https://orcid.org/0000-0003-0917-4288
Figures
Abstract
Background
Antimicrobial resistance (AMR) is a major global health concern. Klebsiella pneumoniae is a key pathogen in healthcare-associated infections due to the production of extended-spectrum beta-lactamases (ESBLs), which limit therapeutic options.
Objective
To estimate the prevalence and identify factors associated with the presumptive ESBL phenotype in K. pneumoniae isolates from hospitals in the Meta Department, Colombia, between 2018 and 2022.
Methods
We conducted a retrospective analytical cross-sectional study using 4,808 non-carbapenem-resistant K. pneumoniae isolates obtained from routine microbiological surveillance. Bivariate analyses (Chi-square and Mann–Whitney U tests) and multivariable logistic regression were performed to evaluate associations between ESBL production and demographic, clinical, and institutional variables.
Results
A total of 4,808 non-carbapenem-resistant Klebsiella pneumoniae isolates were included after applying exclusion criteria. The overall prevalence of the presumptive ESBL phenotype was 22.9% (n = 1,100). In bivariate analysis, ESBL phenotype was significantly associated with male sex, older age, inpatient care, hospital type, and specimen type, particularly urine and blood samples. In multivariable analysis, male sex, inpatient status, and urine specimens remained independently associated, along with variability across specific hospitals, indicating the influence of institutional factors. The model showed limited discriminative ability (AUC = 0.62) and low explanatory power (pseudo-R² 0.03–0.05).
Conclusions
The presumptive ESBL phenotype in Klebsiella pneumoniae is highly prevalent and increasing over time. Its association with demographic, clinical, and institutional factors highlights the multifactorial nature of antimicrobial resistance and supports the need for strengthened surveillance and targeted antimicrobial stewardship strategies.
Citation: Cocunubo-Cocunubo NG, Pavas-Escobar N, Montilla Rodríguez LM, Soto-Gutiérrez F, Rojas-Zapata AK, Sánchez-Pabón L, et al. (2026) Extended-spectrum beta-lactamase phenotype in Klebsiella pneumoniae isolates from Meta Department hospitals, Colombia, 2018–2022: A retrospective analytical cross-sectional study. PLoS One 21(8): e0352453. https://doi.org/10.1371/journal.pone.0352453
Editor: Tombari Pius Monsi, Rivers State University, NIGERIA
Received: February 5, 2026; Accepted: June 10, 2026; Published: August 4, 2026
Copyright: © 2026 Cocunubo-Cocunubo et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data used during the study are available in the institutional repository (Papyrus) of the Cooperative University of Colombia: http://doi.org/10.57924/ZXEAJN/BMWKAI.
Funding: This study was funded by the Universidad Cooperativa de Colombia, which enabled the execution of the research activities, the collection and analysis of data, as well as the preparation and writing of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Antimicrobial resistance (AMR) is recognized as one of the most critical global public health challenges, as it contributes to increased morbidity and mortality, prolonged hospital stays, and a substantial rise in healthcare costs [1,2]. The rapid emergence and dissemination of resistant bacterial pathogens have led international organizations, including the World Health Organization (WHO), to recommend the integration of AMR surveillance as a central component of antimicrobial stewardship programs (ASPs), aimed at promoting rational antibiotic use and preserving the effectiveness of existing therapies [3,4].
Klebsiella pneumoniae is a clinically significant opportunistic Gram-negative pathogen and a major cause of healthcare-associated infections (HAIs), including pneumonia, bloodstream infections, urinary tract infections, and intra-abdominal infections [5]. It’s remarkable capacity to acquire and disseminate resistance determinants through mobile genetic elements has positioned K. pneumoniae as a key member of the ESKAPE group of pathogens, which are responsible for a substantial proportion of multidrug-resistant infections worldwide [6].
A major concern is the increasing prevalence of resistance to third-generation cephalosporins, commonly associated with the presence of extended-spectrum beta-lactamases (ESBLs). These enzymes confer resistance to a broad range of beta-lactam antibiotics, significantly limiting therapeutic options and often leading to increased reliance on carbapenems, thereby accelerating the emergence of resistance to last-line agents [7,8]. Infections caused by Enterobacterales exhibiting ESBL-associated resistance phenotypes have been associated with delays in appropriate antimicrobial therapy, increased mortality, and higher healthcare costs [9].
