“One Health” perspective on prevalence of co-existing extended-spectrum β-lactamase (ESBL)-producing Escherichia coli and Klebsiella pneumoniae: a comprehensive systematic review and meta-analysis

Background The Escherichia coli (E. coli) and Klebsiella pneumoniae (K. pneumoniae) bacterial isolates that produce extended-spectrum β-lactamases (ESBLs) contribute to global life-threatening infections. This study conducted a systematic review and meta-analysis on the global prevalence of ESBLs in co-existing E. coli and K. pneumoniae isolated from humans, animals and the environment. Methods The systematic review protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) [ID no: CRD42023394360]. This study was carried out following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. One hundred and twenty-six eligible studies published on co-existing antibiotic resistance in E. coli and K. pneumoniae between 1990 and 2022 were included. Results The pooled prevalence of ESBL-producing E. coli and K. pneumoniae was 33.0% and 32.7% for humans, 33.5% and 19.4% for animals, 56.9% and 24.2% for environment, 26.8% and 6.7% for animals/environment, respectively. Furthermore, the three types of resistance genes that encode ESBLs, namely blaSHVblaCTX−M,blaOXA, and blaTEM, were all detected in humans, animals and the environment. Conclusions The concept of “One-Health” surveillance is critical to tracking the source of antimicrobial resistance and preventing its spread. The emerging state and national surveillance systems should include bacteria containing ESBLs. A well-planned, -implemented, and -researched alternative treatment for antimicrobial drug resistance needs to be formulated. Supplementary Information The online version contains supplementary material available at 10.1186/s12941-023-00638-3.


Background
The extended-spectrum β-lactamase (ESBL) produced by Enterobacteriaceae is a common source of antimicrobial resistance (AMR) in both animals and humans [1].These ESBL-producing bacteria are often associated with multidrug resistance (MDR) against other antibiotic classes [2].They develop resistance to third-and fourthgeneration cephalosporins, as well as aztreonam, which is among the last medications available to treat infections caused by Escherichia coli and Klebsiella pneumoniae [1,[3][4][5].However, they are not resistant to carbapenems or cephamycin [4].
Escherichia coli and K. pneumoniae are two examples of ESBL-producing Enterobacteriaceae (ESBL-E) that were the first to be linked to infections in the medical context [6,7].Escherichia coli and K. pneumoniae can cause a wide range of serious illnesses, which are becoming increasingly difficult to treat due to the development of antibiotic resistance [8].Carbapenem is an antibiotic used as a last option to treat beta-lactamase Enterobacteriaceae transfer [9].The spread of E. coli and K. pneumoniae extended spectrum cephalosporinase in animal populations, the environment, and the community proves that such bacterial strain are transmitted and persist outside of clinical settings [8,10].
In the Enterobacteriaceae family of commensal or pathogenic bacteria, ESBLs are crucial antibiotic resistance determinants that are passed on through horizontal gene transfer [9].Among several ESBL gene variants, the most prevalent and clinically relevant are bla TEM , bla SHV (sulphydryl variable) and bla CTX−M , (cefotaximehydrolyzing β-lactamase) [4].There are several families of ESBLs, including the bla TEMs -type ESBLs and the bla SHVs -type ESBLs, with variants that differ only by a few amino acid substitutions [4].There are some unique characteristics in each of the ESBL families.The bla TEM−1 , bla TEM−2 and their ESBL derivatives are usually carried by transposons like Tn1, Tn2, or Tn3 [4,11].An association has been found between the spread of bla CTX−M -producing enzymes and E. coli belonging to the ST131 clonal group [12].It has been reported that ST131 isolates contain several bla CTX−M types, including bla CTX−M−15 [13].The empirical treatment of ESBLproducing Enterobacteriaceae infections has become increasingly difficult [14].
The presence of ESBL-producing E. coli and K. pneumoniae in humans, animals, and the environment poses a public health threat.In addition, it is crucial to maintain current data on ESBL-producing E. coli and K. pneumoniae in health systems to minimize the consequences of ESBL-producing bacteria.A few systematic reviews have been published on ESBL-producing E. coli in animals, humans, and the environment in Bangladesh and South America [15,16], as well as ESBL-producing E. coli and K. pneumoniae in the United States of America [17].However, there is limited information on comprehensive data available to estimate the global prevalence of co-existing ESBL-producing E. coli and K. pneumoniae in animals, humans, and the environment.The current study is a systematic review and meta-analysis aimed at providing a comprehensive prevalence of co-existing ESBL-producing E. coli and K. pneumoniae, in animals, humans, and the environment based on available data published globally.

