Antimicrobial resistance prevalence in clinical and aquatic environmental ESKAPE: a systematic review with meta-analysis
1Department of Medical Education Development. Universidade Professor Edson Antônio Velano (UNIFENAS) Medical School. Belo Horizonte (Brazil)
2Department of Radiology. University of Wisconsin-Madison. Madison (EUA)
3Departamento de Engenharia Sanitária e Ambiental. Universidade Federal de Minas Gerais (UFMG). Belo Horizonte (Brazil)
4Pathos Ltda. Belo Horizonte (Brazil)
5Instituto René Rachou (IRR). Belo Horizonte (Brazil)
6Hospital das Forças Armadas. Faculdade de Medicina do Centro Universitário (Uniceplac). Brasília (Brazil)
7Departamento de Genética, Ecologia e Evolução. Universidade Federal de Minas Gerais (UFMG). Belo Horizonte (Brazil)
8Departamento de Engenharias e Computação. Universidade Estadual de Santa Cruz (UESC). Ilhéus (Brazil)
9Departamento de Engenharia Hidráulica e Ambiental. Universidade Federal do Ceará (UFC). Fortaleza (Brazil)
10Infection Control Ltda. Belo Horizonte (Brazil)
*Corresponding author: A. B. M. Vaz: Department of Medical Education Development. Universidade Professor Edson Antônio Velano (UNIFENAS) Medical School. Belo Horizonte (Brazil). Full postal address: Líbano street, 66, zip code 31.710-030, Belo Horizonte, Brazil. Telephone: +55 31 99756 4868 aline.vaz@unifenas.brAbstract
Antimicrobial resistance (AMR) in ESKAPE pathogens represents a major global health threat. Although these organisms are well established as causes of healthcare-associated infections, aquatic environments may function as reservoirs and transmission pathways for resistance. This systematic review aimed to estimate the prevalence of AMR in ESKAPE pathogens isolated from water and wastewater and to compare resistance patterns with those observed in human clinical isolates. The review followed PRISMA guidelines and was registered in PROSPERO (CRD420251020930). PubMed, Embase, and the Cochrane Library were searched to January 14, 2025. Eligible studies were original research reporting antimicrobial susceptibility data for ESKAPE pathogens isolated from both aquatic environmental matrices and clinical samples. Pooled resistance prevalence was estimated using generalized linear mixed models, with heterogeneity assessed using τ² and I² statistics and small-study effects evaluated by funnel plots and Egger’s test. Of 304 records identified, 18 studies met the inclusion criteria. The pooled overall resistance prevalence was 0.46 (95% CI: 0.36–0.57), with heterogeneity (I² = 98.8%). Resistance was higher in clinical isolates (0.67; 95% CI: 0.55–0.77) than in environmental isolates (0.24; 95% CI: 0.14–0.39), and environmental resistance was greater in effluent-impacted waters than in non-effluent sources. Interpretation is limited by methodological heterogeneity, selective isolation approaches in environmental studies, and imprecision due to small and unevenly distributed samples. Overall, AMR in ESKAPE pathogens remains more prevalent in clinical settings, but aquatic environments, particularly wastewater, represent resistance reservoirs, underscoring the need for standardized methodologies within a One Health framework.
Systematic review registration
https://www.crd.york.ac.uk/PROSPERO/view/CRD420251020930, CRD420251020930
Highlights
Antimicrobial resistance was higher in clinical isolates than in aquatic isolates.
Resistance patterns showed extreme heterogeneity across studies.
Effluent-impacted waters showed higher resistance than non-effluent sources.
Higher environmental resistance in some classes reflected methodological artifacts.
Article notes
Competing Interest Statement
The funding for this research was provided through reparation funds originating from Vale mining company.
Funding Statement
The funding for this research was provided through reparation funds originating from Vale mining company.
1.Introduction
Antibiotics are bioactive compounds capable of inhibiting bacterial growth or causing cell death through mechanisms such as disruption of cell wall or membrane integrity, inhibition of protein or nucleic acid synthesis, and selective toxicity toward microbial targets [1]. However, the widespread and often inappropriate use of these agents has intensified selective pressure on microorganisms, enabling them to develop antimicrobial resistance (AMR), defined as the capacity to survive and proliferate despite exposure to drugs that were previously effective [2]. Today, AMR poses a clear and present danger to global public health, threatening to roll back a century of medical advancements [3].
