Treatment gaps and risk factors for multidrug-resistant Enterobacterales in Belize: a 6-year (2018

Abstract

Background:

Antimicrobial resistance (AMR) in Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae is a global threat, yet data from Belize and the wider Caribbean are scarce. Understanding local resistance patterns is critical for effective treatment and stewardship.

Objectives:

To determine antibiotic susceptibility patterns, quantify multidrug resistance (MDR) in outpatient and inpatient settings, identify predictors of empirical treatment failure for urinary tract infections (UTIs), and analyze temporal and geographic trends.

Design:

Nationwide retrospective analysis of clinical isolates from Belize’s public health system (2018–2023).

Methods:

We analyzed 4591 clinical isolates. MDR was defined as non-susceptibility to ⩾3 drug categories. Multivariable logistic regression identified predictors of the risk of empirical treatment failure for UTIs. Joinpoint regression analyzed temporal trends. Geospatial analysis mapped hotspots of incidence and resistance.

Results:

E. coli was predominant (68%). Among 3005 urinary isolates, 1247 (41.5%) were resistant to both first-line oral antibiotics (ciprofloxacin and trimethoprim-sulfamethoxazole). Extended-spectrum β-lactamase (ESBL) production (adjusted odds ratio (aOR) 8.91, 95% CI 7.40–10.73), infection with K. pneumoniae (aOR 2.85, 95% CI 2.32–3.51) or E. cloacae (aOR 4.12, 95% CI 2.78–6.11), and older age were key predictors. Ciprofloxacin resistance in E. coli increased significantly post-2020 (Annual Percent Change 9.8%, p = 0.02). MDR was prevalent in outpatient settings (42.1% for E. coli, 61.2% for K. pneumoniae). Geospatial analysis revealed the Belize District as the incidence hotspot, while the Toledo District was the ESBL prevalence hotspot (29.5%).

Conclusion:

Belize faces a notable AMR crisis characterized by a high community burden of MDR Enterobacterales, a critical empirical treatment gap for UTIs, and accelerating resistance trends.

Plain language summary

Six-year Belize study reveals alarming antibiotic resistance in common gut bacteria, urging immediate action to preserve life-saving medicines

Researchers analyzed antibiotic resistance in three common bacteria (E. coli, K. pneumoniae, and E. cloacae) causing infections in Belize between 2018 and 2023, using 4,591 clinical isolates from patient samples across all six districts. The findings reveal a serious public health crisis. For urinary tract infections, the most common type studied, over 40% of bacteria were resistant to both ciprofloxacin and trimethoprim-sulfamethoxazole—the two main oral antibiotics doctors typically prescribe first. This means standard pills often fail before treatment begins. The bacteria themselves mattered greatly: K. pneumoniae and E. cloacae posed a higher risk of resistance than E. coli. The strongest predictor of treatment failure was ESBL production, an enzyme that degrades multiple antibiotics, increasing the risk of resistance nearly ninefold. Alarmingly, ciprofloxacin resistance accelerated rapidly after 2020, suggesting the COVID-19 pandemic may have contributed to the problem through increased antibiotic use. Crucially, most resistant infections originated in communities and not hospitals, demonstrating that resistance is widespread in the general population. Geographically, while the Belize District reported the highest number of infections, the rural Toledo District had the highest rate of dangerous ESBL-producing bacteria, identifying a resistance hotspot requiring targeted intervention. These findings compel Belize to urgently revise national treatment guidelines, enforce stricter regulations on over-the-counter antibiotic sales, and launch targeted public health campaigns in high-risk districts. Without immediate action, common infections may become increasingly difficult or impossible to treat with standard available antibiotics.

Graphical Abstract

A six-year analysis in Belize reveals a severe antimicrobial resistance crisis. Over 40% of urinary infections are resistant to first-line oral antibiotics, driven by ESBLs. Klebsiella pneumoniae poses the highest multidrug resistance threat. Resistance to key drugs accelerated after 2020. Alarmingly, most resistant infections originate in the community, not hospitals. Geospatial mapping uncovers distinct hotspots for infection burden and resistance intensity. This evidence calls for urgent updates to treatment guidelines and targeted public health action to safeguard effective antibiotics.