Protocol registration
This study was registered in the International Prospective Register of Systematic Reviews (PROSPERO) with registration no: CRD42023394360.

Study design and systematic review protocol
Using the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [18], data extraction, screening, and analysis involved searching database systems for potentially pertinent articles, assessing their relevance, and evaluating whether they were acceptable for review.

Search strategy
This study used several database systems, including ScienceDirect, PubMed, Google Scholar, and Scopus to search for articles published in English from 1 to 1990 until 28 November 2022.Literature searches were conducted using keywords comprising "ESBL, extended-spectrum beta-lactamase, " "extended-spectrum β-lactamase, " "Human, " "Animal, " "Environment, " "Escherichia coli or Klebsiella pneumoniae, " and "E. coli or K. pneumoniae."The last search took place on 28 November 2022.

Study selection
Journal article titles and abstracts were reviewed for suitability for the inclusion and exclusion criteria by three authors namely TR, TM, and OT.Two reviewers independently analysed all the studies from the search by title, abstract, and selected full texts, and a third reviewer arbitrated for any divergence.Full-text journal articles published in English were included, while reviews, conference abstracts, and book chapters were excluded.Journal articles were selected for full-text review if they investigated the co-existence of ESBL-producing E. coli and K. pneumoniae only and reported the number of isolates and the positive number of ESBL isolates from animals, humans, and the environment.

Inclusion and exclusion criteria
Criteria for inclusion of studies were: (i) studies investigating the prevalence of ESBL-producing E. coli and K. pneumoniae in animals, humans, and the environment.The following exclusion criteria was used: (ii) no total number of isolates, (iii) no abstracts, reviews, experiments, and book chapters, and (iv) an unclear ESBL identification method.

Data extraction
The following data were extracted and summarized by two authors independently (TR and TM) from the final selected studies: (1) author name, (2) year of publication, (3) location, (4) year of publication, (5) type of samples, (6) ESBL identification methods, (7) positive samples of E. coli and K. pneumoniae isolates from animals, humans and the environment, (8) total number of ESBL positive E. coli and K. pneumoniae isolates, and (9) genes encoded for antibiotic resistance (ESBL).The recorded data were compiled in Microsoft Excel for further analysis.

Quality assessment of included studies
A critical appraisal tool developed by the Joanna Briggs Institute (JBI) was used to assess each study's bias risk [19].There are nine questions in this JBI instrument, which are: (1) Was the sample frame appropriate to address the target population?(2) Were study participants recruited in an appropriate way? (3) Was the sample size adequate?(4) Were the study subjects and setting described in detail?(5) Was data analysis conducted with sufficient coverage of the identified sample?(6) Were valid methods used for the identification of the condition?(7) Was the condition measured in a standard, reliable way for all participants?(8) Was there appropriate statistical analysis?(9) Was the response rate adequate, and, if not, was the low response rate managed appropriately?Each answer was scored 0 or 1 based on whether the answer was "yes" or "no".NA was used to indicate that the question was not relevant to the study.Studies with scores of 7-8 indicated a low risk of bias, scores of 5-6 indicated a moderate risk of bias, and scores less than 5 indicated a high risk of bias.Only articles with low bias risks were included in this study.

Data synthesis and data analysis
Comprehensive Meta-Analysis Software v.4.0 (https:// www.meta-analysis.com/)was used to conduct the metaanalysis of the prevalence of ESBL-producing E. coli and K. pneumoniae in humans, animals, and the environment.The pooled prevalence of E. coli and K. pneumoniae isolates was measured and subgroup analysis was performed according to continent, country, year of publication, settings, ESBL-resistant genes, and samples collected.Random-effects models were used to generate forest plots showing the study-specific effect sizes with a 95% confidence interval (CI) for the pooled prevalence (PP).The I 2 statistic was used to measure heterogeneity among studies.A value close to 0% indicates no heterogeneity, whereas a value close to 25%, 50%, and 75% corresponds to low, moderate, and high heterogeneity, respectively.The p-values correspond to the heterogeneities between studies from a Chi-squared test of the null hypothesis that there is no heterogeneity.