The molecular strategies underlying AMR are both varied and highly effective. Bacteria employ defenses such as producing enzymes to neutralize antibiotics, altering cell walls to limit drug entry, modifying molecular targets or developing alternative metabolic routes [1,2]. Through these strategies, resistant pathogens are able to persist and circulate not only within hospitals but across communities, making infectious diseases increasingly difficult to manage.
Resistance often begins with a spontaneous genetic mutation, which may become dominant within pathogen populations after the introduction of new antimicrobial agents, although this timeline varies across species and drugs [4]. Evolutionary pressure for resistance is amplified in environments where microbes are exposed to low, non-lethal concentrations of antibiotics, a common scenario in agricultural areas, wastewater treatment plants, and natural aquatic systems [3,5]. In clinic settings, this process is further accelerated by inappropriate prescribing, incomplete treatment courses, and the circulation of substandard medicines [2,3]. This interaction among human, animal, and environmental determinants highlights the necessity of a One Health approach to address antimicrobial resistance effectively [6,7].
Recognizing this growing threat, the World Health Organization (WHO) established its Bacterial Priority Pathogens List (BPPL), which ranks pathogens according to their risk to human health [8]. Among these, the ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp.) are notable for causing severe healthcare-associated infections and for their capacity to acquire and disseminate resistance genes [9–11]. These organisms pose a particular threat to vulnerable patient populations.
Although ESKAPE pathogens have long been recognized as major clinical challenges, there is increasing evidence that they are widely distributed beyond hospital settings, including in soil [12], food products [13], and aquatic environments such as rivers [14], drinking water sources [15] and sewage systems [13,16]. Aquatic environments, in particular, act as important reservoirs and transmission pathways for AMR, as they continuously receive inputs from wastewater effluent, agricultural runoff, agriculture runoff, and animal waste, facilitating the persistence and spread of resistant bacteria [13,17].
Because human, animal and environmental health are intrinsically interconnected, a One Health framework is essential for a comprehensive understanding of AMR. Within this context, identifying environmental reservoirs and comparing resistance patterns across ecological compartments are critical for understanding transmission pathways [18,19].
Accordingly, this systematic review and meta-analysis aimed to evaluate the prevalence of AMR in ESKAPE pathogens from water and wastewater and to compare these estimates with resistance patterns observed in clinical isolates.
2.Methods
This systematic review under the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [20,21]. A prospective protocol was formulated and uploaded to PROSPERO (CRD420251020930).
2.1Search strategy and study selection
This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A comprehensive literature search was performed in PubMed, Embase, and the Cochrane Library to identify all relevant studies published up to January 14, 2025, selected for their broad coverage of biochemical and environmental health literature.
The search strategy combined keywords and subject headings related to antimicrobial resistance, ESKAPE pathogens, and aquatic environments (Supplementary Table 1). No language or publication date restrictions were applied. All records were imported into Zotero for bibliographic management, and duplicates were removed.
Reference lists of included articles were manually screened to identify additional relevant studies. Study selection was conducted in two phases: initial screening of titles and abstracts by four independent reviewers, followed by full-text assessment by two reviewers.
Disagreements were resolved through discussion and consensus. The complete selection process and reasons for exclusion are presented in the PRISMA flow diagram (Figure 1).
2.2Eligibility criteria
Included articles were original research published in English reporting AMR in one or more ESKAPE pathogens, with culture-based resistance data from isolates obtained from both environmental matrices (such as water or wastewater) and human clinical samples. Studies were excluded if they were not original research (e.g., reviews, commentaries, editorials, protocols, or case reports without a comparative component), were not focused on ESKAPE pathogens, lacked isolates from either environmental or clinical sources, or relied exclusively on culture-independent approaches without corresponding antimicrobial susceptibility data. Articles not available in full text or lacking a persistent identifier (e.g., DOI) were also excluded. Eligibility criteria were independently applied by two reviewers, with disagreements resolved through discussion and consensus.
2.3Data Sources and Extraction
Data from all eligible studies were systematically extracted by two independent reviewers using a standardized data collection form capturing bibliographic details (first author and year), pathogen, geographic location (country and continent), sample types (environmental and clinical), number of resistant isolates and antibiotics.
For each pathogen, the number of isolates recovered, and the number identified as resistant to specific antimicrobial agents were extracted. Discrepancies between reviewers were resolved through discussion and consensus.
