Community-acquired pneumonia (CAP) remains an important public-health problem, and the COVID-19 pandemic and non-pharmaceutical interventions (NPIs) may have altered its burden. This study aimed to provide updated CAP burden among adults in Shanghai from 2016–2023.We analysed 61,230 participants aged 20–74 years from the Shanghai Suburban Adult Cohort and Biobank. CAP episodes were ascertained via ICD codes and clinical diagnoses. We calculated incidence rates before, during, and after NPIs, conducted subgroup analyses by age, sex, comorbidity and lifestyle. We used Poisson regression to compare stages, and Cox models to identify risk factors. The Overall CAP incidence was 42.1 per 1,000 person–years (95% CI 41.3–42.8). Incidence declined during NPIs (24.2/1,000 py) and rose after NPIs (95.9/1,000 py). The inpatient-to-outpatient ratio increased to 10.1% during NPIs and fell to 5.7% post–NPI. Among those without underlying conditions, rates were 40.1, 20.1 and 73.6/1,000 py before, during and after NPIs. Incidence was higher in participants ≥60 years and in those with multiple comorbidities, especially respiratory diseases. CAP burden temporarily fell during NPIs but resurged post–NPI, notably among high–risk groups. These findings highlight the need for targeted preventive strategies and continued CAP surveillance in the post-pandemic era.

Orientia tsutsugamushi, the causative agent of scrub typhus, is endemic to the Asia–Pacific region. In South Korea, the Boryong strain is considered dominant; however, nationwide phylogeographic distribution and genetic diversity based on clinical isolates remain incompletely characterized. In this study, 121 O. tsutsugamushi clinical isolates were collected from scrub typhus patients at 11 hospitals across South Korea between 2015 and 2024. Isolates were genotyped using 56-kDa gene sequencing and multilocus sequence typing (MLST) of seven housekeeping genes. Sequence analysis and phylogenetic reconstruction were performed using BLAST, PubMLST, BURST, MEGA11, DnaSP6, and R-based tools. Five 56-kDa genotypes were identified: Boryong (93.4%), Ikeda, Je-cheon, Young-worl, and Yeo-joo. MLST revealed 11 sequence types (STs), including five novel STs. While the Boryong strain and related STs were distributed nationwide, minor strains showed restricted distribution in northern regions. Several isolates sharing the same 56-kDa genotype exhibited different MLST STs, indicating possible recombination or local microevolution. This study provides the first nationwide MLST-based characterization of O. tsutsugamushi in South Korea and demonstrates the dominance of the Boryong strain alongside localized diversity. Our findings underscore the utility of MLST for higher-resolution typing and support the need for continued molecular surveillance to inform regional epidemiology and disease management.

This study examined whether coronavirus disease 2019 (COVID-19) infection experience enhances preventive behaviour (i.e., hand disinfection and mask-wearing), with risk perception acting as a mediating factor. The study included participants aged ≥18 years residing in Japan, enrolled in a 30-wave cohort study conducted from January 2020 to March 2024. Using propensity score matching, 135 pairs of participants with and without infection were extracted, adjusting for dread and unknown risk perception, preventive behaviours, sociopsychological variables, and individual attributes. Comparisons of risk perception and preventive behaviour were made between groups post-infection experience, and mediation analysis was conducted to test whether risk perception mediated the effect of infection experience on preventive behaviour. Following the infection experience, participants in the infection group reported significantly higher scores for one item of unknown risk perception and a greater proportion of mask-wearing. The indirect effect of infection experience on mask-wearing, mediated by the unknown risk perception item, was significant. COVID-19 infection experience increased perceptions of unknowable exposure, which in turn promoted mask-wearing behaviour. Incorporating insights from personal infection experiences into public health messaging may enhance risk perception and promote preventive behaviour among non-infected individuals, offering a novel approach to infection control at the population level.

A first-order Gaussian autoregressive model is considered. The exact finite-sample joint density of the minimal sufficient statistic is derived, for any value of the autoregressive parameter. This allows us to derive explicitly the exact density of the autocorrelation coefficient and its Studentized t-ratio, whose densities were available only in the asymptotic case and not for all values of the parameter and the statistic. This article also demonstrates how to solve a general problem in statistical distribution theory (well beyond the specific case of autoregressive models), that of inverting confluent characteristic functions in multiple variables.