Publication bias
Publication bias was measured using funnel plots to test for symmetry and this was further complemented using the Beg and Mazumdar rank correlation test and Egger's regression test.For all the tests, a p-value of < 0.05 was considered statistically significant.

Search and screening results
A total of 2854 studies were retrieved after the initial search was conducted across four databases, including 688 from PubMed, 925 from ScienceDirect, 980 from Google Scholar, and 261 from Scopus (Fig. 1).The abstracts of 1161 articles were screened based on the inclusion and exclusion criteria.A total of 526 studies were initially considered eligible and were thus subjected to full-text evaluation.After full-text examination, 126 animals (n = 17) were eligible for inclusion [humans (n = 101), environment (n = 5), animals and environment (n = 3)].

Meta-analysis results on overall prevalence
Pooled prevalence estimates (PPE) of ESBL-producing E. coli and K. pneumoniae in animals, humans, and the environment, as well as a summary of the subgroup analysis, are shown in Tables 1 and 2, and 3. A total of 79,497 were confirmed as E. coli, while 37,998 were confirmed as K. pneumoniae.Only 18,923 and 8502 were confirmed as ESBL-producing E. coli and K. pneumoniae, respectively.

Prevalence of ESBL in humans
The global prevalence of co-existing ESBL-producing E. coli and K. pneumoniae in humans was investigated in 101 studies (Supplementary Table S1; Fig. 2), of which the pooled prevalence was 33.0% (0.330; 95% CI: 0.282-0.381)for E. coli and 32.7% (0.327; 95% CI: 0.286-0.371)for K. pneumoniae, ranging from 0.6 to 99.9%.The mean effect size is 0.330 with a 95% confidence interval of 0.282 to 0.381 for E. coli (Fig. 3A); the prediction interval (0.282 to 0.381) reflects the endpoints of this distribution.While the mean effect size is 0,327 with a 95% confidence interval of 0.286 to 0.371 for K. pneumoniae (Fig. 3B), the prediction interval shows that there were some populations where the effect was as low as 0.286 and others where it was as high as 0.371.According to the study year, ESBLproducing E. coli was more prevalent in the 2021-2022 year interval with a PPE of 53% (95% CI: 41.1-65.5%),while ESBL-producing K. pneumoniae were more prevalent in the 2000-2010 period with PPE of 39.7% (95% CI: 31.8-48.1).The period of year interval 1990-2000 had the lowest PPE for ESBL-producing E. coli and K. pneumoniae with 10.9% (95% CI: 04.4-24.5)and 25.4% (95% CI: 7.1-60.0),respectively.

Prevalence of ESBL in animals
The characteristics of all eligible animal studies included in this review were presented in Supplementary Table S2.Using a random-effect model, the pooled prevalence of ESBL-producing E. coli and K. pneumoniae was estimated at 33 In terms of regions, the majority of the studies have performed in Africa (n = 9) (Fig. 4).For ESBL-producing E. coli, the highest estimate of 49.6% (95% CI: 28.2-71.1)was observed from Africa, and pooled prevalence was estimated at 49.6% (95% CI: 28.2-71.1)for ESBL-producing K. pneumoniae.However, Europe and North America were not included in the meta-analysis due to their low number of studies (fewer than three studies).

Prevalence of ESBL from both animals and the environment
The E. coli and K. pneumoniae were screened from 625 isolates of both animals and the environment, 96 of which were ESBL-producing E. coli, and 33 were ESBLproducing K. pneumoniae (Supplementary Table S4).The overall PPE for ESBL-producing E. coli was 26.8% based on three studies (95% CI: 4.9-72.3).The ESBL-producing K. pneumoniae had a PPE of 6.7% (95% CI: 1.5-25.5)(Table 4).

Subgroup analysis by ESBL-resistant genes
We employed a random-effects model to analyze subgroups of ESBL genes evaluated in three or more studies.The bla SHV bla CTX−M, bla OXA and bla TEM genes were included in the meta-analysis for animals, humans, and the environment.As a result of the small number of studies, bla NDM , bla KPC , bla CMY , bla OXA , bla VEB , bla PER , bla GES , bla MOX , bla AmpC , and bla PSE were not included in the meta-analysis.