Keywords

antimicrobial resistance Belize community-acquired infections Enterobacterales multidrug resistance

Introduction

The relentless emergence and global dissemination of antimicrobial resistance (AMR) among Gram-negative pathogens constitute one of the most pressing public health crises of the 21st century.^1,2 Within the Enterobacteriaceae family, Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae are of paramount concern. These organisms, while often commensals of the human intestinal tract, are leading etiological agents of both community- and hospital-acquired infections, including urinary tract infections (UTIs), pneumonia, and bacteremia.^3,4 Their clinical threat is exponentially amplified by their remarkable genetic plasticity, enabling the rapid acquisition and spread of resistance genes, particularly those encoding extended-spectrum β-lactamases (ESBLs) and conferring multidrug resistance (MDR).⁵ The rise of MDR Enterobacterales undermines the efficacy of first- and second-line antibiotics, leading to prolonged illness, increased healthcare costs, and higher mortality rates, a burden felt most acutely in resource-limited settings.⁶

The AMR landscape remains critically under-characterized in Belize as well as other Latin America and the Caribbean countries with significant gaps in local surveillance data hindering effective national and regional response strategies.^7,8 Although Belize has established a national AMR committee, evidence-based studies to inform empirical treatment guidelines and stewardship programs are scarce. Existing regional data suggest alarmingly high rates of AMR. For instance, a Caribbean-wide study found 30% of K. pneumoniae isolates were ESBL producers, while ampicillin demonstrated negligible activity against E. coli.⁸ However, these regional aggregates may mask important national and sub-national epidemiological variations driven by local prescribing practices, healthcare access, and environmental factors.

A particularly concerning yet understudied dimension of the AMR epidemic is the shifting burden of resistance from predominantly hospital settings into the community. Traditionally, MDR Gram-negative bacteria were considered a nosocomial challenge. However, there is growing, albeit fragmented, evidence that ESBL-producing and MDR E. coli and K. pneumoniae are now endemic in many community settings worldwide, complicating the management of common infections like UTIs.⁹ The drivers, magnitude, and clinical implications of this community reservoir in low- and middle-income countries (LMICs) like Belize are poorly quantified. Without understanding the local epidemiology, including which patients are at highest risk for AMR infections, where AMR hotspots exist, and how resistance trends are evolving. Clinicians are left to prescribe empirically, potentially fueling further resistance.

This study, therefore, moves beyond conventional surveillance to address critical knowledge gaps. We aimed to determine the antibiotic susceptibility patterns of E. coli, K. pneumoniae, and E. cloacae in Belize from 2018 to 2023. Our investigation sought to quantify the true burden of MDR in both outpatient and inpatient settings, to identify the demographic and geographic predictors of infections likely to fail first-line empirical therapy, and to analyze temporal trends to pinpoint periods of significant change in resistance patterns. By leveraging a nationwide 6-year dataset, we aimed to generate actionable evidence to bridge the gap between raw susceptibility data and clinical decision-making. The findings are intended to directly inform Belize’s national AMR action plan, guide the development of context-specific empirical treatment algorithms, and highlight priority areas for targeted antimicrobial stewardship and infection prevention interventions, ultimately contributing to the global effort to curb the AMR crisis.

Materials and methods

Study design and setting

We conducted a nationwide, retrospective analysis of AMR using secondary data from the Belize Health Information System (BHIS). The study included all clinical specimens from which Escherichia coli, Klebsiella pneumoniae, or Enterobacter cloacae were isolated as the primary pathogen between January 1, 2018, and December 31, 2023, from public healthcare facilities across all six districts of Belize (Belize, Cayo, Corozal, Orange Walk, Stann Creek, and Toledo). Only the first isolate per patient was included to ensure independence of observations. Patient consent was not required for this study as only fully anonymized secondary data were analyzed.

Data source and management

De-identified and anonymized laboratory data were provided by the Epidemiology Unit of the Ministry of Health and Wellness, sourced from the Central Medical Laboratory (CML) via the BHIS. The raw dataset (4756) included patient age, gender, district of residence, specimen type, inpatient/outpatient status, organism identification, and antimicrobial susceptibility test (AST) results for a panel of 16 antibiotics. Data cleaning involved removing duplicate entries, records with missing critical fields (i.e., organism, specimen, or AST results), and isolates identified as secondary pathogens. After cleaning, the final analytical dataset comprised 4591 unique isolates.

Antimicrobial susceptibility was determined by disk diffusion or automated systems at the CML and interpreted according to the Clinical and Laboratory Standards Institute (CLSI) guidelines in effect at the time of testing. Extended-spectrum β-lactamase (ESBL) production was phenotypically confirmed using combined disk methods.

Inclusion and exclusion criteria

Inclusion criteria: (1) First isolate per patient of E. coli, K. pneumoniae, or E. cloacae from any clinical specimen; (2) Collection date between January 1, 2018, and December 31, 2023; (3) From any public healthcare facility in Belize; (4) Complete antimicrobial susceptibility testing (AST) results available for at least the antibiotics used in MDR classification.

Exclusion criteria: (1) Duplicate isolates (same patient, same organism, identical susceptibility profile within 30 days); (2) Records with missing critical fields (organism identification, specimen type, or AST results for ciprofloxacin and trimethoprim-sulfamethoxazole for urinary isolates); (3) Isolates identified as contaminants or secondary pathogens based on laboratory remarks; (4) Isolates from patients with repeat cultures for the same infection episode.

