isolated, 93.2% of which were S. agalactiae.All S. agalactiae strains were sensitive to penicillin, ceftriaxone, vancomycin and linezolid.The resistance rates to erythromycin, clindamycin, and levofloxacin were more than 50%.

Antimicrobial resistance in Gram-negative bacilli Enterobacteriaceae
A total of 178,979 strains of Enterobacteriaceae were isolated, including 122,384 strains of E. coli (68.4%), 30,316 strains of Klebsiella spp.(16.9%), and 10,271 strains of Proteus spp.(5.7%).The resistance rates of E. coli to imipenem, piperacillin-tazobactam and tigecycline remained lower than 4% while resistance rates to cefazolin, cefuroxime, and ciprofloxacin were higher (Fig. 3).The resistance rates of E. coli to cefazolin, cefuroxime, cefepime and gentamicin decreased gradually over seven years.Similarly, K. pneumoniae showed relatively high resistance rates to cefazolin, cefuroxime and ciprofloxacin, with resistance rates of 40%-60%.The average resistance frequencies of K. pneumoniae to tigecycline and polymyxin B were 6.9% (1027/14,890) and 3.2% (118/3708), respectively.In the early years of 2015-2021, susceptibility test for polymyxin B were conducted only when carbapenem resistance was identified in some hospitals; because of this, the results of polymyxin B are present for informational purposes only.

Antimicrobial resistance within different populations
As shown in Fig. 4, the resistance rates to ampicillin and GEH in E. faecalis and E. faecium from ICU inpatients were significantly higher than in those from outpatients and other inpatients (P < 0.05).
The resistance rates to most β-lactam antimicrobials, aminoglycosides, and fluoroquinolones in E. coli and Klebsiella spp.from ICU inpatients were significantly higher than those from outpatients and other inpatients (P < 0.05).The resistance rates to most β-lactam antimicrobials in E. coli and Klebsiella spp.from pediatric outpatients and inpatients aged < 18 years were higher than those from inpatients aged ≥ 18 years, while the opposite was observed for resistance to aminoglycosides and fluoroquinolones.
Except for tigecycline and polymyxin, the resistance rates to other antimicrobial agents in P. aeruginosa and Acinetobacter spp.from ICU inpatients were significantly higher than those from outpatients and non-ICU inpatients (P < 0.05).Among non-ICU inpatients, the resistance rates to most antimicrobials in isolated from medical inpatients aged ≥ 18 years were higher than in those from surgical inpatients aged ≥ 18 years and  -4.

Multidrug resistance (MDR) and special antimicrobial resistance
MDR was defined as resistantance to at least one antimicrobial agents from three or more different antimicrobial classes.The prevalence of MDR in E. coli, K. pneumoniae, E. faecium and E. faecalis was 72.0% (88,105/122,384), 62.1% (15,929/25,642), 77.0% (21,063/27,343), and 23.4% (5312/22,741), respectively.The prevalence of MDR E. coli and K. pneumoniae in male patients was higher than that in female patients while the prevalence of MDR E. faecalis in male patients was lower than that in female patients (P < 0.05).The prevalence of MDR was significantly higher in isolates from ICU inpatients than in non-ICU inpatients, emergency patients, and outpatients (P < 0.05).Further details are available in the Supplementary Table 5.
Compared with female patients, the prevalence of CRECO, CRKPN and CRABA in male patients was significantly higher; the prevalence of CRECO and CRKPN was nearly two times greater in man than in women.The prevalence of CRECO, CRKPN, CRPAE and CRABA in ICU inpatients was higher than other patients (Supplementary Table 6).