Pooled prevalence rate in humans
The overall incidence rate of ESBL-producing E. coli genes was 35

Pooled prevalence rate in the environment
The PPE for the bla CTXM gene was estimated to be 51.4% (25.8-76.2) and 22.6% (6.5-55.2) for ESBL-producing E. coli and K. pneumoniae, respectively.For studies on the

Risk of publication bias of included studies
Publication bias was measured using funnel plots to test for symmetry, and this was further complemented using the Begg and Mazumdar rank correlation test and Egger's regression test.The funnel plot was used to analyze the publication bias.Figure 6 displays a practically symmetrical distribution of all included studies on both sides of the funnel plot, indicating a relatively low potential for publication bias (p-value < 0.02595) for faecal samples from humans (ESBL-producing E. coli and K. pneumoniae).

Discussion
In this review of data published between January 1, 1990 and November 28, 2022, key findings have been made regarding the burden of co-existing E. coli and K. pneumoniae positively expressing ESBL which provide useful information for health professionals in a variety of fields.There are some big differences between animals, humans and the environment in the prevalence of ESBL positive E. coli and K. pneumoniae.We therefore chose to examine the prevalence of ESBL positive co-existing E. coli and K. pneumoniae from each source and do a subgroup analysis.This study analyzed data on ESBLproducing co-existing E. coli and K. pneumoniae isolated from human, animal, and environmental samples worldwide.Many articles were excluded from our systematic literature review because they only reported the prevalence E. coli or K. pneumoniae separately.However, the aim of this study was to include only studies that reported both co-existing E. coli and K. pneumoniae.

ESBL of E. coli and K. pneumoniae isolated from humans
Escherichia coli and K. pneumoniae species are the most common cause of infections, particularly in countries with underdeveloped healthcare systems [20,21].The ESBL-producing E. coli and K. pneumoniae strains have been identified as prominent multidrug-resistant pathogens linked to hospital-acquired infections [16,20,21].They are also known to cause several common infections in children, including gastroenteritis, urinary tract infections, septicaemia, and neonatal meningitis [16,22].It is also known that K. pneumoniae is one of the most common bacteria that causes of opportunistic healthcare-associated infections, which are made worse by its ability to produce ESBL enzymes [23,24].
The PPE based on 17,513 ESBL positive E. coli and 8165 ESBL positive K. pneumoniae was 33% and 32.7% respectively, from 101 studies in humans.This is comparatively higher than similarly reported PPE from a review conducted in Bangladesh, where PPE of ESBL-producing E. coli was 17% [16], and in South America, where PPE was 2.2% [15].The study conducted in Nepal on ESBL-producing K. pneumoniae had a lower PPE of 5% for ESBLproducing K. pneumoniae [24].The variation could be due to the result of a difference in the time of the study, also be due to differences in geographical properties, sample categories, and types, as well as sample size and identification methods.3 Plot of true effects: A) E. coli, the mean effect size is 0,330, the mean score is 0.82, and B) K. pneumoniae, the mean effect size is 0,327, and the mean score is 0.76 Furthermore, this study found that E. coli had a high PPE of 52.7% on stool samples from 243 positive ESBLproducing E. coli, while K. pneumoniae had a high PPE of 37.7% on urine samples.This is not surprising, as K. pneumoniae is one of the bacteria that cause urinary tract infections [25,26].However, most of the stool samples containing the stx-positive gene of E. coli were typical diarrheal samples [27].

ESBL of E. coli and K. pneumoniae isolated from animals
In this study, 872 ESBL-producing E. coli isolates from 17 studies and 324 ESBL-producing K. pneumoniae from 16 studies were subjected to meta-analysis.The E. coli and K. pneumoniae PPE from this study are 33.5% and 19.4%, respectively.Our results are comparable to a previous systematic review study by Islam et al. [16] that reported prevalence of 22% for for ESBL-producing E. coli in animals.However, PEE in our study was higher than the one obtained in India, where PPE of ESBL-producing E. coli and K. pneumoniae was 9% and 10%, respectively [28].The variation could be the result of a difference in the time of the study.The other reason could be the methodology used to detect the number of published articles included in the current study.
The subgroup analysis at continental level showed that the PPE was higher for ESBL-producing E. coli in Africa and Asia, with 49.6% and 44.5%, respectively.Our findings were consistent with a previous study that reported ESBL in E. coli isolated from environments in Bangladesh at 39% [16].The PPE from chicken samples was 48.0% for E. coli and 43.5% for K. pneumoniae.Other animals such as cattle, goats, birds, dogs, cats, camels, fish, and horses were not included in the meta-analysis due to the low number of samples.Country-specific findings indicated that Nigeria had a PPE of 33.4% and 11.9% for E. coli and K. pneumoniae, respectively.Other countries, including Sudan [29], India [9], Algeria [30], Ghana [31], Turkey [32], Japan [33], Canada [34], Finland [35], Taiwan [36], South Africa [37], Tanzania [38], Tunisia [39], India [40], and the Netherland [41] were not included in the metaanalysis due to the low number of studies.