Variables and definitions

Multidrug resistance

An isolate was classified as MDR if it was non-susceptible to at least one agent in three or more of the following antimicrobial categories: fluoroquinolones (ciprofloxacin), folate pathway antagonists (trimethoprim-sulfamethoxazole), third-generation cephalosporins (ceftriaxone or cefotaxime), aminoglycosides (gentamicin), and β-lactam/β-lactamase inhibitors (ampicillin-sulbactam).

Empirical failure risk for UTIs

For isolates from urine specimens, a conservative Empirical failure risk (EFR) variable was created. An isolate was considered high-risk (EFR_UTI = 1) if it was resistant to both ciprofloxacin and trimethoprim-sulfamethoxazole. This represents the predicted failure of the two most common oral antibiotics used for empirical outpatient treatment of UTIs in Belize.

Community versus healthcare-associated

Isolates were conveniently categorized by the setting in which they were collected as “Outpatient” (community-associated) or “Inpatient” (healthcare-associated).

Statistical analysis

Descriptive statistics were reported as frequencies and percentages for categorical variables. Bivariate comparisons used chi-square tests. To identify independent predictors of high EFR for UTIs, a multivariable logistic regression model was constructed with EFR_UTI as the dependent variable. Independent variables included organism, age group (<50, 50–69, 70+ years), gender, district, year, ESBL status, and inpatient/outpatient setting. Results are reported as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Model fit was assessed using the Hosmer-Lemeshow test.¹⁰

Temporal trends in resistance percentages were analyzed using Joinpoint Regression (Joinpoint Software, Version 5.0.2),¹¹ which identifies points where a statistically significant change in the linear slope (Annual Percent Change, APC) occurs. A maximum of two joinpoints was allowed, and analyses used a log-linear Poisson model.¹²

District-level incidence rates (isolates per 100,000 population) and ESBL prevalence were calculated using 2022 population estimates from the Statistical Institute of Belize. All analyses were performed using R statistical software (Version 4.3.2) and SPSS (Version 28).¹³ Descriptive statistics, bivariate comparisons (chi-square tests), multivariable logistic regression, and the Hosmer–Lemeshow test were performed using SPSS Version 28. Joinpoint regression was performed using Joinpoint Software Version 5.0.2. District-level incidence calculations, ESBL prevalence estimation, and geospatial analyses were performed using R Version 4.3.2 (with packages sf, spdep, and ggplot2). A two-sided p-value < 0.05 was considered statistically significant.¹⁴

Sample size considerations

With 4591 isolates included in the final analytical dataset, we had sufficient power to detect clinically meaningful differences. For the primary analysis of dual resistance in urinary isolates (n = 3005), this sample size provides 80% power to detect an odds ratio of 1.3 for binary predictors with 20% prevalence at α = 0.05 (two-tailed). For subgroup analyses by district (six districts, ranging from 190–2247 isolates), power was more limited for smaller districts (e.g., Toledo: 190 isolates provide 80% power to detect OR ⩾ 2.0). For the multivariable logistic regression model with 11 predictor parameters, the events-per-variable ratio was approximately 30:1 (1247 events of dual resistance), exceeding the recommended minimum of 10:1. For the joinpoint regression analysis, the 6-year time series with annual data points provided sufficient observations to detect one joinpoint with 80% power given the observed effect size (APC change from 1.2% to 9.8%).

The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. The completed STROBE checklist is provided as Supplemental File 1.

Results

Cohort characteristics and descriptive epidemiology

From 2018 to 2023, 4591 isolates (E. coli: n = 3105, 67.6%; K. pneumoniae: n = 1174, 25.6%; E. cloacae: n = 312, 6.8%) met inclusion criteria. Most isolates came from female patients (68.7%) and outpatient settings (92.5%). Urine was the predominant specimen source (65.5%), followed by wound/abscess (17.1%). The highest burden of isolates was from the Belize District (n = 2247, 48.9%), which also had the highest population-adjusted incidence rate (1812.1 per 100,000). Patients aged 50–69 years represented the largest proportion of isolates (34.6%).

The empirical treatment gap for community-acquired UTIs

Among 3005 urinary isolates, 1247 (41.5%) were resistant to both ciprofloxacin and trimethoprim-sulfamethoxazole, indicating a high risk of failure for standard empirical oral therapy. In multivariable analysis, significant independent predictors of this high-risk profile were infection with K. pneumoniae (aOR 2.85, 95% CI 2.32–3.51) or E. cloacae (aOR 4.12, 95% CI 2.78–6.11) compared to E. coli, age 50+ years, ESBL production (aOR 8.91, 95% CI 7.40–10.73), inpatient acquisition (aOR 1.78, 95% CI 1.35–2.34), and isolation in the year 2023 compared to 2018 (aOR 1.43, 95% CI 1.13–1.81; Table 1).