In low- and middle-income countries (LMICs), including Colombia, K. pneumoniae with ESBL-associated resistance phenotypes represents a major cause of HAIs, particularly in intensive care units (ICUs), where high antibiotic pressure, invasive procedures, and prolonged hospitalization favor the selection and transmission of resistant strains [10]. Surveillance data at the national level have consistently reported high resistance rates to third-generation cephalosporins in K. pneumoniae, with similar patterns observed across different hospital settings. In this context, resistance to these antibiotics is frequently used as a phenotypic screening approach for identifying presumptive ESBL phenotype, although confirmatory testing based on clavulanate inhibition is not always available [11].
Despite the availability of national surveillance data, substantial heterogeneity persists in antimicrobial resistance patterns across regions, healthcare institutions, and clinical settings. Variability in geographic, demographic, and healthcare system characteristics within Colombia may significantly influence local resistance dynamics, thereby limiting the applicability of aggregated national estimates for guiding empirical antibiotic therapy at the regional level. Recent evidence indicates that resistance profiles may vary considerably even within the same country, underscoring the need for region-specific analyses to inform clinical decision-making and public health interventions [12,13].
The Meta Department, located in the Colombian Orinoquía region, presents distinct epidemiological and healthcare characteristics; however, data on the burden and determinants of resistance phenotypes compatible with ESBL in K. pneumoniae remain limited. Strengthening regional surveillance is essential to support evidence-based empirical treatment guidelines, optimize antimicrobial stewardship strategies, and inform infection prevention and control policies.
Therefore, the aim of this study was to estimate the prevalence and identify independent factors associated with the presumptive ESBL phenotype in Klebsiella pneumoniae isolates collected from hospitals in the Meta Department, Colombia, between 2018 and 2022, using an analytical cross-sectional design based on retrospectively collected surveillance data
2. Materials and methods
2.1. Study design and data source
An analytical cross-sectional study based on retrospectively collected surveillance data was conducted using a secondary database derived from routine microbiological surveillance. The dataset included records from 11 hospitals in the Meta Department, Colombia, covering the period from January 2018 to December 2022. Within the framework of the National Antimicrobial Resistance Surveillance Program coordinated by the Colombian National Institute of Health, all participating institutions systematically reported microbiological data monthly to the Departmental Public Health Laboratory. This surveillance system operates under standardized national protocols for data collection, validation, and reporting. No direct involvement of human participants occurred, as the study was based exclusively on anonymized secondary data.
2.2. Microbiological methods and ESBL definition
All participating hospitals employed automated systems for bacterial identification and antimicrobial susceptibility testing (AST), such as Vitek® or Phoenix®, in accordance with each laboratory’s standard operating procedures. Susceptibility results were interpreted following the Clinical and Laboratory Standards Institute (CLSI) guidelines applicable to each year of the surveillance period.
For epidemiological surveillance purposes, a presumptive ESBL phenotype was defined based on phenotypic screening criteria, as confirmatory clavulanate inhibition tests were not uniformly available across all participating institutions. According to the Colombian National Institute of Health, microbiological reports used for antimicrobial resistance surveillance were standardized prior to their incorporation into the regional surveillance system [14]. This approach is consistent with antimicrobial resistance surveillance systems in low- and middle-income countries and aligns with the methodological framework recommended by the World Health Organization through the Global Antimicrobial Resistance Surveillance System (GLASS), ensuring international comparability [15–17].
In this study, Klebsiella pneumoniae isolates were classified as exhibiting a presumptive ESBL phenotype if they showed resistance to ceftazidime (CAZ; minimum inhibitory concentration [MIC] > 4 μg/mL) and/or cefotaxime (CTX; MIC > 1 μg/mL), in accordance with INS protocols and CLSI guidelines. This definition reflects a screening-based phenotypic approach and does not confirm ESBL production; therefore, it may include other resistance mechanisms, such as AmpC β-lactamases. To reduce misclassification and avoid the inclusion of carbapenem-resistant Enterobacterales (CRE), isolates showing resistance to any tested carbapenem (ertapenem, imipenem, or meropenem) were excluded from the analysis.
2.3. Study population and inclusion criteria
The surveillance database contained information on the date of isolation, the type and level of the healthcare facility, the hospital ward (including intensive care unit [ICU], inpatient ward, emergency department, or outpatient setting), the type of biological specimen (as reported by the laboratory), bacterial identification, and antimicrobial resistance profile. Initially, all microbiological isolation records from hospitals in Villavicencio and Granada (Meta Department, Colombia) during the study period were considered.