ESBL of E. coli and K. pneumoniae isolated from the environment
This systematic review and meta-analysis showed that ESBL-producing E. coli and K. pneumoniae were isolated from various types of environmental samples, such as water, manure, hospital sewage, and raw vegetables.The PPE of ESBL-producing E. coli was 56.9% (31.6-79.0)while the PPE of ESBL-producing K. pneumoniae was 24.2%.This could be due to the possibility of traces of previously consumed antibiotics and antimicrobial-resistant bacteria in faecal and urine waste.This waste and wastewater can contribute to the spread of antimicrobialresistant bacteria in the environment possibly having a negative impact on human health.Many of these species have genes for antibiotic resistance, which are eventually incorporated into genetic mobile platforms that can spread among bacterial populations in soil and water.
The PPE of ESBL-producing E. coli from environments was 39%, which is comparable to a previous study conducted in Bangladesh that reported a pooled prevalence of 39% for ESBL-producing E. coli from environments [16].According to the literature, there is no meta-analysis on ESBL-producing K. pneumoniae in the environment.Most of the studies were conducted from 2011 to 2020 with a PPE of 75.6% for ESBL-producing E. coli and 11% for ESBL-producing K. pneumoniae.Data recorded from the included articles showed that the studies were from only five countries, namely Nigeria, India, Tunisia, Nepal, and Thailand.

ESBL genes in E. coli and K. pneumoniae isolated from animals, humans, and the environment
In recent years, there has been an increase in ESBL-producing E. coli and K. pneumoniae, mostly identified using phenotypic methods [42,43].We report the first global PPE of co-existing ESBL-producing E. coli and K. pneumoniae.A number of studies have investigated bla TEM , bla SHV , bla OXA , and bla CTX−M genes, with bla CTX−M being the type most commonly associated with ESBL [43][44][45].These genes reside on plasmids, which can be transferred horizontally to other bacteria [46].
Among the bla-genes screened in this study, bla CTX−M and bla SHV were the most detected from ESBL-producing E. coli and K. pneumoniae from animals, humans, and the environment.The bla TEM have been reported from vegetables in Finland [47] and in southern Thailand [48].In this review, bla CTX−M was the gene most detected in animals, humans, and the environment.It has also been reported that K. pneumoniae and E. coli from dogs and cats possess bla CTX−M type genes [49,50].Lately, bla CTX−M enzymes are the most common ESBL type because they have environmental origins [51].The bla CTX−M enzyme is divided into five subgroups based on its amino acid composition: bla CTX−M−1 , bla CTX−M−2 , bla CTX−M−8 , bla CTX−M−9 , and bla CTX−M−25 [46].Globally, carbapenem-resistant Enterobacteriaceae are increasing due to the growing usage of carbapenems to treat ESBLproducing infections [52].
In this systematic review and meta-analysis, we also report the presence of bla KPC in humans.These findings warrant further investigation of the presence of carbapenem-resistant genes since such resistant E. coli and K. pneumoniae are among top priority pathogens on the list of World Health Organization [WHO] for the development of antimicrobials [53].