Discussion
This is the first summary of antimicrobial resistance trends in uropathogens from CHINET system.The surveillance results showed that 76.2% of the bacteria isolated from urinary tract specimens were Gram-negative bacteria; E. coli, E. faecalis, E. faecium and K. pneumoniae were predominantly identified isolates.The dominant bacteria differed among populations of various ages and genders.The resistance patterns varied among different bacteria, and there were also differences in their changes over time.
Most of the strains were isolated from hospitalized patients.Uropathogens were more common in females and half were isolated from people > 60 years of age.E. coli, E. faecium, K. pneumoniae, P. mirabilis and S. agalactis were more frequently isolated from urine specimens from female patients, while E. faecalis, P. aeruginosa, E. cloacae, A. baumannii and S. aureus were more common in male patients.The proximity of the urethral orifice to the anus and short urethra in females may be the reason for a greater susceptibility to infection by intestinal bacteria [2,3].P. aeruginosa and A. baumannii, as common pathogens causing hospital-acquired infections [4,5], were more common in male patients.It is worth further analysis to determine whether UTIs are hospital acquired or community acquired.
The most common Gram-negative bacilli isolated from urine samples were E. coli and K. pneumoniae, accounting for 46.7% and 9.8% of isolates, respectively, with high resistance rates to cefazolin, cefuroxime, and ciprofloxacin.The resistance rates to tigecycline, amikacin and imipenem were relatively low, which was consistent with the CHINET antimicrobial resistance monitoring results from other types of samples [6].Fluoroquinolones cannot be easily catabolized in vivo and are mainly excreted in urine in a prototype form, so they are often used as empirical drugs for UTIs.Monitoring showed that the resistance rates to fluoroquinolones in E. coli and K. pneumoniae ranged from 50% to 70%.This is lower than that reported in the literature [7], where it was found that the resistance rate of E. coli from the urine of diabetes patients in Somalia to ciprofloxacin was 77.8%.Cai et al. reported that the resistance rates to ciprofloxacin were 29.9% and 52.4% in E. coli and Klebsiella spp.isolated from female outpatients affected by uncomplicated cystitis at three hospitals in Italy [8].This indicates that the resistance rate varies among different populations in different regions.We found that the resistance rate to fluoroquinolones was lower in patients ≤ 18 years of age.In the analysis of the etiology of UTIs and antibiotic resistance of isolated strains at a pediatric hospital in Warsaw, the resistance rates of E. coli and Klebsiella spp. to ciprofloxacin were 11.9% and 31.1%,respectively.This suggests that the individualized treatment for UTIs should depend on host factors, severity of illness, and risk for multidrug resistance [3].
The prevalence of nosocomial infections caused by MDR pathogens is increasing and has attracted widespread attention, with complex treatment and high mortality rates [9].In this 7-year retrospective study, the prevalence of MDR in E. coli, K. pneumoniae, E. faecium, and E. faecalis was 72.0%, 62.1%, 77.0%, and 23.4%, respectively, higher than that reported in the Jiaxing Region [10].Although in this study, the prevalence of MRSA constantly decreased from 40.6% in 2015 to 22.9% in 2021, it was still relatively high compared with that in some European countries [11,12].MRSA bacteria often carries both aminoglycoside and quinolone resistantance genes and exhibits different degrees of resistance to β-lactams, aminoglycosides, quinolones, and other antibiotics [13].Based on this study, in terms of medication selection for UTIs caused by MRSA, rifampicin, vancomycin or linezolid can be considered for UTIs caused by MRSA.The widespread use of carbapenems has led to a notable increase in the resistance rate, and carbapenem-resistance organisms (CROs) have emerged as a critical priority antibiotic-resistant pathogens worldwide [14,15].In this study we found a continuous increase in the prevalence of CRABA that was higher in male patients than in females; it even exceeded 80% in ICU inpatients.Clinicians should rationally select antibacterial drugs based on the results of bacterial culture and drug susceptibility testing, and closely monitor the treatment of CRO infection.
Carbapenem-resistant Enterobacteriaceae (CRE) have increased gradually and caused widespread concern and attention [16].In this study, the prevalence of CRKPN isolated from urine was 18.5%, slowly increasing from 2015 to 2021; this trends was slightly lower than that from all specimens monitored in CHINET during the same period [17].Tigecycline and polymyxin are relatively effective antimicrobials in in vitro drug sensitivity assays of CRE strains, but as their concentrations are low in urine and cannot reach effective therapeutic concentrations, they are not recommended for the treatment of UTIs caused by CRE.The Infectious Diseases Society of America [18] recommends ciprofloxacin, levofloxacin, trimethoprim-sulfamethoxazole, nitrofurantoin, or a single dose of an aminoglycoside are preferred treatment options; ceftazidime-avibactam, meropenem-vaborbactam, imipenem-cilastatin-relebactam, or cefiderocol are alternative options for UTIs caused by CRE.
The predominant Gram-positive cocci isolated from urine samples were E. faecalis and E. faecium (10.4% and 8.7%, respectively).According to monitoring results, E. faecalis was more prevalent than E. faecium in outpatients, E. faecium was more prevalent than E. faecalis in inpatients, and E. faecium accounted for nearly 80% of Gram-positive isolates in ICU inpatients.This is was consistent with the reports on antibiotic resistance patterns in uropathogens from selected hospitals in China and Poland [19,20].The resistance rates of E. faecalis to ampicillin and nitrofurantoin were less than 10%, while those of E. faecium to ampicillin and levofloxacin exceeded 90%.Empiric antibiotic therapy should be used with different priorities for UTIs caused by Enterococcus in different populations.Enterococcus is sensitive to vancomycin, linezolid, and teicoplanin, which are the most effective antimicrobial drugs against UTIs caused by this genera; however, some Enterococcus have developed resistance to vancomycin, linezolid, and teicoplanin in recent years, complicating the treatment of Enterococcus infections.
In this study, the prevalences of vancomycin-resistant E. faecalis and E. faecium were 0.1% and 1.3%, respectively, while the prevalences of linezolid-resistant E. faecalis and E. faecium were 2.2% and 0.3%, respectively.The prevalence of VRE was lower than that in Europe and America [21][22][23], however, it should still be monitored with greater vigilance.
The resistance rates of Enterobacteriaceae and other bacteria from inpatients to most antimicrobial drugs were higher than those of isolates from outpatients and emergency patients.The resistance rates to most antimicrobial agents in ICU inpatients were significantly higher than those in non-ICU inpatients, outpatients and emergency patients; this may be because hospitalized patients are more likely to have a history of antibacterial drug use, which has a screening effect on antibiotic-resistant bacteria.ICU patients usually have a longer length of hospitalization and a higher risk of exposure to immunosuppressants, chemotherapy drugs and antibiotics [24,25].The incidence of multidrug-resistant bacterial infections in ICU inpatients is rising [26], making it essential to implement antimicrobial stewardship interventions in the ICU setting.
In summary, we analyzed the distribution trend and antimicrobial resistance of uropathogens in China through a 7-year retrospective study using the CHI-NET antimicrobial resistance surveillance program.Data were collected from 51 hospitals in different provinces and cities in China from 2015 to 2021, making the results extensively representative, though there were some limitations in our analysis.Firstly, this study was retrospective, meaning selection bias and incomplete data (such as that for polymyxin B resistance) may have been present.Secondly, we did not distinguish the differences in bacterial species and drug sensitivity among different urine collection methods such as catheter urine, and midstream urine, cystocentesis urine, etc. Lastly, we did not obtain more detailed information to associate the severity and duration of UTIs with bacterial species and drug resistance to provide more reference value for clinical treatment.Despite this, the findings provide a valuable reference for empiric treatment of UTIs.