Table 1.

Independent predictors of dual resistance to first-line oral antibiotics for urinary tract infections in Belize (2018–2023).

Predictor	Level	Adjusted odds ratio	95% CI	p-Value
Organism	K. pneumoniae (vs E. coli)	2.85	2.32–3.51	<0.001
	E. cloacae (vs E. coli)	4.12	2.78–6.11	<0.001
Age group	50–69 (vs <50)	1.31	1.11–1.55	0.001
	70+ (vs <50)	1.52	1.23–1.88	<0.001
ESBL	Positive (vs Negative)	8.91	7.40–10.73	<0.001
Setting	Inpatient (vs Outpatient)	1.78	1.35–2.34	<0.001
District	Cayo (vs Belize)	0.82	0.68–0.99	0.036
Year	2023 (vs 2018)	1.43	1.13–1.81	0.003

ESBL, extended-spectrum β-lactamase.

Temporal trends in key resistance markers

Joinpoint regression analysis revealed a significant change in the trajectory of ciprofloxacin resistance among E. coli isolates. From 2018 to 2020, the resistance rate was stable (APC 1.2%, 95% CI −10 to 14%, p = 0.8). A joinpoint was identified in 2020, after which resistance increased significantly at an annual rate of 9.8% (95% CI 5–15%, p = 0.02) through 2023 (Figure 1). A consistent, significant upward trend in ESBL production among K. pneumoniae isolates was also observed over the entire period (APC 4.5%, 95% CI 2.1–7.0%, p = 0.01).

Figure 1.

Joinpoint regression trend for ciprofloxacin resistance in E. coli, Belize (2018–2023).

Geospatial heterogeneity in incidence and resistance

Analysis of district-level epidemiology revealed distinct and dissociated patterns of infection burden and resistance (Figure 2 and Table 2). The Belize District had the highest isolate incidence rate (1812.1 per 100,000), serving as the primary testing and referral hub. In contrast, the Toledo District, despite a lower overall incidence (500.0 per 100,000), had the highest prevalence of ESBL-producing isolates (29.5%) among all districts, significantly higher than the national average of 23.3% (p = 0.02). This identifies Toledo as a distinct resistance hotspot despite not being an incidence hotspot.

Figure 2.

District-level heterogeneity in infection burden and resistance intensity. (a) Choropleth diagram of Belize showing the incidence rate (isolates per 100,000 population) of E. coli, K. pneumoniae, and E. cloacae across the six districts for 2018–2023. (b) Choropleth plot showing the prevalence of ESBL production among isolates across districts. The Belize District is the primary incidence hotspot, while the Toledo District is identified as a distinct resistance (ESBL) hotspot. Base diagram data from the Statistical Institute of Belize.

Table 2.

District-level epidemiology of bacterial isolates and ESBL production.

District	Population (Est.)	Total isolates (N)	Incidence (per 100k)	ESBL+ isolates (N)	ESBL prevalence (%)
Belize	124,000	2247	1812.1	552	24.6
Cayo	105,000	912	868.6	178	19.5
Corozal	49,000	408	832.7	92	22.5
Orange Walk	53,000	323	609.4	67	20.7
Stann Creek	45,000	511	1135.6	131	25.6
Toledo	38,000	190	500.0	56	29.5

ESBL, extended-spectrum β-lactamase.

The outpatient multidrug resistance burden

MDR was prevalent across all settings but was notably high in the community. Among outpatient isolates, MDR prevalence was 42.1% for E. coli, 61.2% for K. pneumoniae, and 51.4% for E. cloacae. As expected, MDR rates were significantly higher in inpatient isolates for each organism (58.9%, 74.0%, and 65.6%, respectively; p < 0.001 for all comparisons; Figure 3). ESBL production followed a similar gradient, being the primary driver of MDR, particularly for K. pneumoniae (outpatient ESBL rate: 38.5%).

Figure 3.

Prevalence of MDR and ESBL production by clinical setting and organism.

Stacked bar charts comparing the percentage of isolates classified as multidrug resistant (MDR, defined as non-susceptibility to ⩾3 drug categories) and the percentage confirmed as ESBL producers, stratified by bacterial species (E. coli, K. pneumoniae, E. cloacae) and patient setting (outpatient vs inpatient). Error bars represent 95% CIs for the proportions. Asterisks denote significant differences (p < 0.001) in MDR and ESBL rates between outpatient and inpatient isolates for each organism, as determined by chi-square tests.

Discussion

This nationwide, 6-year analysis provides an objective and novel assessment of the epidemiology of AMR among key Enterobacterales in Belize. Our findings paint a concerning picture of a high and evolving burden of resistance, characterized by a substantial community reservoir of MDR pathogens, a critical gap in effective empirical therapy for common infections, and distinct spatial and temporal patterns that demand targeted public health action. The data robustly support our alternative hypothesis, confirming that E. coli, K. pneumoniae, and E. cloacae isolated in Belize are predominantly multidrug resistant.

