IntroductionMycobacterium tuberculosis has plagued humanity for centuries. Recognized from popular culture and known by many names throughout history, tuberculosis (TB) is a disease that has left an indelible mark on society and shaped medical history. Over time, countless obscure and often unsuccessful attempts to cure it have been documented [1]. Despite this long and winding history, TB remains a feared but curable disease that, in theory, could be eliminated. Nevertheless, TB continues to be the leading infectious killer worldwide, with a record high number of cases reported by the World Health Organization (WHO) in 2023 [2]. Yet, paradoxically, treatment of drug-susceptible TB has remained largely unchanged for decades. The foundation of today’s standard TB treatment was laid through landmark clinical trials beginning in the 1940s and continuing through the 1980s [1]. From the first randomized controlled trial of streptomycin in 1947 to the development of the modern four-drug regimen—isoniazid, rifampicin, pyrazinamide, and ethambutol—these studies shaped not only the treatment of TB but also the broader methodology of clinical research [1]. Together, they represent a triumph in public health. But when we look back at these studies in today’s context, it is hard not to ask: could these trials that defined modern TB treatment even be conducted today? To explore this, we revisited five landmark trials that contributed to the conceptual development of the standard four-drug TB regimen, from the earliest randomized trial to more recent short-course combination studies [[3], [4], [5], [6], [7]].Historical clinical trials defining TB treatmentThe randomized trial of streptomycin monotherapy conducted in 1947, widely regarded as the beginning of modern clinical trials, not only demonstrated the drug’s ability to cure TB but also revealed the rapid emergence of resistance, prompting the search for companion therapies [3]. By 1948, combination therapy with streptomycin and para-aminosalicylic acid (PAS) was shown to improve cure rates and prevent resistance [4]. The addition of isoniazid significantly enhanced efficacy, leading to a triple regimen (streptomycin, PAS, and isoniazid) but requiring up to 24 months of treatment [5]. In the 1960s, ethambutol was introduced as a safer alternative to PAS, forming a regimen of streptomycin, isoniazid, and ethambutol, typically given for 18 months [7]. The introduction of rifampicin in the late 1960s was a milestone. The British Medical Research Council (MRC) trials in the 1970s demonstrated that combinations of streptomycin, isoniazid, rifampicin, and pyrazinamide could reduce treatment duration to 6 months [6]. This historical trajectory, described in depth elsewhere [1,8,9], led to the now globally used, fully oral, four-drug, 6-month regimen for drug-susceptible TB.Ethical and methodological reflectionsOur intention in revisiting some of these landmark trials is not to diminish their importance, but to examine their designs and reporting through the perspective of contemporary research standards. Among the five reviewed trials, informed consent was not mentioned. Strikingly, the early trials explicitly stated: “Patients were not told before admission that they were to get special treatment. C [control] patients did not know throughout their stay in hospital that they were control patients in a special study” and “Patients were not told that they were taking part in a special investigation” [3,4]. There was no mention of ethical review—unsurprising, perhaps, since these studies were conducted before research ethics committees became standard. Still, potential risks to study participants were very sparsely discussed in the trial method sections. Although adverse effects and toxicities were reported as part of some of the trials’ findings and discussions, prestudy considerations of patient risks and prospective safety precautions, such as predefined criteria for stopping treatment or rescue interventions, seemed minimal or absent—measures we would now consider mandatory. While the risk–benefit ratio in these trials was generally presented as favourable, and most likely was, given the high mortality of untreated TB and the promise of novel treatments, methodological differences in comparison to modern studies could be observed. Primary endpoints were mostly unclearly defined or not explicitly stated, while modern trials, in contrast, typically use clearly defined endpoints such as culture conversion, survival, or WHO-defined outcomes. Placebo controls were not employed, with trials generally using active comparators—most likely due to the high mortality of untreated TB. Nevertheless, trials employed blinded evaluations of chest radiographs for outcome assessment, seemingly limited to this outcome. Randomisation was used across all the trials, with allocation concealment implemented in several, including sealed envelopes or centralized Randomisation via telegram, although detailed reporting was limited. Similarly, intention-to-treat analyses were not applied, and patients lost to follow-up or who deviated from the trial protocol were often excluded from analyses (dropout or exclusion rates approaching 30% in some studies) [6,7]—a practice that would now be considered a high risk of biasing the results. The statistical methods used in the trials were predominantly descriptive, comparing group-level outcomes without sample size justifications, confidence intervals, or formal inferential statistics, which are now considered essential. Not so surprisingly, data privacy, transparency, or trial registration was not mentioned. Although documentation of the trial protocol, participant safeguarding protocols, as well as other supporting documents, may have existed at the time or in concurrent publications, this reflects the research standards of the era. In terms of generalisability, earlier trials recruited inpatients from sanatoria or hospitals under specific settings, limiting applicability to community settings. Later studies, however, appeared to reflect more diverse, everyday settings, offering insights for public health implications. Across all the trials, vulnerable populations (adolescents, patients with severe TB, and potentially economically or socially disadvantaged) were included, though rarely acknowledged as such, and without explicit protections that would now be ethically mandated.From past to present: implications for modern TB trialsWe do not suggest that these studies were poorly conducted, but rather that they reflect the standards of their time (Table 1). Additionally, our reflections above raise a paradox: in trying to protect vulnerable patients today, do we inadvertently exclude them from research and thus from the potential benefits of innovative research? Should well-designed observational studies bridge this gap, or should we rethink trial inclusivity? The disconnect between highly regulated, complex trial designs and the realities of clinical care risks excluding patients who reflect the diversity and complexity of routine clinical settings. Historical TB trials included patients with severe disease and, seemingly, also vulnerable patients, groups often excluded in modern studies due to strict eligibility criteria. While these early trials lacked today’s ethical safeguards, their broad inclusion and operational simplicity may have enhanced generalisability and public health impact. Calls have been made for the inclusion of vulnerable populations, such as pregnant women, in contemporary TB trials [10]. Similar developments have already occurred in HIV research, which may serve as a role model. However, the risk-benefit considerations for including such populations must be carefully differentiated based on the specific disease states of TB (e.g. TB infection vs. TB meningitis), as ethical and clinical implications can vary substantially. Recent TB trials have already begun to re-embrace these principles by enrolling more diverse patient populations and adopting pragmatic designs, demonstrating that ethical rigour can be achieved without compromising feasibility or relevance, especially in a disease so strongly influenced by social and structural factors. The endTB trial included adolescents, individuals with HIV (regardless of CD4 count) and substance use disorders, and retained participants who became pregnant, groups often excluded from earlier TB studies [11]. Considering other methodological constraints, the recent Nix-TB trial, an open-label, single-group study of ‘only’ 109 patients with refractory multidrug-resistant (MDR) and extensively drug-resistant TB, has also started a revolutionary change in the treatment of TB, despite its relatively small sample size to modern standards [12]. Another contemporary trial on TB meningitis, which included highly vulnerable and severely ill patients, has also strongly influenced modern clinical practice [13]. While early TB trials used fixed designs, evaluating one or a few treatments at a time in a slow, step-by-step process, modern trials, such as a recently proposed global platform trial for TB meningitis or PARADIGM4TB, adopt adaptive, pragmatic designs that test multiple treatments concurrently [14]. These designs enable adjustments based on interim results, allowing for the rapid and efficient generation of evidence to inform clinical care. Yet despite these advances, the standard regimen has proven remarkably durable. The current 6-month ‘HRZE’ regimen (isoniazid, rifampin, pyrazinamide, and ethambutol) remains the most effective treatment from a public health standpoint, yielding consistently low relapse rates in patients with drug-susceptible TB. Emerging concepts include stratified treatment approaches based on disease phenotype (e.g. ‘easy-to-treat’ vs. ‘hard-to-treat’ TB) or even fully personalised therapy [15]. Nevertheless, the current ‘one-size-fits-all’ regimen has set a high threshold for new regimens to surpass. Whether this is due to the foundational trials, the inherent superiority of the regimen itself, or perhaps the fact that it simply has not been sufficiently challenged, remains difficult to determine.Table 1. Evolution of clinical trial standardsa.Methodological componentsHistorical trialsContemporary standardsProtocol designAssumably often informally and only internally documentedPublicly accessible protocol and SAP; preregistration requiredFunding disclosureNot disclosed, without conflict statementMandatory funding source and COI disclosureEthical reviewNo formal oversightMandatory review by IRB/RECInformed consentAbsent or implied; rarely documentedWritten, voluntary, documented, and revocableEligibility criteriaBroad, sometimes poorly definedPrecisely defined, includes rationale for inclusion/exclusionPopulation equityVulnerable groups often included without safeguardsSpecial protections and justification for patient inclusionRandomization and allocationSimple randomization with unclear allocation concealmentCentralized (electronic) randomization with full concealmentBlindingSeldom implemented or formally describedDouble/triple blinding when feasibleSample size calculationNot reportedPredefined with power justificationIntervention standardizationDrug regimens consistent, but delivery protocols variedGCP standards across sitesMonitoring and safety oversightNo DSMB or stopping rulesIndependent DSMB, stopping rules, SAE monitoringData collection and managementPaper-based, high risk of bias or missing dataElectronic data capture, audit trails, predefined CRFsStatistical analysisDescriptive only, no CIs or formal hypothesis testingITT and per-protocol analysis, CIs, adjusted models, pre-specificationTrial registrationNot requiredMandatory preregistration (e.g. ClinicalTrials.gov, CTIS/EudraCT)Transparency and reportingSelective publication, less stringent peer review(s) processesCONSORT guidelines, full results in registry, open-access encouragedData sharingNot requiredIPD sharing plans increasingly requiredPatient and public involvementNot requiredIncreasing integration in trial design, ethics, disseminationPosttrial accessNot requiredRequired by funders and ethics boards for ongoing treatment provisionCOI, conflict of interest; CONSORT, Consolidated Standards of Reporting Trials; CRF, case report form; CTIS, Clinical Trials Information System, DSMB, data and safety monitoring board; EudraCT, European Union Drug Regulating Authorities Clinical Trials Database; GCP, good clinical practice; IPD, individual participant data; IRB, institutional review board; ITT, intention-to-treat; REC, research ethics committee; SAE, serious adverse event; SAP, statistical analysis plan.aThe authors drafted this table as a comparative example to illustrate evolving standards in clinical trial methodology. For detailed and up-to-date trial conduct standards, readers are referred to regulatory and methodological resources, such as the International Conference on Harmonization Good Clinical Practice (ICH-GCP) guidelines, the CONSORT statement, the Clinical Trials Regulation, and the WHO’s International Standards for Clinical Trial Registries, and among others.The historical trials offer food for thought. Their pragmatic simplicity and focus on operational feasibility yielded results that were not only impactful but also scalable. Despite limited statistical tools and no modern infrastructure or technology, these studies provided clear answers to important clinical questions—something modern research often struggles to do. The trials also demonstrated the possibilities of multinational collaboration and included patients most affected by TB, without narrowing eligibility to the healthiest few. The vision of drug development has evolved with not-for-profit drug developers (e.g. TB Alliance), cross-sector (public–private partnerships), and global consortia (e.g. UNITE4TB or PanACEA) leading major clinical trials, while the MRC trials, in contrast, were initiated and conducted by a single national organisation. While methodology and ethical standards were undoubtedly different, the public health motivation was clear and urgent. As medical standards and research practices evolve, so do our expectations of study participant autonomy and safety, research transparency, and the results and impacts on a patient and societal level.ConclusionLooking back, we must ask ourselves: what lessons do these pioneering studies still hold, and how much have we moved forward? In an era of heightened regulatory oversight, ethical scrutiny, and sophisticated trial designs, how do we balance scientific ambition and the responsibility to protect those most at risk? These questions are particularly pressing in TB research, where global disease burden intersects with poverty, stigma, and structural inequality. Encouragingly, contemporary TB trials show that building truly modern regimens is not just possible, but happening. Many landmark TB trials, past and present, have been investigator-initiated, ranging from the MRC studies to recent advances such as Study 31, Nix-TB, TB-PRACTECAL, and endTB. These studies have significantly influenced practice, despite limited commercial backing, while manufacturers may have later benefited from the widespread adoption of the drugs. This highlights the crucial role of publicly funded, academically driven research in diseases with high global burden but limited market incentives. Reflecting on the past shows the progress we have made, and the work that still lies ahead to deliver the long-overdue innovation that patients with TB deserve.CRediT authorship contribution statementVictor Naestholt Dahl: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Conceptualization. Nils Wetzstein: Writing – review & editing, Methodology, Investigation, Conceptualization.Declaration of competing interestThe authors declare that they have no conflicts of interest.Financial reportNot applicable.Data availabilityNot applicable.AcknowledgementsWe express our gratitude to all TB researchers, healthcare workers, and patients, past, present, and future, whose dedication continues to drive progress in the care, understanding, and control of TB worldwide.