2.4Data analysis
The primary outcome of this meta-analysis was the prevalence of AMR, defined as the proportion of resistant isolates for each specific pathogen-antibiotic combination within each study. For each included study, we extracted the total number of isolates tested, the number of resistant isolates, the specific antibiotic, and the sample source (environmental or clinical). Each unique study–microorganism–antibiotic–source was treated as an independent effect size (Supplementary table 2).
Pooled resistance prevalence was estimated using binomial–normal generalized linear mixed models (GLMMs) with a logit link function [22], accounting for within- and between-study variability. Pooled logit-transformed estimates and corresponding 95% confidence intervals (CIs) were back-transformed to proportions. Between-study heterogeneity was assessed using Cochran’s Q test [23], τ² (maximum likelihood ratio test), and the I² and H² statistics [23]. Homogeneity (τ² = 0) was evaluated using Wald and likelihood-ratio tests. Analyses were conducted for the full dataset and stratified by clinical and environmental sources. Forest plots were generated by antibiotic class, displaying pooled prevalence estimates for clinical and environmental samples, with antibiotics ordered by overall mean resistance. Small-study effects were assessed using contour-enhanced funnel plots of logit-transformed proportions and Egger’s regression test, applied when at least five independent studies were available. All analyses were performed in R (version 4.2.2) [24] using the metafor [25] and meta [26] packages.
3.Results
3.1Search and screening results
From this subset, 18 studies met the inclusion criteria and were included in the review. The main reason for exclusion during full-text screening were the absence of phenotypic AMR data, exclusive use of genomic or metagenomic approaches, pre-selection of resistant isolates precluding incidence comparisons, or study scope outside our criteria (e.g., ICU-only or exclusively veterinary samples). One additional study was qualitatively relevant but excluded from the quantitative synthesis due to the lack of isolate incidence data [27]. Consequently, the final meta-analysis was based on data from 18 included studies [28–45].
4.Study characteristics
4.1Isolate distribution by pathogen
The included studies varied considerably in scale, with sample sizes ranging from four isolates [42] to over 300 isolates [45]. The majority of studies investigated Enterococcus faecium [28,30,31,34,35,44], Pseudomonas aeruginosa [29,32,33,37,40], or Klebsiella pneumoniae [38,42,43]. In contrast, other ESKAPE pathogens were less frequently represented; Staphylococcus aureus [41], Acinetobacter baumannii [39], and other Klebsiella species [45] were each addressed in a single study. No eligible studies reported isolates belonging to the genus Enterobacter (Supplementary Table 3).
4.2Geographic distribution
The geographical distribution of the included studies is illustrated in Figure 2. Most studies were conducted in Europe, followed by Asia. Africa and Oceania were represented by an equal, smaller number of studies, and North America was the location for a single study. At the national level, Australia was the most frequent study location with three publications. A single study was contributed by each of the following countries: China, Croatia, the Czech Republic, Ethiopia, France, Greece, India, Iran, Italy, Japan, Mexico, Nigeria, Portugal, South Korea, and Spain. The research included in this review covers a period of more than two decades, with the earliest study published in 1999 and the most recent in 2024.
5.Discussion
5.1Search and screening results
This systematic review and meta-analysis compared AMR in ESKAPE pathogens from aquatic and clinical sources, identifying 18 eligible studies for inclusion. The limited number of studies highlights a persistent gap in AMR research, particularly the lack of direct comparative analysis integrating environmental and clinical isolates within a One Health framework [7].
Although culture-based approaches capture only a small fraction of environmental microbial diversity [46], they provide direct evidence of phenotypic resistance by demonstrating the survival of viable bacteria under antimicrobial exposure. This functional confirmation allows a more reliable assessment of active resistance potential across environmental and clinical compartments [47].
5.2Study characteristics
5.2.1Isolate distribution by pathogen
In this review, the most frequently reported ESKAPE pathogens were E. faecium, P. aeruginosa, and K. pneumoniae, all well-documented opportunistic pathogens associated with multidrug resistance and healthcare-associated infections [13,48]. Their predominance in studies comparing clinical and environmental samples likely reflects their ecological adaptability, allowing persistence across diverse niches and facilitating recovery under laboratory conditions. E. faecium is a ubiquitous organism found in sources ranging from animal gastrointestinal tracts to water, soil, and wastewater, largely due to its tolerance to adverse conditions such as variations in temperature, salinity, and pH [12,13]. Similarly, P. aeruginosa is an ecologically versatile species common in water and soil, whose metabolic flexibility and capacity to form biofilms support persistence in both natural and hospital ecosystems [49]. K. pneumoniae also occupied diverse niches, being frequently detected in water and soil and capable of surviving under both aerobic and anaerobic conditions [14].