The most critical finding is the scale of the empirical treatment gap for community acquired UTIs. With over 40% of urinary isolates resistant to both first line oral agents (ciprofloxacin and trimethoprim sulfamethoxazole), this resistance pattern signals a crisis in outpatient management (Table 1). This rate far exceeds those reported in many surveillance studies from higher-income settings and underscores the precarious state of empirical therapy in Belize.⁹ Our multivariable model elucidates the drivers of this gap: ESBL production was the strongest predictor, increasing the odds of dual resistance nearly ninefold. This aligns with the global understanding of ESBLs as master switches conferring co-resistance to multiple drug classes.¹⁵ Notably, the organism itself was a major independent risk factor, with K. pneumoniae and E. cloacae posing a significantly higher threat than E. coli. This has immediate clinical implications, suggesting that knowledge of the causative organism, if rapidly available, should heavily influence empiric choices. Furthermore, the significantly higher risk associated with the year 2023 compared to 2018 indicates a worsening trend, a progression that threatens to render current empirical guidelines obsolete.

The empirical treatment gap documented in Belize mirrors findings from other resource-limited settings. In a recent study from Nepal, Thapa et al.¹⁶ reported that E. coli and K. pneumoniae were the predominant uropathogens in a tertiary care hospital, with amoxicillin resistance exceeding 95% for both species, comparable to the near-complete loss of oral penicillin efficacy observed in Belize. Notably, the Nepalese study found preserved susceptibility to gentamicin (7.4% resistance in E. coli) and nitrofurantoin (12.2% resistance), aligning with our observation that nitrofurantoin retains utility for cystitis in Belize. However, important differences emerged: The Nepalese study reported lower ciprofloxacin resistance (32.7% vs our observed rates of approximately 33%–37% depending on year) and substantially higher amikacin resistance (50.0% vs <15% in Belize), highlighting the necessity of locally derived susceptibility data rather than extrapolation from geographically distant settings. Both studies underscore a shared challenge across low- and middle-income countries: the erosion of oral options for outpatient UTI management, compelling reliance on intravenous agents or last-resort oral alternatives like nitrofurantoin.¹⁶

The temporal trend analysis provided statistical rigor to this observation of worsening resistance. The joinpoint regression identified 2020 as a significant inflection point for ciprofloxacin resistance in E. coli, marking a shift from a stable, albeit high, prevalence to a period of significant annual increase (Figure 1). This chronological correlation with the COVID-19 pandemic is highly suggestive of a causal relationship, likely mediated by increased and often inappropriate antibiotic prescribing observed globally during the pandemic, both in healthcare settings and through self-medication.^17–19 This finding underscores how broad healthcare shocks can accelerate AMR trends and highlights the need for reinforced stewardship during and after such crises.

Our analysis challenges the traditional hospital-centric view of MDR by quantifying the substantial outpatient burden. While inpatient isolates, as expected, exhibited higher MDR and ESBL rates, confirming the healthcare system as an amplifier of resistance, the baseline community rates were also substantially high (Figure 3). Among the pathogens in our study, K. pneumoniae, the most problematic, had 61% of outpatient isolates MDR. This establishes that the community in Belize serves as a major reservoir for resistant pathogens, meaning that MDR infections are being imported into hospitals rather than solely generated within them. This shifts the intervention priority toward community-focused stewardship, regulation of over-the-counter antibiotic sales, and public education.

Complementing our findings on Enterobacterales, a parallel nationwide analysis of P. aeruginosa in Belize documented escalating multidrug resistance, with MDR prevalence increasing from 28.3% in 2018 to 41.8% in 2024, an average annual increase of 2.3%.²⁰ That study also identified significant spatial clustering of MDR P. aeruginosa in urban districts (adjusted odds ratio 2.32 for urban vs rural), mirroring the geographic heterogeneity we observed for ESBL-producing Enterobacterales. The convergence of resistance trends across different pathogen families suggests common drivers, including antibiotic selection pressure and healthcare-associated transmission, reinforcing the need for comprehensive, multi-pathogen surveillance and stewardship interventions in Belize.

The geospatial analysis revealed another layer of complexity, identifying a dissociation between the burden of disease and the intensity of resistance (Figure 2 and Table 2). The Belize District, home to the national referral hospital, had the highest incidence of isolates, reflecting centralization of diagnostic testing and care. In contrast, the Toledo District emerged as a clear hotspot of resistance, with the highest ESBL prevalence despite a lower overall incidence. This pattern suggests that drivers of resistance in Toledo may differ from those in urban centers, potentially relating to agricultural antibiotic use, differences in community prescription practices, or distinct local transmission networks. While this study cannot determine causality, the higher ESBL prevalence in a district with more rural communities warrants investigation into potential environmental drivers, such as contamination of water sources or food chains, or differences in access to and use of specific antibiotic classes. This finding argues against a one-size-fits-all national AMR strategy and instead supports the need for targeted, district-specific interventions.