Antimicrobial resistance (AMR) is one of the fastest growing threats to global human health. Neonates are particularly vulnerable to AMR infections and this problem is exacerbated in LMICs where access to healthcare services and resources including antibiotics are limited. Carbapenem-resistant infections are a major healthcare challenge as carbapenems are often a last-line defence against MDR infections. Bacteria colonizing hospital environments serve as a reservoir of- and potential transmission risk for nosocomial infections.ObjectivesDetermine the prevalence of carbapenemase-producing bacterial species colonizing hospital surfaces in neonatal, maternity and delivery wards in 13 hospitals across Bangladesh, Pakistan and Nigeria overall, by surface type and by facility.MethodsHigh-touch and low-touch hospital surfaces across 13 hospitals in Bangladesh, Pakistan and Nigeria were sampled on-site using sterile cotton swabs and saline monthly over three consecutive months. Swabs were stored using Amies charcoal media and transported to the UK where they were cultured for 24–48 h at 37°C on antibiotic-supplemented (vancomycin and ertapenem) and non-supplemented chromatic detection agar media (Liofilchem®). Species from bacterial growth from non-supplemented agar were identified via MALDI-TOF MS and PCR was employed for samples with growth on agar supplemented with vancomycin and ertapenem to screen for carbapenemase genes blaNDM, blaKPC, blaOXA-48-like, blaVIM and blaIMP.ResultsA total of 1854 hospital surface samples have been processed to date. Microbial growth was obtained from n=1636/1854 surface samples. The most common species cultured were Pseudomonas stutzeri (n=717/1854), Enterobacter hormaechei (n=165/1854) and Klebsiella pneumoniae (n=135/1854). Carbapenemase-producing isolates (n=744) were found in n=476/1854 surface samples, including blaNDM (n=249/1854), blaKPC (n=11/1854), blaOXA-48-like (n=60/1854), blaVIM (n=221/1854) and blaIMP (n=82/1854). The surfaces most commonly colonized by carbapenemase producers were in or around sinks including sink bowls/basins, taps and floors surrounding sinks. The species comprising the greatest number of isolates per gene were K. pneumoniae for blaNDM and blaKPC (n=59/394 and n=7/14, respectively), K. pneumoniae for blaOXA-48-like (n=21/76) and P. stutzeri for blaVIM and blaIMP: (n=122/263 and n=46/93, respectively).ConclusionsThe high contamination rate of hospital surfaces with clinically relevant bacterial species across African and Asian hospitals is concerning. The prevalence of species harbouring carbapenemase genes is also of particular concern in terms of neonatal sepsis as they may be providing a route of transmission for AMR infections to neonatal patients born and treated within the hospitals. Future work will include further hospital surface sampling in addition to performing comparisons between isolates from hospital surfaces and pathogens found to cause neonatal sepsis from the same hospital sites, to investigate potential transmission networks.

In accordance with the 2022 recommendations of the WHO and the 2024 National Guidelines of the Russian Federation, alongside the standard 6 month chemotherapy regimen for drug-sensitive TB (DS-TB), a 4 month regimen comprising rifapentine (1200 mg/daily) instead of rifampicin, and moxifloxacin instead of ethambutol, is permissible. However, the use of this regimen in clinical practice may be limited by the high dose of rifapentine, particularly in patients with comorbidities. Therefore, developing alternative shortened therapy regimens with a favourable safety profile and proven efficacy remains an urgent task.ObjectiveTo compare the effectiveness of treatment of pulmonary TB with preserved susceptibility of M. tuberculosis to isoniazid, rifampicin and fluoroquinolones between the standard 6 month regimen and the new short 4 month regimen with rifapentine (900 mg/daily) instead of rifampicin and levofloxacin instead of ethambutol.MethodsThe treatment of 92 TB patients with preserved drug susceptibility to anti-TB drugs was analysed. The patients were divided into two equal groups that were comparable in terms of age, gender and prevalence of the disease. The main group underwent a 4 month regimen comprising an intensive phase (8 weeks) of rifapentine (900 mg/daily), isoniazid, pyrazinamide and levofloxacin, followed by a continuation phase (9 weeks) of rifapentine (900 mg/daily), isoniazid and levofloxacin. The control group was assigned a standard 6 month schedule. The preliminary effectiveness of the 4 month regimen was assessed using a cumulative score determined at the primary combined endpoint (8 weeks), according to the following criteria: sputum culture conversion, positive X-ray dynamics and improvement in the intoxication syndrome.ResultsThe cumulative score at the primary combined endpoint (8 weeks) was 34.89 ± 3.2 in the 4 month regimen group and 25.56 ± 1.5 in the control group (Student’s t-test, P=0.019).ConclusionsPreliminary results indicate that a 4 month regimen involving rifapentine (900 mg/daily), isoniazid, pyrazinamide and levofloxacin is more effective than a 6 month regimen for treating DS-TB.

AbstractA 79-year-old female diagnosed with myelodysplastic syndrome (MDS) and germline GATA2 mutation, on compassionate cobimetinib, was admitted with subacute cough and dyspnea. Chest imaging demonstrated a new, large, left hilar mass and consolidation with scattered diffuse mediastinal, supraclavicular, and hilar lymphadenopathy. A core biopsy of the right supraclavicular lymph node was performed. Acid-fast bacilli (AFB) cultures from both the lymph node and blood were positive, while all fungal cultures were negative. Two distinct AFB colonies were observed on solid media and identified as Mycobacterium avium complex and Mycobacterium kansasii. Disseminated non-tuberculous mycobacterial infections involving two distinct species are rare, pose treatment challenges, and may correlate with cobimetinib administration for MDS, as well as GATA2 germline mutations.

Healthcare is a significant contributor to global emissions, with the sector accounting for large proportion of emissions worldwide. The NHS aims to achieve net-zero carbon emissions, necessitating constant evaluation of clinical practices that contribute to environmental waste. A key antimicrobial stewardship goal; IV to oral switch encompasses these key areas, with oral antibiotics presenting a potentially cheaper, time-efficient and less environmentally damaging option while maintaining effectiveness.ObjectivesThe aim of this evaluation was to compare the nursing time taken for preparation and administration of IV antibiotics versus oral antibiotics and to assess and quantify the associated clinical waste and potential environmental impact. A further aim was to compare the cost between preparations.MethodsA literature review was conducted to inform project design and contextualize findings. A data collection sheet was developed to record antibiotic type, route, preparation time, administration time and waste materials used. Observations were conducted over three days on a surgical ward within a district general hospital. Nursing staff were timed using a stopwatch during their preparation and administration of antibiotics. Equipment use was categorized (e.g. syringes, infusion bags, PPE) and quantified for each observed dose.ResultsThe data collected included 36 IV preparations, 28 IV administrations and 18 oral administrations. The average time required to prepare and administer IV antibiotics was 14.12 min, compared to 0.27 min for oral antibiotics, a 98.1% difference. Over one month, this equated to 295.81 h of nursing time for IV antibiotic tasks, versus 5.66 h if all doses were oral. Waste analysis showed IV antibiotics generated significant single-use plastic waste, including multiple syringes, water for injection vials, giving sets and PPE per dose. Flucloxacillin produced the highest volume of waste due to its high dosing. In contrast, oral antibiotics generated minimal waste. The associated cost was considerably higher per IV dose when compared to their oral equivalent.ConclusionsA substantially greater impact on nursing time was observed with IV antibiotic use versus oral. Oral antibiotics were additionally observed to cost less and produce minimal waste, when compared with IV, particularly where high daily doses were required. These findings are relevant given the existing workload pressures faced by nursing staff, where reducing time spent on IV preparation and administration could free capacity for direct patient care. Furthermore, the results highlight the importance of adopting more sustainable antibiotic prescribing practices. One solution is to reduce IV antibiotic burden by implementing prompt IVOST. Further research is needed to explore the wider impact across health boards and support national policies aligned with NHS sustainability goals.