Conclusions
E. coli, Enterococcus, and K. pneumoniae were the most common pathogens causing UTIs in China.The bacterial species and antimicrobial resistance in different patient populations are different.Strengthening of monitoring and surveillance of bacterial resistance in urine samples is needed to minimize the further spread and evolution of multidrug-resistant superbugs such as CRE and VRE.

Bacterial strains and species identification
All aerobic bacteria isolated from urine specimens of outpatients and inpatients were obtained from the CHI-NET surveillance system for seven years between 2015 and 2021.Duplicate strains from the same patient were eliminated.Species identification was performed by a commercialized automated system at each participating site and confirmed by matrix-assisted laser desorption ionization-time of flight mass spectrometry at the central laboratory.The quality control strains were S. aureus ATCC25923, E. coli ATCC25922, P. aeruginosa ATCC27853, and E. faecalis ATCC29212.

Antimicrobial susceptibility testing
According to a CHINET uniform protocol, the broth microdilution method was used for minimum inhibitory concentration (MIC) testing by an automated antimicrobial susceptibility system, and the Kirby-Bauer disk diffusion method was used to supplement some antimicrobial agents that were not available in automated systems.Throughout the seven-year sampling period, the methodology remained unchanged in every participating hospital.
All results were interpreted by using breakpoints for susceptibility and resistance according to the Clinical and Laboratory Standards Institute (CLSI) 2022 M100-32 guidelines [27], except for tigecycline and polymyxin B. The MICs of tigecycline were interpreted following the breakpoint established by US Food and Drug Administration [28], while the The MICs of polymyxin B were interpreted following the European Committee on Antimicrobial Susceptibility Testing (EUCAST) MIC interpretive breakpoints for colistin [29].

Fig. 1 Fig. 2
Fig. 1 Distribution of the most common isolates across different ages and genders.Gender distribution is shown as a line chart and age distribution is shown as a stacked bar chart

Fig. 3
Fig. 3 Resistance profiles of E. coli, K. pneumoniae, P. aeruginosa, and A. baumannii isolated from urine towards common antimicrobial agents over seven years. a Resistance rates of E. coli for antimicrobial agents; b Resistance rates of K. pneumoniae for antimicrobial agents; c Resistance rates of P. aeruginosa for antimicrobial agents; d Resistance rates of A. baumannii for antimicrobial agents

Fig. 4
Fig. 4 Comparison of antibiotic resistance of E. faecalis, E. faecium, E. coli, Klebsiella spp., P. aeruginosa, and Acinetobacter spp. between the different populations.a Antibiotic resistance of E. faecalis between different populations; b Antibiotic resistance of E. faecium between different populations; c Antibiotic resistance of E. coli between different populations; d Antibiotic resistance of Klebsiella spp. between different populations; e Antibiotic resistance of P. aeruginosa between different populations; f Antibiotic resistance of Acinetobacter spp. between different populations

Table 1
Distribution of 261,893 bacteria in urine specimens over 7 years 10.
patients.The dominant strains varied among patients from different departments.As shown in Supplementary Table1, the top five bacteria isolated from outpatients were E. coli, K. pneumoniae, E. faecalis, P. mirabilis, and S. agalactiae; the top five bacteria isolated from emergency patients and non-ICU inpatients were E. coli, E. faecium, K. pneumoniae, E. faecalis, and P. aeruginosa, while they were E. faecium, E. coli, K. pneumoniae, E. faecalis, and P. aeruginosa in ICU inpatients.E. faecium was the most common in ICU inpatients, but it ranked 6th in outpatients.