The high prevalence of ESBL-producing Enterobacterales in wound specimens (35.9% for E. coli, 42.6% for K. pneumoniae in our study) aligns with a separate Belizean analysis of surgical site and wound infections, which reported that cesarean sections were associated with a 2.4-fold increased odds of ESBL infection compared to other abdominal procedures (adjusted OR 2.4, 95% CI 1.3–4.5), and that diabetic amputations had a 3.1-fold increased odds of methicillin-resistant S. aureus.²¹ These procedure-specific risk patterns, identified across independent datasets, provide a stronger evidence base for targeted perioperative antibiotic prophylaxis and enhanced infection prevention measures in high-risk surgical populations in Belize.

Recommendations

Based on the findings of this study, we propose the following targeted, evidence-based recommendations to strengthen Belize’s national response to antimicrobial resistance:

Revise National Empirical Treatment Guidelines: The high prevalence of dual resistance to ciprofloxacin and trimethoprim/sulfamethoxazole in urinary isolates (41.5%) necessitates an urgent update to national UTI treatment guidelines. For adults with suspected Gram-negative UTI in outpatient settings, nitrofurantoin (given its high retained susceptibility in E. coli) should be strongly prioritized for cystitis. For cases requiring broader coverage or when K. pneumoniae or E. cloacae is suspected, guidelines should recommend a third-generation cephalosporin as the default, with clear instructions to reassess therapy based on culture and susceptibility results.

Implement Differentiated District-Level AMR Action Plans: The identification of Toledo District as an ESBL hotspot, despite lower healthcare access, warrants targeted investigation and intervention. The national AMR Committee should support district health teams in Toledo to conduct focused surveys on local antibiotic prescribing and dispensing practices, agricultural use, and water and sanitation. Public health campaigns on antimicrobial stewardship should be tailored and intensified in this region.^19,22,23

Strengthen Antimicrobial Stewardship Programs (ASPs) with a Community Focus: Given the substantial burden of MDR in outpatient isolates, ASPs must extend beyond hospital walls.^19,20 The Ministry of Health and Wellness should:

i. Develop and enforce stricter regulations on over-the-counter sales of key antibiotics like ciprofloxacin and cephalosporins.

ii. Launch national public awareness campaigns on the dangers of antibiotic misuse and the importance of completing prescribed courses.

iii. Provide continuing medical education to community-based healthcare providers on the local resistance patterns revealed in this study, emphasizing the high risk associated with K. pneumoniae and ESBL producers.

Enhance Laboratory-Based Surveillance and Feedback: The BHIS and CML should be leveraged to create regular (e.g., quarterly) antibiogram reports at the national and district levels. These reports should be disseminated to all prescribers and should highlight the empirical treatment gap metrics. Furthermore, investment in rapid diagnostic tests (e.g., for ESBL detection) at regional hospitals could facilitate more precise, same-day treatment decisions.

Strategies for small island nations: Translating evidence into regional action

The challenges illuminated by this Belizean analysis are not unique; they reflect a broader crisis confronting small island developing states (SIDS) across the Caribbean and beyond. These nations share common vulnerabilities, limited healthcare infrastructure, constrained laboratory capacity, high tourism-related antibiotic consumption, and the challenge of implementing comprehensive antimicrobial stewardship with limited human resources. However, their small size and interconnected health systems also present unique opportunities for coordinated action. The findings from Belize offer a blueprint for how other small island nations can confront the AMR crisis through contextually appropriate, scalable strategies.

First, small island nations must prioritize establishing standardized electronic laboratory surveillance systems that generate real-time resistance data. Belize’s Central Medical Laboratory model, despite its resource constraints, demonstrates that centralized testing with nationwide data aggregation is feasible in small-island contexts. Other SIDS should invest in strengthening a single reference laboratory capable of producing high-quality, standardized susceptibility data, with simplified electronic reporting systems that feed directly into national and regional surveillance platforms. This approach maximizes limited technical expertise while ensuring data comparability across islands, enabling early detection of emerging resistance patterns before they become entrenched.

Second, the substantial community burden of MDR documented in this study compels small island nations to extend antimicrobial stewardship beyond hospital walls. In settings where formal healthcare systems may reach only a portion of the population, community engagement becomes paramount. Island nations should develop culturally tailored public awareness campaigns that address local beliefs about antibiotics and leverage community health workers, religious institutions, and schools as dissemination channels. Given the high rates of outpatient antibiotic resistance, regulations restricting over-the-counter antibiotic sales must be enforced through multisectoral collaboration among health ministries, pharmacy boards, and consumer protection agencies.²² Some islands have successfully implemented “antibiotic guardian” programs that enlist community members as stewards of antibiotic effectiveness, a model worth replicating regionally.