As the first half of 2025 progresses, yellow fever (YF) has again made a concerning resurgence in the Americas. On March 31, 2025, the Pan American Health Organization (PAHO) issued an epidemiological alert warning of a notable increase in YF cases across several countries.1 As of April 26, 2025, 212 confirmed human cases had been reported, with 85 deaths, corresponding to an estimated case fatality rate of approximately 40%. Brazil alone has documented 110 cases and 44 deaths; Colombia follows with 60 cases and 24 deaths; Peru has confirmed 35 cases with 12 deaths; and Bolivia reported 3 cases and one fatal case (Figure 1). The red dots in Figure 1 represent confirmed YF cases, with the size of each dot proportional to the number of cases in each location, scaled relative to the total case count across the region. These numbers have since increased, suggesting an urgent and evolving public health crisis.Open in ViewerFigure 1. Confirmed human cases of yellow fever in the Americas Region in 2025. The red dots represent confirmed yellow fever cases, with the size of each dot proportional to the number of cases in each location, scaled relative to the total case count across the region.Source: Adapted with permission from PAHO.1PAHO, Pan American Health Organization.What is particularly troubling in this resurgence is the pattern of geographic spread. While YF cases in 2024 were primarily clustered in the Amazonian regions of Bolivia, Brazil, Colombia, Guyana, and Peru, 2025 has seen the virus spread beyond these traditionally endemic zones. New cases have emerged in São Paulo (Brazil) and the Tolima and Caldas departments (Colombia)—outside the Amazonian basin. This geographic shift signals potential changes in viral ecology and transmission dynamics. While no large urban outbreaks of YF have been reported in the Americas since 1942, with the exception of some inconclusive origin human cases. Most of the cases reported from recent outbreaks, as those occurring in Brazil, in 2016–2018 and Colombia in 2024–2025, have been successfully attributed to exposure to the sylvatic cycle, with analyses tracing the virus’s spread to peri-urban areas leading to sporadic human spillovers.2 In the last decade, YF cases from Brazil have been concentrated in men, racialized and marginalized groups, and rural workers, underscoring the role of sociodemographic and occupational risk factors.3 The case fatality rate in the most recent outbreaks remains high, between 40% and 50%, urging active surveillance and targeted prevention measures.Orthoflavivirus flavi, formerly known as Yellow fever virus (YFV), is a mosquito-borne flavivirus that has historically maintained an enzootic cycle involving nonhuman primates and sylvatic mosquitoes, particularly species of the genus Haemagogus and Sabethes. Human infections typically occur as spillovers when individuals enter or live near forested areas. However, in the past, urban outbreaks—driven by Aedes aegypti mosquitoes—caused devastating epidemics, particularly before the development and widespread use of vaccines. Although no sustained urban transmission has occurred in recent decades, the conditions for such a scenario remain dangerously present.Urban centers across South America are densely populated and support large populations of A aegypti mosquitoes. Environmental and social changes—driven by macro-determinants such as economic activity, land use, climate change, and human migration—can, as observed in the past, alter the geographic distribution of mosquito vectors and their vectorial capacity.4 Vector presence in non-endemic regions alone is not enough to cause YF outbreaks or sustained transmission. Such a spread requires a combination of favorable environmental and social factors. Sylvatic transmission depends on reservoir hosts such as nonhuman primates, while urban cycles can persist through human-mosquito-human transmission. Though cities at high altitudes like Bogotá or La Paz are unlikely to see urban transmission, many areas in the region have climate conditions—such as suitable temperature, altitude, and rainfall—that could support the virus’s expansion into new urban and peri-urban zones.Viral dynamics and human risk lie in the complex entanglement of vector biology, population immunity, viral adaptation, and ecological isolation. Understanding this interplay is not only a scientific endeavor—it is now a public health imperative. A critical gap in our preparedness is the limited understanding of the current vector competence of urban A. aegypti populations for YFV. Earlier laboratory studies demonstrated that A. aegypti are competent vectors,5 yet ongoing genetic and environmental changes may have altered their susceptibility and transmission efficiency. Given the dynamic nature of arboviral evolution and mosquito adaptation, comprehensive studies assessing the vector competence of urban mosquito populations to YFV must be prioritized. These studies should evaluate YFV infection and dissemination rates in mosquitoes and how climate change, insecticide resistance, and urbanization could reshape vectorial capacity. Without such data, predictions about urban YF outbreaks remain largely speculative. Another potential surveillance tool is to conduct virological surveillance of primate populations and use satellite imagery to monitor deforestation in endemic areas. This enables targeted identification of high-risk zones for YF outbreaks.The concept of “bridge vectors”—mosquito species that inhabit both sylvatic and urban environments and feed on multiple host species—is particularly relevant. Haemagogus leucocelaenus, H. equinus, H. janthinomys, H. lucifer, and Sabethes chloropterus are efficient YFV transmitters in forested environments. However, their potential role as intermediaries in peri-urban and rural settings remains underexplored. The emergence of cases in São Paulo and Tolima—regions characterized by complex mosaics of fragmented forests, agricultural zones, and expanding urban sprawl—suggests these species may play a more prominent role as bridge vectors than previously recognized. Mapping their geographic distributions, host-feeding behavior, and population dynamics could provide critical insights into how YFV expands its territory without igniting full-scale urban outbreaks.Another underappreciated factor is the role of ecological isolation in shaping the evolutionary trajectory of YFV. Within distinct forest patches, isolated primate and mosquito populations may create micro-environments that foster viral diversification.6 This diversification can affect virulence, transmissibility, and antigenic properties, potentially impacting diagnostic accuracy and vaccine effectiveness.7 Monitoring YFV genetic diversity across enzootic regions is, therefore, essential. It enables the early detection of emerging lineages that may evade immune responses in previously vaccinated individuals or exhibit altered vector compatibility. These insights can inform vaccine strain selection and help anticipate future outbreak potential.The current epidemiological picture underscores the need for robust sentinel surveillance, particularly in areas where natural ecosystems intersect with human settlements—ecotones. Zoonotic spillovers from epizootic foci are most likely to occur in these transitional zones, and early detection could prevent urban amplification.8 Implementing real-time surveillance systems in such high-risk zones could involve monitoring nonhuman primate morbidity and mortality, strategically deploying mosquito traps across ecological gradients, engaging local communities in syndromic reporting, and tracking viral evolution through genomic surveillance using publicly available data-sharing platforms, such as GISAID EpiArbo.9 A sentinel system focused on these emergence zones could function as an early warning beacon, allowing health authorities to preemptively deploy vaccines and vector control measures. Timely detection of cases can help identify areas of active transmission, allowing for more effective and targeted vaccination campaigns. Other effective strategies, such as sterile male releases or genetically engineered transgenic mosquitoes, may also hold promise.10,11Despite the effectiveness of the YF vaccine, coverage remains patchy in many at-risk regions, partly due to logistical challenges, political instability, and vaccine hesitancy. The World Health Organization’s Eliminate Yellow Fever Epidemiology (2017–2026) includes vaccination rate targets for 2026 that may be insufficient in regions where R > 0.5.12 Nonetheless, the YF vaccine remains the most crucial and effective tool for countries to combat transmission, particularly in endemic sylvatic cycles. As long as YFV persists in forest ecosystems—sustained by nonhuman primate hosts and competent sylvatic vectors—there will always be a risk of human infection, especially as deforestation and encroachment increase human-wildlife contact. Routine immunization must remain a cornerstone of public health policy in all at-risk countries and should ideally be integrated into broader zoonotic disease prevention programs. While mass vaccination campaigns are effective during outbreaks, they remain reactive measures. Sustained immunization—particularly for children, migrants, and travelers—is the only proactive shield against the urbanization of the disease.13 A 2024 phase I clinical trial evaluated the safety and immunogenicity of a next-generation live-attenuated YF vaccine produced in a Vero cell line. Results showed comparable performance to the licensed YF-VAX vaccine, offering a promising option for future scalability.14Governments across the region must recognize that the response to YF requires sustained, long-term efforts grounded in a comprehensive approach that addresses the full range of biological, environmental, social, and structural disease determinants. While reactive measures may sometimes be warranted, they must be based on sound scientific evidence. For example, the recent announcement by Colombia’s Ministry of Health to impose mobility restrictions on unvaccinated individuals could disproportionately impact vulnerable populations and would be difficult to enforce in practice. Yellow fever’s resurgence in 2025 is not merely a regional public health problem but a global warning. It reminds us that zoonotic diseases remain a persistent threat, our surveillance systems are often reactive, and ecological changes erode the natural buffers that once contained these pathogens. We must update and expand vaccination campaigns in endemic and peripheral areas to respond adequately. We must invest in ecological and entomological research to understand the evolving dynamics of vector competence and emergence zones. Health teams must be trained to conduct effective entomological surveillance of both sylvatic and urban vectors. We must establish surveillance systems integrating human, animal, and environmental health data—a proper One Health approach.15 We must engage local communities and train health workers to recognize early signs of YF and other arboviral illnesses. It is also essential to communicate that primates do not transmit YF to avoid the sacrifices of these emblematic animals. Finally, we must strengthen regional collaboration to share data, coordinate responses, and ensure equitable access to vaccines and diagnostic tools.More broadly, the COVID-19 pandemic left behind critical lessons that some governments are now beginning to overlook. These include the importance of sustaining the public health workforce, recognizing vaccination as a universal public good, strengthening health surveillance systems, and maintaining the operational capacity of public health laboratories. These capacities must be preserved and strengthened at the national and local levels to ensure timely and effective responses. Unfortunately, signs of regression in these priorities are on the public agenda in several countries. The current situation with YF is just one of many public health alerts emerging across the Americas—alongside measles, pertussis, and other re-emerging threats—that underscore the urgent need to reinforce regional preparedness and response systems.The 2025 YF outbreak reminds us that YF is not a disease of the past. Like the forests it inhabits, it remains deeply entangled with the present—and dangerously poised to shape the future.AcknowledgmentsNone.ORCID iDsGabriel Parra-Henao https://orcid.org/0000-0003-4535-6521José A. Usme-Ciro https://orcid.org/0000-0002-8093-0544Andrés F. Henao-Martínez https://orcid.org/0000-0001-7363-8652