"One Health" perspective
"One-Health" concept refers to the collaborative effort between local, national, and global disciplines for the purpose of attaining optimal health for humans, animals, and the environment [54][55][56].Antimicrobial use practices in companion animals are similar to those in humans; that is, drugs are primarily administered for treating clinical infections, such as post-surgery infections, with some being used as prophylaxis- [57].
A major obstacle to addressing antimicrobial resistance has been the blame game between human medicine and the agriculture sectors regarding who is responsible for the increase in antimicrobial resistance due qof bacteria causing zoonotic diseases [58].This study confirmed the co-existence of ESBL-producing E. coli and K. pneumoniae prevalence in animals, humans, and the environment (water, manure, hospitals sewage, and raw vegetables).Our results highlight the significance of "One Health, " as ESBL-producing E. coli and K. pneumoniae were detected in humans, animals, and the environment.Moreover, the bla CTX−M , and bla SHV were detected in ESBL-producing E. coli and K. pneumoniae isolated from animals, humans, and the environment, hence the "One Health" concern.Future researchers should be encouraged to use the "One Health" approach to develop methodologies that explicitly examine interlinks between human-animalenvironment frameworks, with a particular emphasis on antibiotic resistance in zoonotic diseases.

Limitations
To assess ESBL-producing E. coli and K. pneumoniae levels globally, this study relied on peer-reviewed publications.Reviews, theses, and preprint articles that potentially include essential information on ESBL-producing E. coli and K. pneumoniae were excluded from our review as per PRISMA guidelines for writing systematic reviews and meta-analysis.There may have been articles published in other languages that were missed since the search strategy was limited to articles published in English.The pooled prevalence of some countries and continents was not calculated because there are few published studies.Some antimicrobial resistance genes were also not included in this meta-analysis due to the small number of studies.In a meta-analysis, publication bias may be evaluated using funnel plots, regressions, and Begg's adjusted rank correlation test [59].However, meta-analyses that include fewer than 10 studies or high amounts of heterogeneity between studies can lead to misleading results from these evaluation tools.In the presence of a high level of heterogeneity, it is very difficult to evaluate the true results of statistically significant publication bias tests.Since there is high heterogeneity across all analyses, readers should be cautious when interpreting pooled analyses and subgroups.

Conclusion
This study is the first meta-analysis that estimates the prevalence of co-existing ESBL-producing E. coli and K. pneumoniae in animals, humans, and the environment worldwide.We further observed that bla CTX−M , and bla SHV were the most frequently detected genes from ESBL-producing E. coli and K. pneumoniae infecting animals, humans, and the environment.This study presents robust and valuable data that can serve as a useful reference for doctors, veterinarians, and environmental scientists, as it informs them about the prevalence of co-existing ESBL-producing E. coli and K. pneumoniae recovered from humans, animals, and the environment."One-health" surveillance is vital for source tracking and mitigating the spread of antimicrobial-resistant bacteria.ESBL-containing bacteria should be included in emerging state and national surveillance systems.Globally, alternative treatments for antimicrobial drug resistance should be well researched, planned, and implemented.

Fig. 1 A
Fig. 1 A PRISMA flow diagram illustrating the process of selecting studies

Fig. 2
Fig. 2 World map showing number of articles from different countries that reported the ESBL-producing E. coli and K. pneumoniae in humans (https:// www.mapchart.net/world.html(accessed on 10 March 2023)

Fig.
Fig.3Plot of true effects: A) E. coli, the mean effect size is 0,330, the mean score is 0.82, and B) K. pneumoniae, the mean effect size is 0,327, and the mean score is 0.76

Fig. 4
Fig. 4 World map showing number of articles from different countries that reported the ESBL-producing E. coli and K. pneumoniae in animals (https:// www.mapchart.net/world.html(accessed on 10 March 2023)

Fig. 5
Fig. 5 Forest plot showing the pooled estimates of (a) ESBL-producing E. coli and (b) ESBL-producing K. pneumoniae from animals' samples.The squares demonstrate the individual point estimate.The diamond at the base indicates the pooled estimates from the overall studies

Fig. 6
Fig. 6 Funnel plot showing the evidence of publication bias of the studies conducted on faecal samples from humans (ESBL-producing and K. pneumoniae)

Table 2
Subgroup analysis of ESBL-producing Escherichia coli and Klebsiella pneumoniae in animals animals consisted of the bla SHV ,bla CTXM ,bla OXA , and bla TEM genes.In total, four genes or gene groups were analyzed.The PPE was observed on ESBL-producing E. coli for the bla SHV ,bla CTXM ,bla OXA and bla TEM at 29.

Table 4
Subgroup analysis of ESBL-producing Escherichia coli and Klebsiella pneumoniae in animal and the environment