The original database included patient-level identifiers (internal codes, national identification numbers, and names), which allowed the identification of multiple records from the same individual. Prior to analysis, the dataset was reviewed and deduplicated to ensure that only one isolate per patient was included. Subsequently, all personal identifiers were removed to guarantee data anonymization and compliance with ethical standards.
For the final analysis, only records corresponding to Klebsiella pneumoniae isolates were included. Records were excluded if they showed documented carbapenem resistance, corresponded to laboratory quality control procedures, or lacked complete information on bacterial identification or antimicrobial susceptibility results.
The primary outcome was the presumptive ESBL phenotype, defined phenotypically as resistance to ceftazidime (CAZ) and/or cefotaxime (CTX), based on established minimum inhibitory concentration (MIC) cutoff values. This definition corresponds to a screening-based criterion derived from microbiological surveillance data and does not imply confirmed ESBL production.
2.4. Data analysis
Categorical variables were summarized using absolute frequencies and proportions. Quantitative variables were assessed for distribution and described using the median and interquartile range (IQR) or the mean and standard deviation, as appropriate. Data were exported to Microsoft Excel and analyzed using Prism software (GraphPad, version 10.0.0 for Mac/iOS, Boston, Massachusetts, USA).
Bivariate analyses were performed to evaluate associations between independent variables (age, sex, hospital type, level of care, hospital ward, and specimen type) and the outcome of interest, defined as the presumptive ESBL phenotype. Differences between categorical variables were assessed using the chi-square (χ²) test or Fisher’s exact test, as appropriate. Quantitative variables (e.g., age) were compared using the Mann–Whitney U test due to non-normal distribution. Measures of association were reported as odds ratios (ORs) with 95% confidence intervals (95% CIs). A p-value < 0.05 was considered statistically significant.
Subsequently, a multivariable logistic regression model was constructed to identify independent factors associated with the presumptive ESBL phenotype. Given the structure of the available surveillance data, variables included in the model were selected based on both statistical criteria (p < 0.20 in bivariate analysis) and epidemiological relevance. Although clinically relevant variables such as prior antibiotic use, comorbidities, and length of hospital stay were not available in the dataset, the model was designed to evaluate structural and demographic factors associated with the outcome. Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC), and explanatory capacity was evaluated using Cox & Snell and Nagelkerke pseudo-R² statistics
2.5. Ethical considerations
The present study was based exclusively on the analysis of secondary data derived from routine microbiological surveillance systems in public health, systematically collected by participating institutions under previously established institutional and national protocols. There was no direct involvement of human participants, and no biological samples were collected for research purposes. The study protocol was reviewed by the Bioethics Subcommittee of the Universidad Cooperativa de Colombia (Villavicencio Campus), which determined that the research met the criteria for exemption from ethical approval, as it involved the use of fully anonymized secondary data and posed no risk to individuals. Accordingly, the study was classified as research without risk, and a formal waiver of ethical approval was granted. The database was anonymized by the corresponding public health authority prior to access by the investigators, with all information that could allow direct or indirect identification of individuals removed. At no point during or after data access did the authors have access to identifiable information. The study was conducted in accordance with national regulations for health research (Resolution 8430 of 1993) [18] and international ethical standards, including the Declaration of Helsinki [19] and the guidelines of the Council for International Organizations of Medical Sciences (CIOMS) in collaboration with the World Health Organization (WHO) [20]. To ensure institutional confidentiality, each participating hospital was assigned a letter of the alphabet (A–K). The quality of reporting was assessed using the STROBE checklist.
3. Results
3.1. Flowchart of selection, exclusion, and classification of Klebsiella pneumoniae isolates according to presumptive extended-spectrum β-lactamase (ESBL) production
As shown in Fig 1, a total of 41,163 bacterial isolates (100%) were initially recorded, of which 7,471 (18.1%) corresponded to Klebsiella pneumoniae. Subsequently, exclusion criteria were applied to ensure data quality and analytical consistency. A total of 1,741 isolates (23.3%) identified as carbapenemase-producing strains were excluded due to their distinct resistance profile, along with 922 records (12.3%) due to missing or incomplete data. After these exclusions, 4,808 isolates (64.4%) were included in the final analytical dataset. Within this cohort, isolates were classified according to a presumptive extended-spectrum β-lactamase (ESBL) phenotype based on phenotypic screening criteria. Overall, 1,100 isolates (22.9%) were classified as presumptive ESBL producers, whereas 3,708 (77.1%) were categorized as non-ESBL.
Flowchart showing the selection of Klebsiella pneumoniae isolates included in the study and the application of exclusion criteria, resulting in the final analytical dataset and classification according to presumptive ESBL phenotype. ESBL = extended-spectrum beta-lactamase. Source: Original from the authors.