Military metaphors are pervasive in the language of antibiotics and resistance [1]. Bacterial pathogens are foreign invaders with whom we are at war. To fight them, we deploy antibiotic chemical weapons. Powerful antibiotics are big guns, narrow-spectrum antibiotics are silver bullets, and broad-spectrum antibiotics are nuclear bombs. We are in a losing battle against antibiotic-resistant superbugs and are quickly running out of ammunition. Metaphors shape how you feel about an issue, and how you feel often determines the actions you take. Framing resistant bacteria as foes to be feared traps us within a fundamentally adversarial us-versus-them framework, where the only possible outcomes are their eradication—or ours.Instead, a growing body of research is re-examining antibiotic resistance through the lens of microbial ecology and evolution. This work has revealed that bacterial resistance significantly predates human use of modern antibiotics, with resistance genes encoded by many benign environmental and commensal microbes often before they are detected in pathogens [2]. Unchecked antibiotic use can dramatically amplify these resistance reservoirs, select for their dissemination into pathogens, and engender significant negative side effects on human health [3,4]. Given the dwindling approval of new antibiotics [5], we must rethink the brute force approaches that attempt (futilely) to fight resistance by searching for the next bacteria-killing molecule. Instead, we should pursue complementary strategies that mitigate selection of antibiotic resistance by improving stewardship and developing novel chemotherapeutic, probiotic, and prebiotic strategies for suppressing bacterial virulence and supporting commensalism.To contrast the military metaphors, we metaphorize these ecological approaches using the diplomatic terms of disarmament, nonproliferation, and soft power. Strategies for targeted killing of pathogens must invariably remain part of our medical toolkit, but by pursuing these complementary strategies, we argue that we can better foster sustainable coexistence with our microbial cohabitants and tip the scales against pathogenesis.Ancient originsMost of our clinically important antibiotics were discovered as the natural products of soil-dwelling microbes, particularly the actinomycetes. In 1973, Davies and Benveniste [6] proposed that the origins of antibiotic resistance likely lie with these ‘producer’ microbes, as they would need elements to protect themselves from the antibiotics they produced. These elements, by definition, would be antibiotic resistance genes. The producer hypothesis postulates that, over time, nonproducers (including pathogens) have acquired these antibiotic resistance genes via horizontal gene transfer. With the evolution of antibiotic production estimated to have occurred hundreds of millions of years ago [7], antibiotic resistance must be just as old.The ancient origin of resistance in soil-dwelling microbes explains several features of resistance that seem puzzling, even paradoxical, when viewed solely from a clinical perspective. First, it explains why soil-dwelling actinomycetes encode resistance to more antibiotics than most clinical pathogens; seven or eight antibiotics on average and 15 at the upper end, including synthetic compounds recently approved for clinical use [8]. It also explains why soils are an immense and diverse reservoir of previously unknown resistance genes not yet seen in pathogens, and how soil bacteria and human pathogens can encode identical resistance genes [9].Collateral damageZooming in on the subset of antibiotic-resistant bacteria associated with humans, an ecological perspective also reveals the unintended knock-on effects of antibiotic use. Pathogens are not the only microorganisms that make a home in the human body. Each human is colonized by 500 to 1000 species of bacteria, and the number of bacterial cells is roughly the same as the number of human cells [10,11]. These microbes form relatively stable ecosystems called microbiomes, which are just as diverse and complex as macroscopic ecosystems. The densest and most diverse human-associated microbiome is in the colon: the gut microbiome. A healthy gut microbiome is critical to human health and development, providing essential functions including nutrient processing, colonization resistance against pathogens, and immune system regulation [12].Antibiotics are developed with the intention of killing a select few pathogenic species; however, they do so by targeting conserved molecular features shared by thousands of species across broad branches of the tree of life (e.g. cell wall, ribosome, and RNA polymerase). As a result, they can indiscriminately kill both ‘good’ commensal species and ‘bad’ pathogens when administered to humans. Taking antibiotics to treat a gut infection may kill the pathogen, but at the cost of potential significant collateral damage—another military metaphor—to commensal microbiota. Substantial antibiotic perturbations can lead to microbiome dysbiosis, characterized by acute or persistent imbalanced states of the commensal ecosystem. Dysbiosis has been associated with numerous human pathologies, including inflammatory bowel disease, obesity, and gastrointestinal cancers [13]. Postantibiotic microbiome dysbiosis can also lead to the expansion of opportunistic pathogens in the gut, such as Clostridioides difficile.Antibiotic use also contributes to antibiotic-resistant bacteria within the gut microbiome. In the context of war, killing any number of the enemy seems like a victory; however, like a hydra sprouting two heads after one is chopped off, repeated antibiotic exposures will make antibiotic-resistant bacteria stronger [14]. Importantly, this is not limited to pathogens. The same evolutionary dynamics of selection and adaptation also drive increasing resistance in nonpathogenic species. Often, this resistance is encoded on mobile genetic elements, which can lead to more frequent and broader dissemination of resistance by horizontal gene transfer.Diplomatic strategiesNo matter how many bacteria-inhibiting drugs we develop, antibiotic resistance is never going away. The history of antibiotic use and resistance in pathogens has taught us that resistance is not a question of ‘if,’ but rather, only a matter of ‘when’ [15]. Instead, we must develop complementary strategies that enable us to better coexist with resistant bacteria while suppressing their virulence. New approaches draw on two key insights from microbial ecology: first, the real problem is not resistance alone, but resistance coupled with pathogenesis. Second, pathogenesis is not binary between ‘good’ commensals and ‘bad’ pathogens but a spectrum between them. For example, pathobionts like Escherichia coli can be harmless members of a healthy gut microbiome or a disease-causing gastrointestinal, urinary, or blood pathogen, depending on the host and environmental context, and the induced expression of specific repertoires of virulence factors. In contrast to the martial strategies of conventional antibiotic deployment for bacterial growth inhibition or killing, ecology-first strategies borrow from tenets of diplomacy and cooperation to attempt to mitigate antibiotic resistance by tilting bacteria away from pathogenesis toward avirulence and commensalism.De-escalation (antibiotic stewardship)The back-and-forth between humans deploying new antibiotic drug classes and bacteria evolving new resistance mechanisms has been called an evolutionary arms race. That implies bacteria are a strategic enemy consciously responding to our moves with countermoves; however, the expansion and spread of resistance is an automatic evolutionary response to the selective pressure imposed on bacteria by prolific antibiotic use. The development of more powerful antibiotics to eradicate resistant bacteria continues to escalate the antibiotic arms race. Instead, we should strive for de-escalation by controlling and reducing antibiotic use.In clinical settings, stewardship programmes encourage prudent use of antibiotics. They establish guidelines for ‘the optimal selection, dosage, and duration of antimicrobial treatment that results in the best clinical outcome for the treatment or prevention of infection, with minimal toxicity to the patient and minimal impact on subsequent resistance’ [16]. In conjunction, it is critical to reduce antibiotic overuse and misuse. Overuse occurs when antibiotics are used unnecessarily, such as in patients with viral infections, noninfectious processes, bacterial infections that do not require antibiotics, and bacterial colonization [17]. It also occurs when medically important antibiotics (i.e. from classes important to human medicine) are used in food-producing animals to improve growth rates [18]. Antibiotics are also frequently misused, such as when broad-spectrum antibiotics continue to be used even after culture data indicates the pathogen is not susceptible to that regimen or a narrower-spectrum compound would be just as effective. Phage therapy, which kills antibiotic-resistant pathogens using highly specific bacteriophages, is also a promising alternative to traditional antibiotics.Disarmament (antivirulence)At its root, the problem with antibiotic-resistant infections is not resistance per se but pathogenesis. An alternative to antibiotics are antivirulence therapeutics that disarm pathogens of the virulence factors pathogens produce to cause infection and evade the host immune response (e.g. toxins, adhesins, immune evasion and modulation factors, and siderophores) [19,20]. Interestingly, antivirulence strategies historically precede antibiotic use, with Emil von Behring developing an antiserum directed against diphtheria toxin in 1893. A related strategy, referred to as plasmid interference and plasmid curing, disarms pathogens at the genetic level by eliminating problematic plasmids that encode for virulence and antibiotic resistance genes [21,22].These approaches have several distinct advantages compared with conventional antibiotics: first, because commensal bacteria generally do not encode for virulence factors, there is less potential impact on the broader gut microbiome. Second, although antibiotics target central growth pathways, these strategies do not destroy host bacterial populations outright. That may mean there may be less evolutionary pressure for the development of antivirulence and plasmid-curing resistance. Although several antivirulence agents have entered clinical trials, most are still in the preclinical stage [20].Soft power (colonization resistance)Whether a pathobiont is a virulent pathogen or a benign commensal can depend on its ecological context. Commensal microbiota can suppress the proliferation of incoming pathogens and the expansion of opportunistic pathobionts via several mechanisms of colonization resistance. These include nutrient competition, altering host environmental conditions, and enhancing the intestinal epithelial barrier [23]. By modulating the gut microbiome using probiotics and prebiotics, or replacing dysbiotic microbiota with those from healthy donors through faecal microbiota transplantation, we can exert a kind of soft power that influences pathobionts to adopt commensal lifestyles, rather than pathogenic ones.For example, patients can be asymptomatically colonized for prolonged periods by toxigenic C. difficile. Our recent work has identified certain commensal gut species that protect against disease by suppressing C. difficile toxin production while allowing its stable growth and colonization in the gut [24]. By identifying such commensal-pathogen dynamics, we can design microbiome-targeting therapeutics that alter the gut microbiome in a way that steers pathogens toward avirulent states.ConclusionLike any good diplomatic strategy, the best approach to mitigate antibiotic resistance will be multipronged. When used in combination, many of the approaches above could reinforce each other: reducing antibiotic use increases the gut microbiome’s colonization resistance against C. difficile and neutralizing the C. difficile toxins with antivirulence drugs bolsters the action of commensal species that suppress toxin production. Similar approaches may also be used beyond the gut microbiome to address antibiotic-resistant infections from the oral, skin, and genital microbiomes. However, we stress that although they are frequently based on well-established ecological principles and in vivo studies, the therapeutic potential of these approaches remains speculative. Few have been proven in the field and, if implemented, should be tailored to the specifics of each patient, pathogen, and disease.We likely will always need targeted antibiotics to kill the deadliest human pathogens, especially in immunologically vulnerable patients. However, as in human diplomacy, we must strive to explore all other options to suppress pathogenesis and only resort to armed conflict as a last resort. Nonantibiotic strategies of de-escalation, disarmament, and soft power promise to be effective for patients in the short term while conserving our antibiotic arsenal for those who need it most. By understanding and leveraging the ecological and evolutionary processes underlying antibiotic resistance, pathogenesis, and commensalism, we may be able to more sustainably and peacefully coexist with the diversity of microbes who live in and on us.CRediT authorship contribution statementKevin S. Blake: Writing – review & editing, Writing – original draft, Conceptualization. Gautam Dantas: Writing – review & editing, Writing – original draft, Funding acquisition, Conceptualization.Declaration of competing interestThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.AcknowledgementsWork in the authors’ laboratory was supported in part by awards to G.D. through the National Institute of Allergy and Infectious Diseases (NIAID; award numbers U01AI123394, R01AI155893, R01AI184858), and the National Center for Complementary and Integrative Health (NCCIH; award number U01AT012998) of the National Institutes of Health (NIH).