Third, the geospatial heterogeneity observed in Belize underscores that even within small nations, resistance patterns vary meaningfully. SIDS should adopt differentiated intervention strategies that recognize sub-national variation in resistance burdens. Districts or islands with high ESBL prevalence, like Toledo, warrant targeted investigations into local drivers, whether agricultural runoff, water quality, or prescribing practices, followed by tailored responses. This precision approach ensures that limited public health resources are directed where they are most needed, rather than diluted across broad, one-size-fits-all campaigns.

Fourth, small island nations must leverage their regional platforms to amplify individual efforts. The Caribbean Public Health Agency (CARPHA) and the Organization of Eastern Caribbean States provide existing mechanisms for harmonizing treatment guidelines, aggregating resistance data, and coordinating procurement of essential antibiotics. Belize’s finding that over 40% of urinary isolates fail first-line oral therapy should catalyze regional dialogue about revising empirical treatment algorithms across the Caribbean. Joint procurement arrangements could ensure consistent access to second-line agents like nitrofurantoin and fosfomycin, while regional laboratory networks could facilitate external quality assurance and technology transfer.

Finally, small island nations must embed AMR surveillance within broader health system strengthening initiatives. The post-2020 acceleration of ciprofloxacin resistance in Belize highlights how external shocks, pandemics, natural disasters, and economic crises can rapidly worsen resistance patterns. SIDS, which face disproportionate climate-related and economic vulnerabilities, should integrate AMR monitoring into disaster preparedness frameworks to ensure that antibiotic use during crises remains judicious and that surveillance systems withstand disruptions. Investing in rapid diagnostic technologies appropriate for resource-limited settings, such as low-cost platforms for ESBL detection, can empower clinicians at district hospitals and health centers to make evidence-driven prescribing decisions.

The path forward for Belize and its island neighbors lies not in blind emulation of high-income country strategies but in harnessing their unique characteristics, small populations, connected health systems, and strong community networks to build nimble, responsive AMR containment programs. By translating local data into regional action, small island nations can collectively confront the silent pandemic of antimicrobial resistance, preserving the efficacy of life-saving antibiotics for future generations.

Limitations

While this study provides critical insights, several limitations should be considered. First, as a retrospective analysis of laboratory records, it lacks detailed clinical data, including patient comorbidities, treatment histories, antibiotic exposure prior to culture, and clinical outcomes (e.g., treatment failure, mortality). This prevents us from linking specific resistance profiles to patient-level health impacts. Second, the data are sourced exclusively from the public health sector; isolates from private clinics and hospitals are not included, which may limit the generalizability of the findings to the entire population of Belize. Third, although laboratory methods were standardized at the CML, subtle changes in procedures or interpretive criteria over the 6-year study period could introduce non-biological variation in susceptibility results. Finally, the geospatial analysis used district-level population estimates, which may mask finer-scale heterogeneity (e.g., urban vs rural differences within a district) in the drivers of resistance.

Conclusion

This 6-year nationwide analysis elucidates a critical and evolving antimicrobial resistance crisis in Belize, moving beyond simple surveillance to reveal the drivers, distribution, and consequences of resistance among key Enterobacterales. We demonstrate that multidrug resistance, particularly driven by ESBL production, is deeply entrenched not only in hospitals but as a prevalent feature of community-acquired infections. The significant empirical treatment gap for UTIs, the marked acceleration of fluoroquinolone resistance post-2020, and the identification of distinct geographic hotspots for infection burden and resistance intensity collectively provide a powerful evidence base for action.

The findings compel a paradigm shift in the national AMR strategy—from a focus predominantly on hospital infection control to a dual approach that equally prioritizes community-based stewardship and regulation. This can only be achieved through the immediate implementation of updated, data-driven treatment guidelines, targeted interventions in high-risk regions such as the Toledo District, and a sustained commitment to strengthening surveillance and public education. By translating these epidemiological insights into policy and practice, Belize can more effectively confront the threat of AMR, safeguarding its population’s health and contributing to the global fight against this silent pandemic.