3.2. Descriptive characteristics of presumptive ESBL-producing Klebsiella pneumoniae isolates
A total of 4,808 Klebsiella pneumoniae isolates were analyzed, of which 1,100 (22.9%) were classified as presumptive extended-spectrum β-lactamase (ESBL) producers and 3,708 (77.1%) as non-ESBL (Table 1). Regarding demographic characteristics, male patients predominated (56.1%) compared to females (43.9%). The median age was 61 years (IQR: 43–73), with a slightly higher median observed among ESBL cases (63 years) compared to non-ESBL isolates. In terms of patient type, most isolates corresponded to adults (92.0%), followed by geriatric patients (42.8%), while pediatric (6.4%) and neonatal (1.6%) groups represented smaller proportions. With respect to institutional origin, most isolates were obtained from private hospitals (62.6%), compared to public hospitals (37.4%). At the institutional level, hospitals D (30.6%), F (30.5%), and J (23.7%) accounted for the highest proportion of isolates. According to hospital service, isolates were primarily obtained from inpatient wards (36.2%), followed by intensive care units (30.8%) and emergency departments (29.3%), while the outpatient setting represented a smaller proportion (3.7%). Regarding specimen type, the most frequent sources were urine (33.3%), tracheal secretions (20.3%), and blood (17.5%), followed by other secretions (9.2%) and sputum (5.9%). Finally, the temporal distribution showed a higher frequency of isolates in 2021 (27.8%), followed by 2019 (21.3%), 2020 (19.2%), 2022 (16.2%), and 2018 (15.5%).
3.3. Bivariate analysis of the association between selected variables and presumptive ESBL production in Klebsiella pneumoniae
As shown in Table 1, the bivariate analysis identified multiple variables associated with presumptive extended-spectrum β-lactamase (ESBL) production in Klebsiella pneumoniae isolates. Regarding demographic characteristics, male sex was associated with a higher likelihood of ESBL production compared to females (OR = 1.31; 95% CI: 1.14–1.50; p < 0.001). A significant association was also observed with age (p = 0.03), with a higher median age among ESBL cases. No statistically significant associations were identified according to patient type (p > 0.05), although higher proportions were observed in pediatric and geriatric groups. With respect to the institution of origin, heterogeneity was observed across hospitals. Significant associations were identified in hospitals A (OR = 2.59; 95% CI: 1.52–4.42; p < 0.001), C (OR = 2.88; 95% CI: 1.37–6.03; p < 0.001), D (OR = 1.39; 95% CI: 1.01–1.91; p = 0.04), I (OR = 3.04; 95% CI: 1.89–4.90; p < 0.001), and J (OR = 2.18; 95% CI: 1.59–3.00; p < 0.001). At an aggregate level, isolates from private hospitals showed higher odds of ESBL production compared to public hospitals (OR = 1.55; 95% CI: 1.34–1.79; p < 0.001), as illustrated in Fig 2. Regarding hospital service, isolates from inpatient settings were significantly associated with ESBL production (OR = 1.82; 95% CI: 1.22–2.71; p < 0.001), whereas no significant differences were observed for ICU or emergency departments. This pattern is shown in Fig 3A. In relation to specimen type, significant associations were identified for urine (OR = 2.56; 95% CI: 1.77–3.69; p < 0.001) and blood (OR = 2.08; 95% CI: 1.42–3.04; p < 0.001), as presented in Fig 3B. Finally, the temporal analysis showed significant associations in 2018, 2019, 2020, and 2022, with 2022 showing the highest magnitude (OR = 2.53; 95% CI: 2.06–3.12; p < 0.001) compared to 2021 as the reference category. The bivariate analysis demonstrates that presumptive ESBL production in K. pneumoniae is significantly associated with demographic, institutional, clinical, and microbiological factors, showing patterns consistent with those observed in Figs 2 and 3, particularly in relation to hospital type, care setting, and specimen type.
The figure compares the proportion of presumptive ESBL-producing Klebsiella pneumoniae isolates between public and private hospitals during 2018–2022. Bars represent percentages of ESBL-positive isolates in each hospital type. ESBL = extended-spectrum beta-lactamase. Source: Original from the authors.
(A) Distribution of Klebsiella pneumoniae isolates according to hospital ward. (B) Distribution according to clinical specimen. Bars show the total number of isolates and the proportion of presumptive ESBL-producing isolates in each category. ESBL = extended-spectrum beta-lactamase. Source: Original from the authors.