In contrast, S. aureus and A. baumannii exhibit narrower ecological ranges. S. aureus is primarily adapted to colonize human skin and mucosa and is typically detected at lower concentrations in environmental samples [15,50], whereas A. baumannii remains predominantly associated with hospital environments, despite occasional recovery from soil or water [51].
No eligible studies reported Enterobacter species. Although members of this genus have been detected in soil, wastewater, and aquatic ecosystems [16], their absence may reflect lower prevalence in human infections compared to other ESKAPE pathogens [52], as well as challenges in species-level identification using conventional methods.
5.2.2Geographic distribution
The geographic distribution of the included studies indicates a marked bias in AMR research on ESKAPE pathogens, with most studies conducted in high-income countries. This pattern likely reflects differences in research infrastructure, funding availability, and maturity of AMR surveillance programs. Such disparities are captured by the Global One Health Index for AMR (GOHI-AMR), which consistently shows higher scores for high-income countries [53], reinforcing the impact of unequal monitoring capacity and research development on the global AMR evidence base.
6.Conclusion
This meta-analysis reveals a complex and highly heterogeneous landscape of AMR in ESKAPE pathogens across clinical and environmental settings. Our findings confirm that resistance prevalence is significantly higher in clinical isolates, reflecting the intense selective pressures in healthcare environments characterized by high antibiotic use and vulnerable patient populations. However, this overall pattern is strongly influenced by substantial methodological and ecological heterogeneity among studies. This high statistical heterogeneity observed was not merely a sampling artifact but a genuine reflection of variability in the primary literature, driven by the pooling of distinct environmental matrices and the lack of standardized laboratory protocols.
These limitations were most evident in the anomalous finding that rifamycins, nitrofurans, polymyxins, and streptogramins showed higher resistance in environmental samples. This counterintuitive result did not represent a true ecological signal but rather an artifact arising from a small number of studies focused on wastewater hotspots and employing enrichment techniques that inflate resistance estimates. Consistently, Egger’s test indicated that funnel plot asymmetry for aminoglycosides, fluoroquinolones, and β-lactams was not indicative of publication bias but instead reflected uneven representation of individual antibiotics within aggregated classes.
In summary, although antimicrobial resistance remains more concentrated in clinical settings, our synthesis highlights a critical limitation in current AMR research: the lack of standardized sampling and analytical frameworks severely constrains data integration. The contribution of environmental reservoirs to the clinical AMR burden cannot be accurately quantified until harmonized methodologies are adopted. Future studies should prioritize standardized protocols to improve comparability and enable a clearer understanding of transmission dynamics between environmental and clinical compartments.
Supporting information
Data Availability
All data produced in the present study are available upon reasonable request to the authors
Funding and Conflict of interest
The funding for this research was provided through reparation funds originating from Vale mining company.
Declaration of generative AI use
During the preparation of this work, the author(s) used Resea in order to improve the clarity, conciseness, and overall readability of the manuscript. After using this service, the author(s) reviewed and edited the content as needed and took full responsibility for the content of the published article.
Supplementary files
Supplementary Figure 1. Comparison of pooled antimicrobial resistance proportions between clinical and environmental bacterial isolates by antimicrobial class and agent.
Supplementary Figure 2. Forest plots of pooled antimicrobial resistance proportions by antibiotic class comparing clinical and environmental bacterial isolates.
Supplementary Figure 3. Funnel plots of pooled antimicrobial resistance estimates for clinical and environmental isolates by antibiotic class.
Supplementary table 1. Search terms
Supplementary table 2. Characteristics of included studies and antimicrobial resistance profiles of bacterial isolates from clinical and environmental sources.
Supplementary table 3. Overview of clinical and environmental samples, sample sources, number of isolates, geographic distribution, ESKAPE pathogens, isolation methods and identification techniques in the included studies.
Supplementary table 4. Pooled antimicrobial resistance and heterogeneity metrics by site (clinical and environmental) and dataset type (all, with effluent, without effluent.
Supplementary table 5. Publication heterogeneity on the patterns of antimicrobial resistance in clinical and environmental samples.