Standard culture methods for antimicrobial susceptibility testing (AST) are manual and laborious. Fast, higher throughput, digital and automated AST is urgently required for veterinary and human health. Astratus Limited was spun out from the University of Reading to commercialize our novel, affordable microcapillary AST platform. The modular instrument design allows medium to ultra-high-throughput rapid AST options for both direct-from-urine and isolate testing. We previously presented the analytical performance of direct-from-urine microcapillary AST with >450 human urine samples showing 96% sensitivity (95% CI: 92%–99%) and 93% specificity (95% CI: 89%–96%) for detecting microbial growth in urine with >105 cfu/mL. Direct-from-urine microcapillary AST showed a concordance with broth microdilution of 557/572 bacteria/antibiotic combinations (97.38%) for urine samples with a single Gram-negative bacteria. The mean time to AST result was 5.7 h from 85 positive urine samples. Analytical performance data for Gram-positive bacteria is equally critical. Our objective was to analytically validate microcapillary AST with Gram-positive bacterial isolates.MethodsAnalytical validation of microcapillary AST was conducted with a total of 425 antibiotic/Gram-positive isolates: Staphylococcus and Enterococcus isolates from human urine samples (Hampshire Hospitals, UK); and Staphylococcus isolates from canine urine, pus and wound samples (Batt Laboratories, UK). Human isolates were tested side-by-side, microcapillary AST was compared to the in-house reference broth microdilution method (BMD). Canine isolate microcapillary AST results were compared to Batt Laboratories reported results. In this study 18 antibiotics were tested simultaneously using microcapillary AST for continuous growth measurement: Ampicillin; Amoxicillin clavulanic acid; Nitrofurantoin; Clindamycin; Doxycycline; Gentamicin; Oxacillin; Enrofloxacin; Penicillin G; Erythromycin; Cephalexin; Cefoxitin.ResultsHigh categorical agreement was observed indicating excellent accuracy for both human (128/135; 95%) and canine (265/290; 91%) Gram-positive isolates. Closer agreement was seen for human isolates tested with microcapillary AST compared to in-house BMD than for canine isolates tested with microcapillary AST compared to laboratory-reported results. However, overall accuracy remained high for canine isolates. The difference in accuracy was found to be related to an increased number of discordant results for certain antibiotics within the canine isolate dataset. For seven of the antibiotics tested, accuracy exceeded 90%. Four antibiotics showed lower agreement (ampicillin, penicillin G, erythromycin and cephalexin).ConclusionsDetailed analysis of discordant results is planned alongside side-by-side comparison of canine isolates tested with microcapillary AST compared to in-house BMD to remove the laboratory-to-laboratory variation. Further optimization of the system may be required to ensure that the accuracy of the AST results for the 4 antibiotics matches the high accuracy of the AST results of the other antibiotics so that we can ensure that we provide faster, accurate antimicrobial susceptibility testing results to enable more targeted antimicrobial prescribing. Our aim is to replace manual, labour intensive methods with a single, simple-to-use, higher throughput platform unlocking productivity and relieving laboratory workforce pressures across human and veterinary medicine.

ABSTRACTGepotidacin is a novel first-in-class triazaacenaphthylene antibiotic developed for the treatment of uncomplicated gonorrhea and uncomplicated urinary tract infections (uUTIs). To support uUTI, in vivo pharmacokinetics (PK)/pharmacodynamics (PD) studies have been conducted in the murine neutropenic thigh infection model evaluating gepotidacin against 17 isolates of Escherichia coli and 7 isolates of Klebsiella pneumoniae having MICs of 0.25 to 16 µg/mL. Exposure data were fit using a population PK model, and efficacy data were fit with an inhibitory effect sigmoid Imax model using free-drug area under the concentration-time curve (fAUC)/MIC as the primary index. The ratios associated with response were determined for each isolate, and the median fAUC/MIC (excluding strains with <1-log of growth from baseline) was 13. To retain the data for all isolates, an exploratory analysis was conducted using different criteria to determine the target for strains with <1-log of growth (6 of the 24 strains tested). While the median fAUC/MIC was similar regardless of the criteria used, this analysis highlights the importance of critically reviewing PK/PD data for trends related to isolate characteristics and/or individual study outputs. As previously reported, the systemic targets determined from PK/PD studies were applied to urine concentrations for probability of target attainment analyses, which led to successful clinical trials and regulatory approval for gepotidacin in the treatment of uUTI. However, further work is needed to confirm the translational validity of these approaches on a broader scale and their application in establishing PK/PD targets for cystitis.

The primary challenge for global and national TB control programmes is TB resistant to rifampicin and isoniazid (MDR-TB), including cases with additional resistance to fluoroquinolones (pre-XDR-TB). Treatment regimens for both MDR-TB and pre-XDR-TB include chemotherapy protocols containing bedaquiline and linezolid. A key aspect of the effective use of these drugs in TB control programmes is the low level of drug resistance to them. Prescribing effective treatment regimens is essential for reducing the burden of drug-resistant TB and preventing the spread of XDR-TB, particularly in countries with a high burden of MDR-TB. Currently, phenotypic methods are the main approach for drug susceptibility testing (DST) of bedaquiline and linezolid.ObjectivesTo determine the level of drug resistance to bedaquiline and linezolid (XDR) among various categories of pre-XDR-TB patients.Materials and methodsData on the number of TB cases and DST results were obtained from the Federal TB Register in 2024. We determined the coverage and results of drug susceptibility testing for bedaquiline and linezolid among new cases, relapses and chronic cases of TB with pre-XDR.ResultsIn 2024, 36 104 new TB cases were registered in the Russian Federation. Bacteriological confirmation was obtained in 53.0% of cases (19 124/36 104). Among confirmed cases, resistance to fluoroquinolones (pre-XDR) was detected in 8.3% (1589/19 124) of patients. Among patients with pre-XDR-TB, testing for bedaquiline resistance was performed in 80.8% of cases (1284/1589), with resistance detected in 6.5% (84/1284). Testing for linezolid resistance was performed for 81.2% (1290/1589) of patients, with resistance found in 3.3% (42/1290) of cases. Among patients with TB relapses (n=8711), bacteriological confirmation was obtained in 56.9% of cases (4953/8711). The frequency of pre-XDR in this group was 18.4% (911/4953). The coverage of testing for bedaquiline and linezolid resistance was 82.9% (755/911) and 83.9% (764/911), respectively. The resistance rate to bedaquiline was 7.3% (55/755) and to linezolid - 6.0% (46/764). In the group of chronic TB patients (n=11 217), bacteriological confirmation was obtained in 55.8% of patients (6262/11 217). The proportion of pre-XDR was highest (P<0.05) in this cohort - 28.1% (1763/6262). The coverage of testing for bedaquiline and linezolid resistance was 72.1% (1272/1763) and 73.5% (1295/1763), respectively. The frequency of resistance to bedaquiline reached 12.1% (154/1272) and to linezolid - 7.6% (99/1295).ConclusionsThis analysis demonstrates significant heterogeneity in the drug resistance structure of Mycobacterium tuberculosis depending on patient categories. The prevalence of pre-XDR and XDR-TB was statistically significantly lower among newly diagnosed cases compared to patients with TB retreatment and especially chronic TB patients. Determining drug resistance to bedaquiline and linezolid is essential for various cases of chronic TB to prescribe adequate personalized treatment regimens.

Routine surveillance for resistant organisms in hospitalized patients’ gastrointestinal tracts is used to prevent spread. Results can also be used to inform empirical antibiotic recommendations. However, the association between enteric resistance on surveillance and resistance in subsequent invasive infections in children is not evidencedObjectivesWe studied the capacity of resistance identified from enteric surveillance to predict resistance in Gram-negative bloodstream infection isolates (GN-BSII).Patients and methodsWe obtained respectively 10 years and 7 years 8 months of data from Great Ormond Street and Alder Hey Children’s Hospitals (GOSH and AHCH, respectively), comprising GN organisms identified on enteric surveillance and blood cultures. For each unique GN-BSII with surveillance result available in the preceding 5 days to 6 months, the presence of resistance to specified antibiotics was determined. 2×2 tables were constructed for each antibiotic. Within BSII with surveillance available in the preceding 5 days to 1 year, changes in test performance with increasing lookback were explored. Sensitivity analyses explored approaches with organism matching and application of EUCAST rules. Organism-group-weighted resistance probability predictions were produced from the organism matching approach.ResultsA total of 1119 individuals had GN-BSII, of which 91% (n=1022) had surveillance available. 1830 isolates were included. Escherichia coli, Klebsiella pneumoniae, Enterobacter spp. and Pseudomonas aeruginosa represented 56% of BSII. 53% and 77% of patients (AHCH and GOSH) had surveillance in the prior 6 months. Rates of resistance to gentamicin, piperacillin/tazobactam and ciprofloxacin were higher in BSII from patients with preceding resistance on surveillance samples than in patients without in GOSH (36% versus 7%, 36% versus 15%, 46% versus 13%; PPV versus NPV) and AHCH (38% versus 9%, 36% versus 15%, 31% versus 10%). While meropenem resistance was uncommon in AHCH (2% BSII), in GOSH surveillance was predictive (30% versus 5%). Sensitivity was 33%–66% across sites and antibiotics. The rates of ESBL were higher in BSII from patients with preceding ESBL-producers on surveillance (31% versus 4% GOSH; 36% versus 3% AHCH). Extending the inclusion window to one year had minimal impact on sensitivity (<2% in both sites). Organism matching the prevalent species increased specificity at the expense of sensitivity, with increases in PPV. However, this was only a minimal improvement in predictive power.ConclusionsIn two paediatric hospitals, the presence of resistance on surveillance was predictive of GN-BSII resistance across a range of antimicrobials. Predictive values indicate clinically significant differences. Predicting GN-BSII resistance by organism matching offered limited benefit, suggesting pattern of resistance is more informative than pathogen/resistance combinations. Gains by looking back beyond 6 months were limited. The similarity of results between two hospitals with different populations and surveillance practices is notable and may imply some generalizability. The application of these findings for empirical treatment in suspected sepsis will depend on clinical scenarios. Most episodes of suspected sepsis are blood culture negative, but many are caused by microorganisms resident in patient flora. Carriage of resistance is still expected to be relevant. Resistance in GI flora is relevant for predicting resistance in GN-BSII.

Antimicrobial resistance (AMR) is a critical global health threat, projected to cause up to 10 million deaths annually if left unaddressed.1 Antimicrobial stewardship (AMS) provides a structured framework to optimize prescribing practices and curb resistance.2 Dentistry plays a significant role in this crisis, with inappropriate antibiotic prescribing and widespread self-medication fuelling resistance.3 Dental stewardship programmes offer evidence-based frameworks to optimize antibiotic prescribing in dental practice and mitigate the AMR crisis through targeted interventions and improved prescribing protocols.3ObjectivesTo systematically review awareness, knowledge and perceptions of AMR among Indian dental professionals, and to examine AMS implementation, challenges, facilitators and recommendations for strengthening stewardship in dentistry.MethodsA systematic review was conducted in accordance with PRISMA guidelines. Literature was searched across PubMed, Scopus, CINAHL, Web of Science and Google Scholar for studies published between 2015 and 2025. Predefined search terms and Boolean operators were applied. Eligible studies included those involving dental professionals, students and dental settings in India, as well as AMR awareness, knowledge, perceptions of AMR, or AMS-related interventions. Non-human research and studies outside dentistry were excluded. Ethical approval was not required as no patient-identifiable data were used.ResultsA total of 414 records were screened, with 14 studies meeting the inclusion criteria for quantitative and qualitative synthesis. Quantitative findings showed that all studies (100%) referenced guidelines and pathways, while 85.7% assessed prescribing practices and 57.1% included compliance monitoring. However, multidisciplinary teams and audit/feedback approaches were reported in only 7.1%, highlighting significant gaps in AMS implementation. Antibiotic prescribing patterns revealed Amoxicillin/Penicillin accounted for 62% of prescriptions, broad-spectrum agents 31% and others 7%. Ten studies reported problematic behaviours, with 80% empirical prescribing and 70% both prophylactic overuse and misuse of indications. Qualitative analysis identified four themes: poor AMR knowledge, a knowledge-practice disconnect, heterogeneity in prescribing practices and systemic educational/institutional barriers. Collectively, these findings highlight the urgent need for tailored stewardship strategies and integration of AMS into dental education and practice in India and promote dental stewardship in practice.ConclusionsThis systematic review demonstrates a significant gap between theoretical AMR awareness and clinical practice among Indian dental professionals. Findings reveal critical deficiencies in prescribing knowledge, profound knowledge-practice disconnect and inappropriate antibiotic use, exacerbated by systemic barriers. Dental stewardship requires immediate multi-pronged strategies: educational reform, institutional antimicrobial stewardship programmes with monitoring and national evidence-based guidelines. Future research must focus on implementation science, developing practical real-world interventions, such as digital decision support tools, to bridge the identified knowledge-practice gap across dental stewardship settings effectively.

AbstractAntimicrobial resistance (AMR) is a global health challenge that disproportionately affects countries and regions with limited resources. Current efforts to address AMR largely emphasize antimicrobial stewardship (AMS), community engagement and scientific research, while often overlooking systemic factors such as corruption, poverty and inflation, which significantly influence the capacity of individuals and communities to respond effectively. In this article, we suggest that AMR responses must extend beyond traditional AMS and community engagement to adopt a genuinely people-centred approach—one that empowers individuals not only to achieve optimal health but also to navigate broader social and economic constraints. To our knowledge, this is the first report to critically examine the combined impact of corruption, poverty and inflation on AMR.

ABSTRACTCeftolozane/tazobactam is a key antibiotic for Pseudomonas aeruginosa infections. Our objective was to determine whether continuous infusion (CI) of ceftolozane/tazobactam can achieve aggressive pharmacokinetic/pharmacodynamic (PK/PD) targets that suppress resistance and improve outcomes in severe Pseudomonas aeruginosa infections. A retrospective analysis of adult patients receiving CI ceftolozane/tazobactam and therapeutic drug monitoring (TDM) of both compounds was performed. Population PK/PD modeling identified the most accurate method for estimating ceftolozane/tazobactam clearance based on kidney function and Monte Carlo simulations investigated the relationship between various CI dosing regimens and aggressive PK/PD target attainment of ceftolozane/tazobactam. The 2021 non-race-based Chronic Kidney Disease Epidemiological Collaboration (CKD-EPI) equation with body surface area indexation provided the most reliable clearance estimates. Simulations showed that CI regimens of 4–6 g/2–3 g daily achieved optimal target attainment (≥90%) across all kidney function strata for MICs up to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoint. Cumulative fraction of responses remained robust against key resistance phenotypes: MDR (12.6%; 83.8%), pan-beta-lactam-nonsusceptible (8.2%; 78.2%), and difficult-to-treat resistant (6.7%; 71.7%). Even among isolates with MIC >4 mg/L, aggressive target attainment was reached in 15%–40% of cases. This study suggests that CI ceftolozane/tazobactam, informed by TDM and optimized for aggressive PK/PD targets, offers a promising strategy to maximize efficacy and suppress resistance in severe Pseudomonas aeruginosa infections. These findings warrant prospective clinical trials of CI-based, exposure-guided therapy.

:Echinocandins are recommended as an initial treatment for invasive candidiasis. Although safety and efficacy profiles of both anidulafungin and caspofungin are well established, direct head-to-head comparisons have not been reported before.Objective:Compare efficacy and safety of anidulafungin versus caspofungin among patients with invasive candidiasis.Design:Retrospective observational study.Methods:Adult patients with invasive candidiasis who were treated with either anidulafungin or caspofungin for ⩾5 days were retrospectively reviewed over a period of 6 years. The primary endpoint was global response, defined as clinical and microbiological success at the end of treatment duration.Results:A total of 223 patients who received either anidulafungin (n = 176) or caspofungin (n = 47) were initially included. Propensity score matching (based on age, malignancy, level of care, presence of candidemia, and other factors) was performed to improve comparability of the two groups. As a result, 32 patients in the caspofungin arm and 79 patients in the anidulafungin arm were included in the final analysis. Around three-quarters of the cohort had candidemia, and the most common isolated Candida species were C. albicans and C. glabrata. Response rates were comparable between both groups, with the primary outcome of global response showing no significant difference (56.3% for the caspofungin group vs 63.3% for anidulafungin, p = 0.490). Similarly, no differences between the two groups were observed in terms of 90-day all-cause mortality (p = 0.672) or any other secondary endpoints.Conclusion:Our data suggest that anidulafungin and caspofungin have comparable global response among patients with invasive candidiasis. Additionally, both studied echinocandins showed no significant difference in 90-day all-cause mortality. However, due to the limited sample size, larger studies are needed to confirm these results.

:Salmonella species are major pathogens responsible for gastroenteritis and typhoid fever. In 2017, Salmonella enterocolitis caused over 95 million cases of diarrhoea and 50,000 deaths globally, with India bearing more than 50% of the typhoid burden.Objectives:To test the association of monthly mean and maximum temperature, precipitation and absolute humidity with the incidence of Salmonella bacteraemia in a metropolitan city in South India.Design:A retrospective time-series analysis.Methods:This study employed a retrospective time-series analysis to evaluate the influence of meteorological variables, including temperature, absolute humidity and precipitation, on Salmonella bacteraemia in a metropolitan city in South India.Results:Between 2014 and 2019, a total of 492 blood culture-confirmed cases of Salmonella bacteraemia were identified in Mysuru, India. S. typhi was predominantly among younger patients, while non-typhoidal Salmonella was more frequent in older age groups. Resistance was highest to nalidixic acid (84%), with a rising trend in ciprofloxacin resistance. Increased mean temperature (lags 1–3 months) and absolute humidity were positively associated with Salmonella bacteraemia, while temperature variability was protective, and monsoon rainfall significantly increased the risk. Cumulative exposure–response analyses further showed elevated risks at higher humidity (>26 g/m³), temperatures (>34°C) and extreme precipitation (>250 mm), although confidence intervals were wide and most associations did not reach statistical significance.Conclusion:This single-centre retrospective time-series analysis has demonstrated that meteorological variables impact the incidence of Salmonella bacteraemia, which could lead to increased antibiotic use and contribute to the development of resistance.

ABSTRACTNeisseria gonorrhoeae and Chlamydia trachomatis are the most common bacterial sexually transmitted infections. The World Health Organization (WHO) estimates that there were approximately 211 million cases of chlamydia and gonorrhea in 2020. Currently, no single drug is effective against both pathogens. Ceftriaxone (CRO) is the only recommended treatment for gonococcal infections but has no activity against C. trachomatis. Azithromycin (AZM) or doxycycline (DOX) is recommended for C. trachomatis infections; however, N. gonorrhoeae has developed resistance to both agents. Without new therapeutic options, these infections risk becoming untreatable. Utilizing a drug repurposing approach, we identified epetraborole (EBO) as a potent inhibitor for N. gonorrhoeae and C. trachomatis. EBO is a boron-containing compound currently in clinical trials for the treatment of non-tuberculous mycobacterial infections. EBO demonstrated potent activity against multidrug-resistant N. gonorrhoeae with MIC ranging from 0.125 to 0.25 µg/mL. Additionally, EBO exhibited anti-C. trachomatis activity at concentrations of ≤1 µg/mL. Furthermore, EBO was capable of eliminating the intracellular burden of both N. gonorrhoeae and C. trachomatis, surpassing the activity of CRO and AZM. Moreover, unlike CRO and AZM, EBO showed limited activity against the normal vaginal microbiota. Finally, in an in vivo mouse model of CRO-resistant N. gonorrhoeae genital tract infection, EBO exhibited a 99.95% reduction in bacterial burden after 2 days of treatment. Collectively, our findings highlight EBO as a promising candidate for the treatment of sexually transmitted infections that warrants further investigation.

The use of antibiotics varies by setting—hospital, outpatient/primary care—with a disproportionate impact of antimicrobial resistance in low-and middle-income countries. The WHO AWaRe (Access/Watch/Reserve) book gives detailed guidance on the optimal use of antibiotics across primary care and hospitals for adults and children with the aim of improving the quality of use of essential antibiotics.ObjectivesTo develop universally applicable, model sets of appropriate and feasible quality indicators based on the WHO AWaRe system for primary care, hospital and general indicators for optimal antibiotic use.MethodsIndicators identified in a scoping review were revised to focus on primary care and hospital facility infections in the AWaRe book. Condition-specific indicators captured measures such as appropriate antibiotic use, total daily dose and proportion of Access or Watch antibiotics. General indicators were developed from prescribing and dispensing guidelines in the AWaRe book and covered themes from proper documentation in medical records to population-level antibiotic use. The indicators were evaluated through a two-round Global Delphi Technique of 104 and 107 panellists, respectively, to determine appropriateness and feasibility in national and local settings, followed by a two-round RAND/UCLA Appropriateness Method (RAND/UCLA) with 12 panellists rating indicators on a global scale. Panels comprised experts from every WHO region. In Round 1 of each method, panellists also rated clarity and could suggest rewording or new indicators. Findings from Round 2 are reported. The median rating was calculated for each indicator to determine agreement. A median of 7–9 was considered appropriate or feasible. If ≥80% of panellists rated within ±1 of the 7–9 median, the indicator was considered appropriate with agreement.ResultsThe indicators covered nine clinical conditions in primary care and nine in hospital settings, including respiratory tract infections, diarrhoea, urinary tract infections (UTIs), sepsis and others. There were 102 indicators (Primary Care: 46; Hospital: 39; General: 17) in Round 2 of the Delphi Technique. Of these, 100% were rated appropriate and 99 (97.1%) feasible in a local context. During the RAND/UCLA method, 136 indicators (Primary Care: 56; Hospital: 60; General: 20) were rated in Round 2, with 131 (96.3%) appropriate and 72 (52.9%) feasible in a global context. From these broad sets, 12 indicators from the Delphi Technique and 31 indicators from the RAND/UCLA method were rated both appropriate and feasible with agreement respectively. Most indicators rated appropriate and feasible with agreement by the Delphi panel measured the proportion of patients receiving Access or Watch antibiotics across clinical conditions. Those rated appropriate and feasible with agreement by the RAND/UCLA panel measured appropriate antibiotic choice, dose, duration and the proportion of Access and Watch antibioticsConclusionsThese model AWaRe-based, universally applicable quality indicators can be locally adapted and measured with different tools to improve the optimal use of antibiotics and inform global and country specific antimicrobial stewardship programmes (AMS).

Meningitis and encephalitis (M/E) are medical emergencies with high morbidity and mortality if not promptly diagnosed and treated. The diagnostic gold standard remains CSF culture via lumbar puncture (LP), but this approach is limited by low sensitivity, long turnaround times and a significant proportion of culture-negative cases despite strong clinical suspicion due to antimicrobial treatment prior to sampling. On the Acute Medical take presentations with suspicion of M/E are common. Delays in CSF sampling and the culture thereof can lead to unnecessary hospital admissions, unnecessary antimicrobial use and increased healthcare costs. Molecular diagnostics such as the BioFire® Meningitis/Encephalitis (ME) panel offer rapid, multiplex PCR-based syndromic detection of 14 common bacterial, viral and fungal pathogens directly from CSF within approximately 1 h. This evaluation sought to assess the performance of the BioFire® ME panel at the point of care (POC) in the Acute Medical Unit (AMU) at Royal Berkshire Hospital, and to consider its impact on clinical workflow during the acute take.MethodsOver a 19 month period, 105 patients admitted to the acute medical take with suspected meningitis/encephalitis underwent BioFire® ME panel testing alongside conventional LP with CSF analysis and culture. Patient demographics, length of hospital stay and diagnostic outcomes were collected. Pooled patient demographic data was collected and the results of the BioFire® detected organisms and the culture based methods compared.ResultsThere were 28 positive cases and 77 negative cases. BioFire® detected pathogens in all 28 positive cases of which 15 were viral and 13 bacterial. CSF culture detected bacterial pathogens from 2 cases. Both CSF-positive cases were confirmed by BioFire® as Streptococcus pneumoniae. Five BioFire® tests were invalid; corresponding CSF results were negative. Positive cases mean admission was 16.5 days. Mean admission for viral infections was 13 days and 23 for bacterial infections. S. pneumoniae (n=6, 21%) was the commonest pathogen, Enterovirus and Haemophilus influenzae (n=5, 18% each), Varicella zoster virus and Herpes simplex virus 1 (n=3, 11% each), Human herpesvirus 6 (HHV-6) and Neisseria meningitidis (n=2, 7% each) and Herpes simplex virus 2 and Streptococcus agalactiae (n=1, 4% each). HHV-6 detection may represent latent viral DNA rather than active infection.ConclusionsThe BioFire® ME panel demonstrated markedly superior sensitivity compared with conventional LP culture in this AMU cohort and provided clinically actionable results within the timeframe of the acute medical take. Its rapid turnaround and broad coverage support earlier decision-making, more targeted antimicrobial use and earlier discharge of those with negative results or with viral meningitis. Importantly, the test was feasible to deploy at POC in the AMU, integrating into routine clinical workflow without delaying patient management. While interpretation of certain viral results (e.g. HHV-6) requires caution, this evaluation supports BioFire® as a practical and effective diagnostic tool for meningitis/encephalitis in the acute setting.

Importance  Accurate species identification and antifungal susceptibility testing are essential for effective aspergillosis treatment. However, non–Aspergillus fumigatus species, such as Aspergillus tubingensis, are often misidentified and understudied, potentially compromising proper prognosis and treatment.Objective  To examine the species and prevalence of triazole resistance among clinical Aspergillus isolates in Southern California.Design, Setting, and Participants  This cross-sectional study collected clinical Aspergillus cultures from September 1, 2019, to June 30, 2023, at Kaiser Permanente Southern California, an integrated health system serving a diverse regional population. Triazole susceptibility testing and whole genome sequencing were performed on selected isolates. A total of 2421 consecutive Aspergillus cultures were included. Eighty putative Aspergillus niger isolates were selected for sequencing, including 44 with positive growth in the presence of at least one clinically relevant triazole.Main Outcomes and Measures  The primary outcome was the genome-based species identification of A tubingensis. The secondary outcome was triazole susceptibility above the A niger epidemiological cut-off values. The hypothesis that A tubingensis is a prevalent, underrecognized, triazole-resistant pathogen was developed during data collection.Results  Of 2421 cultures, 1835 were successfully cultured for Aspergillus. After purification and deduplication, 1505 isolates were screened for triazole resistance. A substantial fraction of putative A niger isolates grew at the A niger epidemiological cut-off for itraconazole (110 of 664 [15.1%]). DNA sequencing revealed that 59 of 80 putative A niger isolates (73.8%) were actually A tubingensis. Elevated triazole minimum inhibitory concentrations were not strongly associated with any known cyp51 mutations among the A tubingensis isolates.Conclusions and Relevance  In this cross-sectional study of Aspergillus isolates from Kaiser Permanente Southern California, A tubingensis was a prevalent but underrecognized cause of aspergillosis in Southern California. Its frequent misidentification, association with invasive infections, and triazole resistance underscore the need for improved diagnostics and species-specific epidemiological investigations.

AbstractObjectivesLife-threatening invasive infections caused by Streptococcus pyogenes and Streptococcus dysgalactiae subsp. equisimilis (SDSE) are unpredictable, frequently fatal, and are increasing in incidence globally. In the absence of evidence from randomized controlled trials (RCTs), clinical management for these conditions varies. Understanding current management approaches and areas of clinical equipoise will inform planning of feasible high-impact RCTs.MethodsTwo web-based surveys were distributed to Staphylococcus aureus Network Adaptive Platform trial clinicians in April 2024. Survey 1 addressed clinical management practices for S. pyogenes and SDSE bacteraemia, research question (domain) priorities, and areas of clinician equipoise. ‘Domains’ included backbone antibiotics (e.g. penicillin and ceftriaxone), adjunctive antibiotics (e.g. clindamycin and linezolid), and immunomodulation (intravenous immunoglobulin). Survey 2 collected aggregate annual numbers of blood culture isolates (bacteraemia episodes) for selected streptococci and enterococci from 2019 to 2023 inclusive.ResultsSurvey-1 respondents (n = 65 clinicians from nine countries) ranked S. pyogenes as the highest priority pathogen for inclusion in a future RCT, and ‘adjunctive antibiotics’ as the highest priority domain. However, only 51% (S. pyogenes) and 65% (SDSE) of respondents were willing to randomize participants with severe disease to receive ‘no adjunctive antibiotics.’ For nonsevere disease, 89% (S. pyogenes) and 85% (SDSE) of clinicians were willing to randomize participants to ‘no adjunctive antibiotics.’ Survey 2 (n = 17 hospitals from six countries) found the mean number of blood culture isolates/site/year to be 24 for S. pyogenes, and 27 for SDSE (2019–2023).DiscussionHigh-quality evidence is needed to guide treatment of patients with invasive streptococcal diseases. Ongoing clinician engagement will enhance research question prioritisation and feasibility for future trials.Graphical AbstractDownload: Download high-res image (398KB)Download: Download full-size image

ABSTRACTMycobacterium abscessus (MAB), a rapidly growing non-tuberculous mycobacterium, is becoming increasingly recognized as a significant pathogen affecting humans. These bacteria particularly impact individuals with cystic fibrosis (CF), non-CF bronchiectasis, and compromised immune systems. Treating pulmonary infections with MAB is challenging due to the bacteria’s inherent and acquired resistance to many antibiotics, including most anti-tuberculosis antibiotics. Antibiotic therapy of MAB infection is lengthy, involves multiple oral and parenteral administered drugs, induces significant toxicity, and, on many occasions, fails to cure. Consequently, developing more effective antibiotics has become a high priority. Preclinical studies to evaluate antibiotic efficacy against MAB are challenging because they fail to establish a progressive and sustained pulmonary infection in commonly used animal models. To address this issue, the course of MAB pulmonary infection was evaluated in 15 immunocompetent or deficient mouse strains. We report bacterial burden and histopathology and classify the models according to their ability to clear or sustain progressive infection beyond 28 days. We also examined the potential of these models for drug screening. Our findings provide a foundation for selecting suitable mouse models of pulmonary MAB infection for drug discovery.

:Febrile neutropenia (FN) is the most common and serious adverse event of chemotherapy for solid and hematological neoplasms, with infection as a major complication. FN occurs in 10%–50% of patients with solid tumors and over 80% with hematological malignancies, with mortality rates up to 11%.Objective:To characterize bloodstream pathogens in post-chemotherapy FN through a systematic review and meta-analysis of studies published from 2013 to February 13, 2024.Design:Systematic review and meta-analysis.Data sources and methods:PubMed, Web of Science, Scopus, and Embase were searched using MeSH, Emtree, and keywords. Risk of bias was assessed using the JBI checklist. Random-effects proportions meta-analysis was performed.Results:Twenty-two studies (n = 23,319) reported 8665 positive blood cultures: 59% Gram-negative (95% CI: 46.8–67.5), 39.7% Gram-positive (95% CI: 31.3–48.2), and 2.5% fungi (95% CI: 0.9–4.1). Random-effects meta-analysis showed high heterogeneity (I2 = 98.92%, p < 0.01). Meta-regression by sample size, economic development, and risk of bias did not explain this variability.Conclusion:Gram-negative pathogens slightly predominate over Gram-positives in bloodstream infections among post-chemotherapy FN patients.Trial registration:PROSPERO (CRD42023472191).

Disinfectants are antimicrobial agents frequently used in healthcare settings. Bacterial susceptibility against a wide range of antimicrobial agents can be changed due to prolonged exposure to sub-inhibitory concentrations of disinfectants.ObjectivesTo investigate the adaptive responses of hospital-acquired MRSA (HA-MRSA) to long-term exposure to alkylating agents such as Glutaraldehyde (GA) and Formaldehyde (FA). In a 25 day serial passaging, isolates were exposed to different inhibitory concentrations. Bacterial isolates were selected based on genetic diversity across the clonal complexes (CCs) and antibiotic resistance profiles. MICs were conducted before and after passaging for all isolates, indicating consistent MIC values for FA and GA in the WT strain.ResultsAfter 25 days, three different strains evolved in GA showed shifting in the MIC values, indicating lower susceptibility for GA. In addition to the MIC analysis, efflux pump activity and biofilm formation capabilities were evaluated for all isolates prior to and post the long-term exposure. Different efflux pump activities were observed with WT samples; however, evolved isolates, belonging to CC22, showed higher efflux activity in response to GA and FA. Additionally, isolates were classified based on their biofilm formation abilities. WT strains exhibited a wide range of forming abilities. Among evolved isolates, all screened ones were prone to changes in their formation abilities; however, the highest changes were more related to FA.ConclusionsThese findings highlight that evolutionary responses are difficult to predict, as they vary depending on multiple factors such as the type of stressor, the CC of the isolate, and its genetic background. For understanding the underlying mechanisms driving these changes, further studies are required, including cross-resistance profiling for different antibiotics, biofilm inhibition assays, qPCR and mutation repair assays. Additionally, sequencing and genetic analysis of the evolved samples to identify SNPs and other mutations responsible for these tolerance capabilities.

Pseudomonas aeruginosa is a significant pathogen of wound infections with a high propensity for MDR, making it a challenging organism to treat. Although resistance rates of individual antibiotics are usually described, the relationships between resistances of various drug classes and co-occurrences of resistances are less understood, specifically in low- and middle-income countries. We assessed both resistance prevalence and co-resistance structure in a tertiary-care setting in Bangladesh to inform empirical therapy and stewardship.ObjectivesTo quantify resistance and MDR prevalence, map statistically robust co-resistance and hub antibiotics, and translate findings into pragmatic empirical stewardship signals.MethodsWe conducted a retrospective cross-sectional study of 113 non-duplicate P. aeruginosa wound isolates at Khulna Medical College Hospital, Bangladesh (Aug 2018–Dec 2022). Susceptibility was tested using Kirby–Bauer disc diffusion interpreted per CLSI. Binary resistance to eight antipseudomonal antibiotics was correlated pairwise (Spearman’s ρ) with bootstrap 95% CIs; multiplicity was controlled with Benjamini–Hochberg FDR (q<0.05). A co-resistance network linked antibiotic pairs with significant |ρ|≥0.30; degree and betweenness centrality defined hubs.ResultsOverall, 69.0% (78/113) of isolates were MDR. Resistance was highest to third- and fourth-generation cephalosporins: ceftazidime 85.0% and cefepime 81.4%. Resistance was lower for imipenem 23.9%, piperacillin/tazobactam 34.5% and amikacin 38.1%. Strong within-class co-resistance was observed for cefepime–ceftazidime (ρ=0.63; q<0.001) and amikacin–gentamicin (ρ=0.61; q<0.001). Cross-class links included ciprofloxacin–imipenem (ρ=0.48; q<0.001) and ciprofloxacin–piperacillin–tazobactam (ρ=0.45; q<0.001); moderate correlations were seen for ciprofloxacin–cefepime (ρ=0.33) and ceftazidime–piperacillin–tazobactam (ρ=0.31). The network was non-random (permutation P<0.001) with moderate clustering. Ciprofloxacin was the dominant hub (degree 6; betweenness 0.42), followed by gentamicin (degree 5; betweenness 0.31) and imipenem (degree 4; betweenness 0.28). Clinically, 89.2% of ciprofloxacin-resistant isolates were MDR, reinforcing the risk of empirical ciprofloxacin monotherapy. Conversely, 41.6% of isolates remained jointly susceptible to amikacin plus piperacillin–tazobactam, consistent with these agents’ peripheral network positions.ConclusionsIn this high-burden setting, MDR P. aeruginosa is common, and resistance to cephalosporins is very high. A correlation-driven network highlights ciprofloxacin, imipenem and gentamicin as connectivity hubs that co-fail with multiple partners, whereas amikacin and piperacillin–tazobactam are more peripheral. Two stewardship-ready signals emerge: 1) avoid hub antibiotics for empirical monotherapy when severe wound infection is suspected, and 2) consider peripheral combinations such as amikacin plus piperacillin–tazobactam pending culture. Prospective, multi-centre validation with molecular profiling is warranted to confirm mechanisms and test the outcome impact of network-guided prescribing.