3.4. Multivariable logistic regression analysis of factors associated with presumptive ESBL production in Klebsiella pneumoniae isolates
As shown in Table 2 and Fig 4, a multivariable logistic regression model was developed to identify independent factors associated with presumptive extended-spectrum β-lactamase (ESBL) production in Klebsiella pneumoniae isolates. Variables were selected based on statistical criteria (p < 0.20 in the bivariate analysis) and epidemiological relevance. For parsimony, only variables that remained statistically significant after adjustment are presented. After adjustment, male sex, inpatient status, and urine specimens remained independently associated with ESBL production. At the institutional level, specific hospitals (A, D, I, and J) also showed statistically significant associations, indicating variability across centers. These factors were associated with increased odds of ESBL production, with adjusted odds ratios greater than 1 and 95% confidence intervals that did not cross unity. The model showed an area under the curve (AUC) of 0.62, indicating limited discriminative ability. In addition, the model demonstrated modest explanatory power, with a Cox & Snell pseudo-R² of 0.0302 and a Nagelkerke pseudo-R² of 0.0458, indicating that the variables included explained approximately 3% to 5% of the variability in the outcome.
Forest plot showing adjusted odds ratios (ORs) and 95% confidence intervals obtained from the multivariable logistic regression model evaluating factors associated with presumptive ESBL production. Variables with confidence intervals not crossing the null value indicate statistically significant associations. ESBL = extended-spectrum beta-lactamase; OR = odds ratio; CI = confidence interval. Source: Original from the authors.
4. Discussion
This retrospective analysis of over 7,000 Klebsiella pneumoniae isolates provides robust epidemiological evidence of a concerning increase in the presumptive extended-spectrum β-lactamase (ESBL) phenotype in the Meta Department, Colombia. The observed prevalence (22.9%) is consistent with reports from Latin American antimicrobial resistance surveillance networks, which have documented a sustained rise in resistance to third-generation cephalosporins among Enterobacterales [21–24]. Compared with local data from approximately two decades ago reporting resistance rates around 12.5%, our findings suggest a substantial increase, highlighting persistent selective pressure and the limited effectiveness of current containment strategies [25].
A notable finding is the marked increase in the presumptive ESBL phenotype in 2022 (33.5%), following a transient decline in 2021. This pattern aligns with global evidence describing a post-pandemic rebound in antimicrobial resistance, likely driven by the widespread use of broad-spectrum antibiotics during the COVID-19 pandemic, which may have intensified selective pressure on bacterial populations [26,27]. These trends underscore the need to reassess empiric treatment strategies and reinforce antimicrobial stewardship programs, particularly in low- and middle-income countries where regulatory and surveillance capacities may be heterogeneous [21].
In the multivariable analysis, male sex, inpatient status, and urine specimens remained independently associated with the presumptive ESBL phenotype. Additionally, variability across hospitals suggests that institutional-level factors, such as infection control practices, antimicrobial prescribing patterns, and diagnostic capacity, may play a relevant role in shaping resistance dynamics. Importantly, variables that appeared significant in the bivariate analysis, such as hospital ward, did not retain statistical significance after adjustment, indicating that these associations were likely confounded by other factors rather than reflecting an independent effect. This finding reinforces the importance of multivariable modeling to disentangle complex epidemiological relationships [28]
The higher prevalence observed in general wards in the bivariate analysis (27.8% vs. ICU 17.6%) should therefore be interpreted with caution, as this association was not maintained after adjustment. This suggests that the observed differences are likely explained by confounding factors rather than by the ward itself, including variations in patient characteristics, sampling practices, and exposure to healthcare interventions.
Furthermore, the association between the presumptive ESBL phenotype and older age groups supports the hypothesis that healthcare exposure and cumulative antibiotic use contribute to the maintenance of resistant reservoirs, as previously described in international studies [13,29]. Older adults are frequently identified as key reservoirs for ESBL-producing Enterobacterales due to repeated contact with healthcare systems, higher comorbidity burden, and increased antibiotic exposure [30]. These findings support targeted infection prevention strategies, including risk-based screening and strengthened surveillance in high-risk populations [31–33].