Supplemental Material

sj-docx-1-tai-10.1177_20499361261458469 – Supplemental material for Treatment gaps and risk factors for multidrug-resistant Enterobacterales in Belize: a 6-year (2018–2023) nationwide retrospective analysis

Supplemental material, sj-docx-1-tai-10.1177_20499361261458469 for Treatment gaps and risk factors for multidrug-resistant Enterobacterales in Belize: a 6-year (2018–2023) nationwide retrospective analysis by Innocent Nwachukwu, Camryn Linares, Griseidy Alcoser, Nivi Novelo and Danladi C. Husaini in Therapeutic Advances in Infectious Disease

Footnotes

Acknowledgements

The authors gratefully acknowledge the Belize Ministry of Health and Wellness for their invaluable support and collaboration. We extend our sincere thanks for granting us access to the Belize Health Information System (BHIS) and for providing the comprehensive national dataset that formed the foundation of this study. We are particularly indebted to the Epidemiology Unit and the staff of the Central Medical Laboratory for their technical assistance and expertise. This research would not have been possible without their commitment to public health surveillance and their dedication to advancing the understanding of antimicrobial resistance in Belize.

Declarations

ORCID iDs

Innocent Nwachukwu

Danladi C. Husaini

Supplemental material

Supplemental material for this article is available online.

References

World Health Organization. Antimicrobial resistance [Internet], https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance (2023, accessed 18 November 2025).

Marino

Maniaci

Lentini

, et al. The global burden of multidrug-resistant bacteria. Epidemiologia (Basel) 2025; 6(2): 21.

Mahon

Lehman

DC.

Textbook of diagnostic microbiology–E-Book. 6th ed. Elsevier Health Sciences, St. Louis, MO, 2018.

Bitew

Dagnew

Dereje

, et al. Burden of multi-drug resistant bacterial isolates and its associated risk factors among UTI-confirmed geriatrics in Gondar town. Sci Rep 2025; 15: 14270.

Nwobodo

Ugwu

Oliseloke Anie

, et al. Antibiotic resistance: The challenges and some emerging strategies for tackling a global menace. J Clin Lab Anal 2022; 36(9): e24655.

Nwobodo

Ugwu

Anie

, et al. Antibiotic resistance: the challenges and some emerging strategies for tackling a global menace. J Clin Lab Anal 2022; 36(9): 24655.

Husaini

Kishan

, et al. Field-based screening of selected oral antibiotics in Belize. Plos One 2020; 15(6): e0234814.

Nagassar

Daniel

Peters

, et al. Antimicrobial resistance in the Caribbean. Caribb Med J 2022; 87(2).

Lee

Choe

Community-acquired urinary tract infection by Escherichia coli in the era of antibiotic resistance. Biomed Res Int 2018; 2018: 1–14.

10.

Surjanovic

Loughin

TM.

Improving the Hosmer-Lemeshow goodness-of-fit test in large models with replicated Bernoulli trials. J Appl Stat 2023; 51(7): 1399–1411.

11.

National Cancer Institute at the NIH. (n.d.). Joinpoint Trend Analysis Software. Joinpoint Regression Program. https://surveillance.cancer.gov/joinpoint/

12.

Kim

Fay

Feuer

, et al. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med 2001; 20: 655.

13.

R Project . (n.d.). The R project for statistical computing. R. https://www.r-project.org/

14.

R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/.

15.

Castanheira

Simner

Bradford

PA.

Extended-spectrum β-lactamases: an update on their characteristics, epidemiology and detection. JAC-Antimicrob Resist 2021; 3(3): dlab092.

16.

Thapa

Shrestha

Adhikari

, et al. Antibiotic resistance patterns in uropathogens: insights from a Nepalese tertiary care setting. Therap Adv Infect Dis 2025; 12: 1–7.

17.

Sulayyim

HJA

Ismail

Hamid

, et al. Antibiotic Resistance during COVID-19: a systematic review. Int J Environ Res Public Health 2022; 19(19): 11931.

18.

Nandi

Pecetta

Bloom

DE.

Global antibiotic use during the COVID-19 pandemic: analysis of pharmaceutical sales data from 71 countries, 2020–2022. EClinicalMedicine 2023; 57: 101848.

19.

Husaini

Parchue

Orellana-Erazo

, et al. Post-COVID-19 medication adherence among HIV/AIDS patients in Belize: a case for consolidating education and monitoring. AIDS Res Treat 2025; 2025(1): 5552340.

20.

Nwachukwu

Ico

Itza

, et al. Escalating burden and spatial clustering of multidrug-resistant Pseudomonas aeruginosa in Belize: a nationwide retrospective analysis, 2018-2024. Microbe 2026; 11: 100696.

21.

Nwachukwu

Rivera

Reyes

, et al. Antimicrobial resistance in surgical site and wound infections in Belize: a seven-year national surveillance analysis, 2018-2024. MicrobiologyOpen 2026; 15(3): e70312.

22.

Husaini

Abubakar

Harris-Thurton

, et al. Over-the-counter sales of antibiotics at community pharmacies in Belize: a cross-sectional simulated client visit study. Integrative J Med Sci 2021; 8: 1-5.

23.

Husaini

Bolon

Smith

, et al. Quality of medicines and public health in Latin America and the Caribbean (LAC): a review. Int J Pharm Healthcare Marketing 2023; 17(4): 654–674.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB