Impact of molecular point-of-care testing on surveillance and public health management of invasive meningococcal disease.

Predictive surveillance and diagnosis of COVID-19: An integrative machine learning and wastewater multi-omics approach.

Treeshrews as a potential reservoir: First detection of dengue virus serotype 2 in Malaysian treeshrew faeces.

France implemented wide-ranging responses to control SARS-CoV2 transmission during the COVID-19 pandemic. Some measures targeted specific social groups considered to be ‘vulnerable’. Yet how diverse populations experienced and shared existing vulnerabilities within these public health responses remains poorly understood. Theoretical grounding of the present analysis relies on Florencia Luna"s ‘layers’ of vulnerability and David Napier and Anna Volkmann’s ‘vulnerability vortex’. Between January and July 2021, the research team led 156 ethnographic interviews among diverse participants residing in the Paris-Seine-Saint-Denis area and the Vendôme territory. We conducted thematic analysis of these interviews to elucidate the dynamic, situational experiences of vulnerability. We present in-depth cases of three participants, elaborating how differently situated people coped with the pandemic and its effects and showing that access to public services or informal forms of support was predicated on existing and dynamic forms of vulnerability, which came sharply into focus during the pandemic. Categorical boundaries between groups, we argue, were not useful in understanding some root causes of vulnerability. Rather, the consideration of existing and emerging cross-cutting forms of precarity and marginality was useful for understanding how individuals coped with the pandemic.

Accurate identification of recent HIV-1 infections is critical for real-time epidemic monitoring. However, conventional Limiting Antigen Avidity Enzyme Immunoassays (LAg-EIAs) are restricted to laboratory settings. A novel rapid recency test based on the limiting antigen avidity principle was developed for point-of-care use. We evaluated the performance of the rapid HIV-1 recency test using 500 longitudinal plasma specimens from 107 seroconverters. The mean duration of recent infection (MDRI) was estimated via binomial regression with maximum likelihood modeling. The false recent rate (FRR) was assessed using samples from individuals with long-term infection, including those on antiretroviral therapy (ART). Concordance was compared with two commercial LAg-EIA kits (Maxim and KingHawk). The rapid assay yielded an MDRI of 123 days (95% CI: 87-138), shorter than Maxim (152 days, 95% CI: 137-172) and KingHawk (131 days, 95% CI: 107-140). Among ART-naïve individuals infected for over 1 year, FRR was 5.3%, similar to Maxim (5.9%) and slightly higher than KingHawk (2.7%). High concordance was observed with Maxim (93.1%, kappa = 0.743) and KingHawk (89.4%, kappa = 0.558). In ART-treated individuals, FRR was significantly higher in the early ART group (69.9%) compared to the late ART group (20.2%, P < 0.001). The novel rapid HIV-1 recency assay demonstrates acceptable MDRI and FRR and strong agreement with commercial LAg-EIA kits. Its simplicity and rapid turnaround make it a promising tool for decentralized surveillance and targeted HIV interventions, especially in resource-limited settings.IMPORTANCERapid detection of recent HIV-1 infections is essential for monitoring ongoing transmission and guiding targeted prevention efforts. However, currently used laboratory-based recency assays require specialized facilities and trained personnel, limiting their use in decentralized or resource-limited settings. In this study, we evaluated a newly developed rapid test that identifies recent HIV-1 infections within minutes using a simple, instrument-free format. The test showed strong agreement with two widely used laboratory assays and demonstrated performance suitable for surveillance applications. Its ease of use, rapid turnaround, and minimal infrastructure requirements make this rapid test a practical tool for expanding real-time HIV monitoring and improving the efficiency of public health responses.

ABSTRACTBackground and AimsAs of late 2023, an estimated 39.9 million people are living with HIV, placing strain on healthcare systems. Machine learning (ML), a branch of artificial intelligence, enables systems to improve performance through data‐driven learning without explicit programming. HIV prognosis is influenced by clinical, epidemiological, and psychosocial factors, and ML algorithms have the potential to integrate these determinants efficiently. This can provide valuable insights into disease progression and risk assessment in terms of viral load, CD4 cell count, treatment initiation, treatment adherence, hospitalization, acquired immunodeficiency syndrome diagnosis, quality of life and mental health. This protocol outlines the existing applications of ML to prognostic modeling in the context of HIV, highlighting how ML can equip physicians with rapid and accurate predictions of disease progression, thereby informing treatment decisions such as clinical prescriptions and social support plans, and optimizing patient outcomes.MethodsThe protocol follows the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses Extension for Scoping Reviews (PRISMA‐ScR) framework. A search strategy has been developed for Medline (PubMed) and will be adapted for searches in Embase, Web of Science, Scopus, IEEE Xplore, and ACM Digital Library. The study selection and data extraction will be conducted in duplicate. The methods for the scoping review are prespecified to ensure transparency.DiscussionThe proposed scoping review will identify effective model types, data inputs, and applications of ML in the context of HIV prognosis. While ML has been integrated into various aspects of HIV research, few studies have focused on predicting prognosis. This review aims to synthesize current uses of ML in prognostic modeling and highlight gaps within the existing technology. The findings from this review will support the development of future ML models that can inform clinical decision‐making, and, in turn, optimize patient care, improve resource allocation, and enhance public health responses to the ongoing HIV epidemic.

Over a 14-day period, four patients developed sepsis within hours of undergoing endoscopic procedures at a gastroenterology outpatient clinic in Geneva (Switzerland), triggering an urgent epidemiological investigation upon notification to authorities. The case clustering raised concerns about a iatrogenic source, prompting a coordinated public health response. This report outlines the investigation led by the Geneva Health Authorities in collaboration with the medical team, emergency care providers, infection prevention specialists, and the bacteriology laboratory at Geneva University Hospitals. A multidisciplinary team, including an epidemiologist, a pharmacist and an infection prevention specialist, conducted an on-site investigation the day after the outbreak was identified, highlighting a probable extrinsic contamination of propofol by Escherichia coli.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13756-025-01655-x.

Seasonal influenza remains a major global health concern, causing substantial morbidity and mortality each year. Early indicators of transmission are critical for timely public health interventions, and reverse transcription PCR (RT-PCR) cycle threshold (Ct) values, reflecting viral load, have recently been proposed as a proxy for community-level trends. In this study, we analyzed national influenza surveillance data from Iran between March 21, 2023, and March 19, 2024. Weekly mean Ct values were calculated from RT-PCR tests of suspected influenza cases, and weekly confirmed case counts were extracted from national reports. Time series patterns were visually inspected, and stationarity was assessed using the Augmented Dickey-Fuller test. Cross-correlation analysis was performed to identify lags between Ct values and incidence. ARIMA and ARIMAX models were fitted to evaluate and predict influenza incidence using Ct as an input, while VAR models were applied to characterize dynamic relationships. Analyses were conducted in Stata version 14. The mean Ct value over the one-year period was 27.16 ± 4.68. Cross-correlation showed that Ct values preceded incidence by about four weeks (r = 0.41, p = 0.003). ARIMA (0,1,1) best fitted Ct values, while ARIMA (2,1,2) best fitted influenza incidence. In ARIMAX, Ct values lagged by four weeks were a strong predictor of weekly incidence (β = 6.29, p < 0.001). These findings indicate that population-level Ct values can serve as an early signal of rising influenza activity, potentially allowing prediction of incidence trends up to four weeks in advance and supporting timely public health responses.

The emergence of novel psychoactive substances (NPSs) has overwhelmed forensic, health care, and regulatory systems. Conventional analytical techniques are ineffective for identifying known compounds but fail against newly synthesized NPSs lacking reference standards. This review explores the roles of artificial intelligence (AI) and machine learning in addressing growing challenges in NPS identification and characterization. The authors reviewed the current forensic workflows and the integration of AI-based approaches, including deep learning models, chemical language models, and spectral prediction tools. Particular emphasis was placed on the DarkNPS framework, which uses Long Short-Term Memory networks and SMILES-based data augmentation to generate millions of plausible NPS structures, and on spectral prediction tools, such as Competitive Fragmentation Modeling for Metabolite Identification (CFM-ID) and novel psychoactive substances-mass spectrometry, for in silico MS/MS spectra generation. Additional emerging AI technologies, such as transformers, graph neural networks, and multimodal frameworks, were also examined. AI-based systems significantly reduced the time and resources required for NPS identification by enabling structure generation, spectral prediction, and prioritization without physical standards. The DarkNPS model successfully predicted structures for >8.9 million plausible NPS compounds, with over 90% of the future market NPS accurately anticipated. In silico MS/MS spectral libraries built using AI tools demonstrated high cosine similarity scores (>0.7) with the experimental spectra, allowing top-hit identification in 75%-90% of the cases. This improved efficiency can facilitate more accurate diagnoses, guide timely treatment decisions, and support public health responses to emerging NPS threats. Integrating AI with traditional analytical chemistry significantly enhanced the speed, scope, precision, and utility of NPS identification, marking a promising shift in forensic toxicology and chemical surveillance.

Effective communication and public engagement are essential components of infectious disease control, yet they remain underdeveloped in the field of immunology. This review explores how immunologists and scientists can contribute to countering misinformation and improving vaccine uptake through inclusive, culturally sensitive engagement. Drawing on historical and contemporary case studies, we examine how trust, cognitive biases, and community involvement shape public responses. We highlight the importance of co-produced messaging and the role of community champions in building trust, particularly among marginalized groups. Vaccine communication is analyzed through the lens of the five Cs: confidence, complacency, convenience, communication, and context. We discuss how demographic and structural barriers, historical mistrust, and politicization of health messaging contribute to declining vaccine uptake and propose tailored strategies to address these challenges. The final section focuses on data presentation as a core foundation of public communication, emphasizing that clarity, transparency, and ethical framing are critical to public understanding. We outline principles for designing trustworthy visuals, mitigating cognitive biases, and embedding context directly within graphics to prevent misinterpretation. Participatory approaches to data communication are shown to improve comprehension and trust, especially when co-developed with affected communities. Together, these domains-engagement, vaccine communication, and data presentation-form a foundation for resilient public health responses. By integrating immunological expertise with inclusive communication strategies, scientists can play a central role in fostering informed decision making and strengthening public cooperation in future outbreaks.

BackgroundIn recent years, over-the-counter (OTC) drug sales have emerged as a novel indicator for symptom monitoring, attracting widespread attention in public health research globally. This study conducted weekly monitoring of five OTC drug categories related to fever respiratory system diagnostic cluster (FRSDC) — antitussive/expectorant drugs, cold medications, antibiotics, pungent and cool exterior-relieving agents, and influenza medicine — in Pinghu City, Zhejiang Province, from 2022 to 2024. Concurrently, weekly FRSDC cases from Pinghu First People’s Hospital were collected.MethodsSpearman correlation analysis was used to quantify associations between OTC sales and FRSDC cases, while decision tree models evaluated the reliability of OTC data for early prediction of FRSDC trends.ResultsResults showed significant positive correlations between all five OTC drugs and FRSDC cases, with synchronous seasonal peaks in winter and spring (Spearman’s correlation coefficients ranged from 0.36 to 0.80, all p value ≤ 0.0001). Even when OTC drug sales preceded FRSDC cases by one or two weeks, strong correlations persisted (Spearman’s correlation coefficients ranged from 0.28 to 0.79, p value ≤ 0.0001). Decision tree analysis revealed that combining antitussive/expectorant drugs and influenza medications effectively predicted FRSDC epidemics with 83.33% accuracy (adjusted p value < 0.05).ConclusionThese findings suggest that monitoring OTC drug sales may serve as a useful early warning indicator for FRSDC, potentially aiding public health response and resource planning.

Background This study analyzed influenza-like illness (ILI) and influenza-positive rate (IPR) in Quzhou from 2016 to 2024 to assess the impact of the COVID-19 pandemic. Methods The weekly surveillance data of ILI were collected from sentinel hospitals and network laboratories. Real-time reverse transcription-polymerase chain reaction (RT-PCR) was used to detect and identify the specific types of influenza viruses. Results A total of 77,912 ILI cases were reported (ILI% = 3.9%). Both ILI% and IPR showed seasonal winter-spring peaks. A marked decline in ILI% (4.7% to 2.7%) and IPR (19.2% to 10.7%) was observed during the COVID-19 pandemic period (2020–2022), followed by a post-pandemic rebound. Children under 15 years old accounted for the highest proportion of ILI cases (65.9%). The 5–14 age group had the highest IPR (33.6%). Influenza A(H3N2) (39.2%) and B/Victoria (36.0%) were the predominant strains. The correlation between ILI% and IPR varied significantly across COVID-19 pandemic phases: weak positive pre-pandemic (rs = 0.325), moderate negative during the pandemic (rs = −0.597), and moderate positive post-pandemic (rs = 0.584). Age-stratified analysis revealed distinct correlation patterns in different COVID-19 pandemic phases, with the strongest correlation in the 15–59 age group (rs = 0.56, p = 0.004). Conclusion The COVID-19 pandemic has significantly influenced influenza activity and altered the relationship between ILI% and IPR. The inconsistent correlation between ILI and IPR, especially during the COVID-19 pandemic, highlights the necessity of integrated virological surveillance and appropriate definition of ILI for effective influenza monitoring and public health response.

Non-O1, non-O139 Vibrio cholerae is an uncommon cause of pneumonia, particularly following freshwater exposure. Non-O1, non-O139 Vibrio cholerae was identified from bronchoalveolar lavage through culture and quantitative polymerase chain reaction (qPCR) in Hungary. During an epidemiological investigation, the source of infection was traced to a designated bathing site at a lake in Central Hungary, where non-O1,non-O139 Vibrio cholerae was isolated from surface water. We conducted whole-genome sequencing and comparative genomic analysis on a clinical isolate (N = 1) and three phenotypically distinct environmental isolates (N = 3). In addition, we reviewed the available literature on pulmonary infections associated with Vibrio cholerae. Core genome multilocus sequence typing (cgMLST) revealed that the clinical and environmental isolates clustered together with zero allelic differences. Multilocus sequence typing (MLST) identified a new sequence type (ST1605), representing a novel combination of known allele variants. In silico analysis of antibiotic resistance genes identified the presence of blaCARB-7. Both the clinical and environmental isolates exhibited identical virulence gene profiles, reinforcing the hypothesis that the infection was acquired from a local water source. This study represents the first investigation of a primary pulmonary Vibrio cholerae infection reported in Europe following a near-drowning event. While Vibrio vulnificus and Vibrio metschnikovii have been implicated in similar pneumonia cases, the precise virulence mechanisms of these species remain poorly understood. Although non-O1, non-O139 Vibrio cholerae infections associated with recreational water exposure are rare in Hungary (1-2 cases per year), this study underscores the importance of ongoing surveillance for the detection of potential outbreaks and to inform public health responses.

BackgroundThis study investigates the severity of COVID-19 infection among pregnant women during three distinct waves of the pandemic in Ahvaz, Iran. Considering the serious impact of COVID-19 on maternal and neonatal health, this research aimed to compare the clinical characteristics and disease severity across these waves.MethodsThis was a retrospective observational cohort study conducted to identify the clinical features and outcomes of pregnant women with PCR-confirmed SARS-CoV-2 infection during three pandemic waves in Ahvaz. Data including demographic variables, disease severity (classified according to WHO clinical criteria), complications, and delivery outcomes were extracted from hospital medical records between February 2020 and December 2021.ResultsA total of 267 pregnant women with confirmed COVID-19 were analyzed across three pandemic waves. In the third wave, the mean maternal age was 30.58 years, mean gravida 2.61, and mean parity 1.23. The average gestational age at admission was 35.49 weeks, and mean BMI was 23.46 kg/m². The mean duration of hospitalization was 7.01 days, and average recovery time 20.09 days. Compared with earlier waves, the third wave showed a higher frequency of positive CT findings, fever, cough, and shortness of breath, indicating greater disease severity.ConclusionThe findings demonstrate changing patterns in the clinical presentation and severity of COVID-19 among pregnant women across successive pandemic waves. These differences may reflect evolving viral characteristics and public health responses. The study underscores the need for tailored maternal care strategies and ongoing surveillance to manage COVID-19 and its potential long-term effects in this vulnerable population.

The spillover of H5N1 clade 2.3.4.4b into dairy cattle has raised concerns over the safety of fluid milk. While no foodborne infection has been reported in humans, this strain has infected at least 70 people and milk from infected cows is known to be infectious by ingestion in multiple other species. Investigation into the public health threat of this outbreak is warranted. This farm-to-table quantitative microbial risk assessment (QMRA) uses stochastic models to assess the risk of human infection from consumption of raw and pasteurized fluid cow’s milk from the United States supply chains. These models were parameterized with literature emerging from this outbreak, then employed to estimate the H5N1 infection risk and evaluate multiple potential interventions aimed at reducing this risk. The median (5th, 95th percentiles) probabilities of infection per 240-mL serving of pasteurized, farmstore-purchased raw, or retail-purchased raw milk were 7.66E-19 (2.39E-20, 4.02E-17), 1.56E-7 (6.67E-10, 1.28E-5), and 1.40E-7 (6.65E-10, 1.13E-05), respectively. Our results confirm that pasteurization is highly effective at reducing H5N1 infection risk. Scenario analysis revealed quantitative real-time reverse transcriptase-polymerase chain reaction (qrRT-PCR) testing of bulk tank milk to be an effective method for numerically reducing risk from raw milk. Additionally, we identify knowledge gaps related to human H5N1 dose-response by ingestion and raw milk consumption patterns. These findings emphasize the importance of mechanistic epidemiologic models for informing public health responses amidst outbreaks with foodborne potential and highlight the need for additional research into raw milk consumption patterns to better understand this exposure pathway.

BackgroundSnakebite envenomings (SBE) are an important and neglected health issue due to their frequency and potential for severe clinical outcomes. Envenomations can cause local and systemic complications, depending on the snake species, amount of venom injected, comorbidities, timing and use of antivenom, and access to health care. Systemic effects may be fatal or lead to permanent sequelae, including strokes resulting from venom-induced vascular and tissue damage. The objective of this study is to investigate the main clinical and epidemiological characteristics of individuals who developed stroke following SBE and to identify predictors of death.Methodology/principal findingsWe conducted a systematic review and individual patient data meta-analysis using a predefined search strategy across MEDLINE/PubMed, LILACS, and SciELO databases, following PRISMA guidelines. A total of 100 studies were included, predominantly case reports and case series, comprising 130 individuals with stroke following SBE. Most patients were male (62.3%) and aged between 40 and 59 years (37.7%). Viperids caused 96.4% of the snakebites, particularly Daboia russelii and Bothrops spp. Most patients (90%) received antivenom therapy. Reported cases of snakebite-related stroke originated from 22 countries, mostly from India (36.9%), Brazil (13.9%) and Sri Lanka (10.8%). Ischemic strokes were more common than hemorrhagic strokes (61.5% vs. 38.5%), and multifocal brain involvement was predominant in both stroke types. Overall case-fatality was 23.4%. Sepsis [OR=6.21 (1.35-33.47); P = 0.001] and thrombocytopenia [OR=3.97 (1.66-10.03); P = 0.02] were predictors of deaths. Hemorrhagic stroke [OR=2.67 (1.15-6.31); P = 0.02], multiple brain lesions in a single hemisphere [OR=7.57 (2.33-33.39); P < 0.001], and subarachnoid hemorrhage [OR=7.00 (1.87-29.4)); P = 0.001] significantly increased the risk of death. Motor sequelae remained the most common long-term outcome (22.4%), occurring significantly more often in ischemic stroke survivors (28.8% vs. 9.4%, P = 0.05). Autopsy findings revealed intense brain alterations generally in parallel with damage in other organs such as the kidneys, lung, and heart.Conclusions/significanceStrokes from SBE represent a potential medical emergency in low- and middle-income countries where snakebites predominate, and lead to high rates of mortality and long-term disability. Recognizing stroke as a disabling and underreported consequence of snakebite is essential for improving clinical outcomes and guiding public health responses. Integrating the knowledge on predictors of death from SBE-relate strokes into health policies will be vital for reducing long-term morbidity and advancing disability-inclusive strategies.

Two methods to estimate the population size of people who inject drugs in the country of Georgia: implications for the EECA region.

This article examines historical polio outbreaks in three Dutch towns (1963, 1966, 1971) to show how vaccination refusal became an expression of and contribution to local solidarity shaped by religion, place, and tradition. In doing so, it demonstrates how medical history contributes to ongoing conversations in medical humanities about vaccination refusal. I argue that refusal was neither simply resistance nor misunderstanding but a deliberate act that reaffirmed local community boundaries. Drawing on archival research, I explore how public health interventions and national media scrutiny made previously unnoticed communities visible, fostering their collective self-awareness and sense of distinctiveness. Following Anna Tsing, I describe the productive yet uneasy interaction between national public health practices and local ways of living as an example of "friction." Historical analysis reveals how friction during vaccination campaigns brought these communities into public view, highlighting tensions between collective responsibility for public health and respect for traditional, place-specific ways of living. By emphasizing the spatial dimensions of refusal, I suggest that effective public health interventions may benefit from greater sensitivity to local cultural contexts.

Dengue fever has become a major global public health challenge due to its rapidly in-creasing incidence. Rapid on-site detection of dengue virus (DENV) is critical for early diagnosis, timely patient isolation, and outbreak control. In this study, dengue virus serotype 2 (DENV-2), the predominant strain circulating in tropical and subtropical regions, was selected as the target pathogen. We established a one-tube rapid detection assay that integrates the HUDSON nucleic acid extraction-free protocol, reverse transcription recombinase-aided amplification (RT-RAA), and CRISPR/Cas12a-mediated trans cleavage activity. The method achieved a detection limit of 1 × 102 copies/μL for simulated infected samples and exhibited no cross-reactivity with other DENV serotypes (DENV-1, DENV-3, DENV-4) or with other arboviruses, including Zika, Japanese encephalitis, yellow fever, and chikungunya viruses. The assay demonstrated high sensitivity and specificity across various sample types, including mosquitoes, rodents, blood, and cultured cells, with results consistent with quantitative PCR (qPCR). Requiring only basic equipment such as a water bath, the system enables on-site detection of DENV-2 within 1 h. This simple, cost-effective, and reliable assay provides a practical tool for field-based DENV-2 surveillance and supports effective public health responses in resource-limited settings.

Innovations in aptamer-based biosensors for detection of pathogenic bacteria: Recent advances and perspective.