The analysis by specimen type further highlights the role of invasive devices and urinary tract infections in the epidemiology of antimicrobial resistance [30]. Similar patterns have been observed in multicenter studies from Europe and North America, where catheter-associated biofilms act as persistent reservoirs for ESBL-producing K. pneumoniae and facilitate intra-hospital spread [34].The high frequency observed in catheter-related samples (51.0%) and urine (27.3%) underscores the importance of “device stewardship,” with protocols emphasizing timely removal of catheters to reduce biofilm-related colonization, as recommended by international guidelines [35,36]. Additionally, the detection of resistance in outpatient samples suggests an increasing convergence between community- and healthcare-associated resistance, a phenomenon widely reported and associated with growing challenges in empirical treatment selection, particularly for urinary tract infections
From a modeling perspective, the observed limited discriminative ability (AUC = 0.62) and low explanatory power (pseudo-R² < 0.05) are consistent with the use of secondary surveillance data lacking detailed clinical variables. These findings indicate that the occurrence of the presumptive ESBL phenotype is influenced by complex and multifactorial determinants not fully captured in the available dataset, rather than reflecting poor model performance per se. Similar observations have been reported in epidemiological studies based on routine surveillance systems, where the absence of clinical, behavioral, and treatment-related variables limits predictive performance but does not compromise the identification of relevant associations.
From a public health perspective, these findings emphasize the need to strengthen regional antimicrobial resistance surveillance systems and to integrate microbiological data with clinical and structural health system indicators. Such integration is essential to inform context-specific interventions, optimize antimicrobial stewardship strategies, and improve infection prevention policies. Continued investment in surveillance infrastructure, along with the gradual incorporation of molecular resistance monitoring, represents a critical component of national and global strategies to address antimicrobial resistance [16,17]. Although phenotypic detection may underestimate underlying molecular diversity, it remains a fundamental tool for population-level surveillance in resource-limited settings.
These findings should be interpreted considering several limitations inherent to the study design and data source. First, the retrospective nature of the study, based on secondary microbiological surveillance data, may be subject to selection bias as well as limitations in data quality and completeness. Second, the definition of the presumptive extended-spectrum β-lactamase (ESBL) phenotype was based on phenotypic screening criteria derived from resistance to third-generation cephalosporins without confirmatory testing. Although this approach is consistent with large-scale antimicrobial resistance surveillance systems, it may have introduced misclassification bias by including other resistance mechanisms, such as AmpC β-lactamases, potentially leading to non-differential misclassification and affecting the accuracy of ESBL frequency estimates. Additionally, the database did not include key clinical variables such as comorbidities, prior antibiotic exposure, clinical diagnoses, or length of hospital stay. The absence of these variables limits the ability to control for residual confounding and restricts causal interpretation, as the analysis was based exclusively on microbiological and structural variables available in the surveillance system. From an analytical perspective, the limited discriminative ability (AUC = 0.62) and low explanatory power (pseudo-R² < 0.05) are consistent with the use of secondary surveillance data lacking detailed clinical information. These findings suggest that the occurrence of the presumptive ESBL phenotype is influenced by complex and multifactorial determinants not fully captured in the available dataset.
Furthermore, the multivariable regression model did not account for clustering at the hospital level. Although isolates were deduplicated to ensure one observation per patient, patients were nested within healthcare institutions, and therefore observations may not be fully independent. Individuals within the same hospital are likely to share similar clinical practices, antimicrobial use patterns, and infection control measures. This lack of adjustment may have influenced variance estimates, potentially resulting in underestimated standard errors and, consequently, an overestimation of the precision of the associations observed.
Despite these limitations, the use of a large, multicenter surveillance dataset provides a robust epidemiological framework for identifying resistance patterns and associated factors at the population level, particularly in settings where integration of clinical and microbiological data remains limited. Future studies should incorporate patient-level clinical variables, apply confirmatory ESBL testing, and consider hierarchical or multilevel modeling approaches to improve internal validity and the accuracy of effect estimates.
In conclusion, Klebsiella pneumoniae isolates from hospitals in the Meta Department exhibit a high and increasing prevalence of the presumptive extended-spectrum β-lactamase (ESBL) phenotype, with evidence of temporal escalation in recent years. Independent associations identified in the multivariable analysis highlight the multifactorial nature of antimicrobial resistance in this setting, rather than a single dominant determinant. These findings underscore the importance of strengthening regional surveillance systems, reinforcing antimicrobial stewardship programs, and integrating microbiological, clinical, and molecular data to better characterize resistance dynamics. Such efforts are essential to support evidence-based decision-making and to design targeted interventions aimed at mitigating the spread of antimicrobial resistance in both healthcare and community settings.
Acknowledgments
Thanks to Norton A. Perez Jr. for compiling all the information into a single database for analysis; recognition to the Meta Department of Health for contributing to this project; and acknowledgments to the Public Health Laboratory for providing the necessary information on microbiological isolate data.
References
- 1. Murray CJ, Ikuta KS, Sharara F, Swetschinski L, Robles Aguilar G, Gray A, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399(10325):629–55. pmid:35065702
- 2. Maragakis LL, Perencevich EN, Cosgrove SE. Clinical and economic burden of antimicrobial resistance. Expert Rev Anti Infect Ther. 2008;6(5):751–63. pmid:18847410
- 3.
World Health Organization. Global action plan on antimicrobial resistance. 2016.
- 4. Quirós RE, Bardossy AC, Angeleri P, Zurita J, Aleman Espinoza WR, Carneiro M, et al. Antimicrobial stewardship programs in adult intensive care units in Latin America: Implementation, assessments, and impact on outcomes. Infect Control Hosp Epidemiol. 2022;43(2):181–90. pmid:33829982
- 5. Revoredo Rego F, Huamán Egoávil E, Zegarra Cavani S, Auris Mora H, Valderrama Barrientos R. Microbiological and resistance profiles of community acquired and nosocomial intra abdominal infections in surgery of National Hospital Guillermo Almenara, Lima, Peru. Rev Gastroenterol Peru. 2016;36(2):115–22. pmid:27409087
- 6. Supplement 1. Clinical Infectious Diseases. 2009;48(9):1334–1334.
- 7. Paterson DL, Bonomo RA. Extended-spectrum beta-lactamases: a clinical update. Clin Microbiol Rev. 2005;18(4):657–86. pmid:16223952
- 8. Bush K, Bradford PA. Epidemiology of β-Lactamase-Producing Pathogens. Clin Microbiol Rev. 2020;33(2):e00047-19. pmid:32102899
- 9. Tamma PD, Aitken SL, Bonomo RA, Mathers AJ, van Duin D, Clancy CJ. Infectious Diseases Society of America 2022 Guidance on the Treatment of Extended-Spectrum β-lactamase Producing Enterobacterales (ESBL-E), Carbapenem-Resistant Enterobacterales (CRE), and Pseudomonas aeruginosa with Difficult-to-Treat Resistance (DTR-P. aeruginosa). Clin Infect Dis. 2022;75(2):187–212. pmid:35439291
- 10. De La Cadena E, Pallares CJ, García-Betancur JC, Porras JA, Villegas MV. Actualización sobre la resistencia antimicrobiana en instituciones de salud de nivel III y IV en Colombia entre enero del 2018 y diciembre del 2021. Biomédica. 2023;43(4):457–73.
- 11. Schuetz AN, Ferrell A, Hindler JA, Humphries R, Bobenchik AM. Overview of changes in the Clinical and Laboratory Standards Institute Performance Standards for Antimicrobial Susceptibility Testing: M100 32nd and 33rd editions. J Clin Microbiol. 2025;63(9).
- 12. Olaitan MO, Orababa OQ, Shittu RB, Oyediran AA, Obunukwu GM, Arowolo MT, et al. Extended-spectrum beta-lactam-resistant Klebsiella pneumoniae in sub-Saharan Africa: a systematic review and meta-analysis from a One Health perspective. BMC Infect Dis. 2025;25(1):843. pmid:40597808
- 13. van Duin D, Doi Y. The global epidemiology of carbapenemase-producing Enterobacteriaceae. Virulence. 2017;8(4):460–9. pmid:27593176
- 14. Villalobos AP, Barrero LI, Rivera SM, Ovalle MV, Valera D. Vigilancia de infecciones asociadas a la atención en salud, resistencia bacteriana y consumo de antibióticos en hospitales de alta complejidad, Colombia, 2011. Biomédica. 2013;34(0):67.
- 15.
World Health Organization. GLASS Manual for Antimicrobial Resistance Surveillance in Common Bacteria Causing Human Infection. 2023.
- 16. O’Brien TF, Stelling J. Integrated Multilevel Surveillance of the World’s Infecting Microbes and Their Resistance to Antimicrobial Agents. Clin Microbiol Rev. 2011;24(2):281–95. pmid:21482726
- 17. Sulis G, Sayood S, Gandra S. Antimicrobial resistance in low- and middle-income countries: current status and future directions. Expert Rev Anti Infect Ther. 2022;20(2):147–60. pmid:34225545
- 18. Ministerio de Salud. Normas científicas, técnicas y administrativas para la investigación en salud. 1993. https://www.minsalud.gov.co/sites/rid/lists/bibliotecadigital/ride/de/dij/resolucion-8430-de-1993.pdf
- 19. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310(20):2191–4. pmid:24141714
- 20. Alcalde Bezhold G, Alfonso Farnós I. El acceso a la historia clínica en la investigación observacional. Revista Española de Medicina Legal. 2018;44(3):121–30.
- 21. Quirós R, Angeleri P, Zurita J, Aleman W, Carneiro M, Guerra S, et al. Impact of Antimicrobial Stewardship Programs in Latin American Adult Intensive Care Units: PROA-LATAM Project. Infect Control Hosp Epidemiol. 2020;41(S1):s520–s520.
- 22. Zhuang Q, Guo H, Peng T, Ding E, Zhao H, Liu Q, et al. Advances in the detection of β-lactamase: A review. Int J Biol Macromol. 2023;251:126159. pmid:37549760
- 23. Lawrence J, O’Hare D, van Batenburg-Sherwood J, Sutton M, Holmes A, Rawson TM. Innovative approaches in phenotypic beta-lactamase detection for personalised infection management. Nat Commun. 2024;15(1):9070. pmid:39433753
- 24. Mohammed AB, Anwar KA. Phenotypic and genotypic detection of extended spectrum beta lactamase enzyme in Klebsiella pneumoniae. PLoS One. 2022;17(9):e0267221. pmid:36173938
- 25. Ayan M, Çelik AK. Klebsiella pneumoniae infections in the intensive care unit: risk factors related to carbapenem resistance and mortality. J Infect Dev Ctries. 2025;19(2):248–57. pmid:40063750
- 26. Boattini M, Bianco G, Comini S, Iannaccone M, Casale R, Cavallo R, et al. Direct detection of extended-spectrum-β-lactamase-producers in Enterobacterales from blood cultures: a comparative analysis. Eur J Clin Microbiol Infect Dis. 2022;41(3):407–13. pmid:34822029
- 27. Ramos DA, Moreno Henao LE, Bolaños OF, Roa Martínez ÁJ. ESBL-producing Enterobacteriaceae, an unnoticed pandemic with challenges in clinical practice, a mini-review. J Infect. 2024;6(3):1–5.
- 28. Jiménez A, Alvarado A, Gómez F, Carrero G, Fajardo C. Risk factors associated with the isolation of extended spectrum betalactamases producing Escherichia coli or Klebsiella pneumoniae in a tertiary care hospital in Colombia. Biomedica. 2014;34(Suppl 1):16–22. pmid:24968032
- 29. Cassini A, Högberg LD, Plachouras D, Quattrocchi A, Hoxha A, Simonsen GS, et al. Attributable deaths and disability-adjusted life-years caused by infections with antibiotic-resistant bacteria in the EU and the European Economic Area in 2015: a population-level modelling analysis. Lancet Infect Dis. 2019;19(1):56–66. pmid:30409683
- 30. Pérez Gutiérrez N, Eird. Resistencia a los antimicrobianos de las enterobacterias en un hospital de la Orinoquía colombiana. Acta Colomb Cuid Intens. 2008;8(2):67–74.
- 31. Daoud Z, Hakime N. Prevalence and susceptibility patterns of extended-spectrum betalactamase-producing Escherichia coli and Klebsiella pneumoniae in a general university hospital in Beirut, Lebanon. Rev Esp Quimioter. 2003;16(2):233–8. pmid:12973463
- 32. Husna A, Rahman MM, Badruzzaman ATM, Sikder MH, Islam MR, Rahman MT, et al. Extended-Spectrum β-Lactamases (ESBL): Challenges and Opportunities. Biomedicines. 2023;11(11):2937. pmid:38001938
- 33. Asokan S, Jacob T, Jacob J, AlSosowaa AA, Cherian T, Peijnenburg WJGM, et al. Klebsiella pneumoniae: A growing threat in the era of antimicrobial resistance. The Microbe. 2025;7:100333.
- 34. Donlan RM. Biofilms and Device-Associated Infections. Emerg Infect Dis. 2001;7(2):277–81.
- 35. Centers for Disease Control and Prevention CDC. CDC’s Core Infection Prevention and Control Practices for Safe Healthcare Delivery in All Settings. https://www.cdc.gov/infection-control/hcp/core-practices/index.html 2024.
- 36. WHO. Global Strategy for Containment of Antimicrobial Resistance. https://www.who.int/publications/i/item/who-global-strategy-for-containment-of-antimicrobial-resistance 2022.