Accidental escapes of pathogens from laboratories continue to cause outbreaks in the community today, posing significant risks to the general public, animal communities and the environment. These incidents, as well as the uncertainties surrounding the origins of the COVID-19 pandemic, highlight the need to consider unnatural origins as part of emerging outbreak surveillance and detection. Identifying recurring patterns and distinctive factors of laboratory-associated disease outbreaks can aid in successfully preventing and mitigating these occurrences. Seventy incidents of laboratory-associated leaks that led to outbreaks in the wider public have been reported (Supplementary Appendix S1). Seven renowned cases that have been comprehensively studied were selected for review: (i) 1955 Polio vaccine incident in western USA, (ii) 1977 H1N1 influenza virus re-emergence in China and the Soviet Union, (iii) 1979 Anthrax release in Sverdlovsk, Soviet Union, (iv) 1995 Venezuelan equine encephalitis epidemics in Venezuela and Colombia, (v) 2003–4 SARS-CoV-1 escapes from Singapore, Taiwan and China, (vi) 2007 Foot-and-Mouth disease virus outbreak in Pirbright, England and (vii) 2019 Brucella leak in Lanzhou, China. These outbreaks were selected because data on their geographical spread, genetics, phylogeny, epidemiological factors (including attack rates, infectious dose, time, location and season of spread) and governmental and institutional responses to the incidents had been previously analysed and published. Thematic analysis of these lines of evidence revealed seven recurring insights described in historically confirmed laboratory-associated outbreaks: unusual strain characteristics, peculiar clinical manifestations or affected demographics, unusual geographical features, atypical epidemiological patterns, delayed government action and communication to the public, misinformation and disinformation spread to the public and biosafety concerns/incidents predating the event. The outbreaks exhibited between 13 and 19 retrospectively identified indicators. These indicators were used to develop preliminary risk criteria intended to support structured, hypothesis-generating assessment of outbreaks, rather than to establish origin.

The activity of respiratory viruses (RVs) displays large variability in tropical regions, posing challenges for public health response strategies. Data from most RVs in south-eastern Mexico remain limited, particularly in the Yucatan Peninsula, the largest tourism hub in the country. This retrospective study analyses the regional epidemiology of RVs in Merida, the largest city in the region, using laboratory test data from a local hospital (January 2018–April 2024). Test results of 143292 RVs were collected, including 121976 for SARS-CoV-2, 19355 for influenza A and B viruses, and 1961 for 17 distinct RVs. We found that non-SARS-CoV-2 RVs circulated year-round, with higher activity in autumn and spring, while SARS-CoV-2 peaked in summer and winter. Influenza A virus, respiratory syncytial virus, and influenza B virus reached their highest activity in autumn, earlier than in other regions of Mexico. Human metapneumovirus peaked during autumn-winter. Rhinovirus/enterovirus and parainfluenza showed year-round activity, with peaks in autumn and spring. Other coronaviruses were more frequent during winter-spring. In post-pandemic years (2022–2023), adenovirus outbreaks emerged, as well as an increased prevalence of non-SARS-CoV-2 RV co-infections. This study highlights the need for region-specific public health strategies, including optimized vaccination schedules, such as for influenza A virus, and enhanced diagnostic surveillance.

This article presents novel methods and theories for estimation and inference about parameters in statistical models using machine learning for nuisance parameter estimation when data are dyadic. We propose a dyadic cross-fitting method to remove over-fitting biases under arbitrary dyadic dependence. Together with the use of Neyman orthogonal scores, this novel cross-fitting method enables root-n consistent estimation and inference robustly against dyadic dependence. We demonstrate its versatility by applying it to high-dimensional network formation models and reexamine the determinants of free trade agreements.

Nationwide screening for parvovirus B19 among blood donors in Hungary has been conducted since 2019. Although B19 is primarily transmitted via the respiratory route, transfusion-related transmission also occurs. This study investigated the impact of COVID-19–related restrictions on B19 incidence. Between January 1 2019 and December 31 2024, a total of 2,043,119 blood donations were screened for B19 DNA using PCR, and the study period was divided into six epidemiological phases.

During the pre-restriction period (Phase I), B19 incidence was relatively low (0.87/10,000 donations). Following the introduction of COVID-19 restrictions (Phase II), highly viremic donations were not detected. Incidence gradually returned in Phase III (0.22/10000) and increased in Phase IV (1.96/10000), suggesting a minor outbreak. A marked surge in December 2023 (23.03/10000) initiated a nationwide epidemic, peaking in March–April 2024 (46.01/10000), before declining by August (Phase VI; 0.54/10000).

COVID-19 restrictions substantially reduced B19 transmission and may have led to increased population susceptibility. This likely contributed to the unusually intense B19 epidemic observed in 2024, which was considerably more severe than contemporaneous outbreaks reported in other countries.

Hyderabad, the fourth-most populous city in India, accounts for the majority of people living with human immunodeficiency virus (HIV) (PLWH) in Telangana, likely comprised of two populations with a disproportionately high national HIV prevalence: gay, bisexual, and other men who have sex with men (MSM) and those who engage in sex work (SW). Research has shown that engaging in SW increases vulnerability to HIV transmission risk for both women and MSM, but less is known about contributors to non-optimal (ART) adherence. We analyzed data from 45 MSM and 49 women living with HIV who were enrolled in the first year of data collection from an mHealth education study in Hyderabad. Modified Poisson regression was used to measure factors associated with ART adherence measured with a visual analogue scale (VAS) (model 1) and pill count (model 2). Less than half (40.9%) reported ever engaging in SW, including 13 women and 25 MSM. The prevalence of non-optimal ART adherence was 14.9% with VAS and 42.4% with pill count. Engaging in SW was not associated with non-optimal ART adherence. Differences in non-optimal ART adherence measured by VAS and pill count suggest that future studies should utilize both methods to better distinguish the measures.

We aimed to describe the evolution of gonorrhea infection and its antimicrobial resistance patterns in the Prairie provinces compared to Canada between 1980 and 2022. Data was collected from publicly available sexually transmitted infection reports in Canada, Alberta, Saskatchewan, and Manitoba. We extracted the number and rates of gonorrhea cases; percentage of cases by sex, age, ethnicity, sexual orientation; and data on cases diagnosed by culture and antimicrobial resistance. Descriptive statistics and age–period–cohort effect analysis were used. Gonorrhea cases in Canada rose from 32.4 per 100 000 in 1992 to 92.3 in 2022. In 2020, 36.9% of gonorrhea cases in Canada were females, compared to 42.8% in Alberta, 55.3% in Saskatchewan and 56% in Manitoba. People aged ≥30 years represented 22.5% of cases in 1980, and 54.1% in 2022. By 2022, the proportion of Canadian cases detected by culture declined to less than 10%, and azithromycin resistance of N. gonorrhoeae isolates was 8.1%. Alberta, Manitoba, and Saskatchewan reported higher rates of gonorrhea compared to Canada, with a higher proportion of female cases in Manitoba and Saskatchewan. Rising antimicrobial resistance rates and decreased culture testing present significant concerns for gonorrhea control and surveillance.

The normalised partial sums of values of a nonnegative multiplicative function over divisors with appropriately restricted lengths of a random permutation from the symmetric group define trajectories of a stochastic process. We prove a functional limit theorem in the Skorokhod space when the permutations are drawn uniformly at random. Furthermore, we show that the paths of the limit process almost surely belong to the space of continuous functions on the unit interval and, exploiting results from number-theoretic papers, we obtain rather complex formulas for the limits of joint power moments of the process.

This review examines the legal, voluntary, and technical mechanisms that govern the ownership of nonpersonal agricultural data generated by IoT-enabled farm machinery, sensors, and related systems. Given that this data is not subject to personal data protection legislation such as General Data Protection Regulation (GDPR), its governance presents distinct challenges requiring alternative governance approaches. Drawing on 63 peer-reviewed studies published over the last decade, this review proposes an integrated conceptual framework comprising legal enforcement, voluntary governance, and technical enforcement mechanisms. A distinctive contribution of the study is to show that data ownership in agriculture becomes meaningful at the moment of data sharing, where rights claims are made visible, contested, or constrained, and that these three governance pathways must be understood jointly rather than in isolation. The analysis demonstrates that although farmers generate vast quantities of nonpersonal data, no existing legal framework explicitly grants them ownership, leaving ownership to be ambiguously allocated or de facto transferred through contracts in ways that limit their ability to contest access or downstream use. Technical mechanisms promise automated enforcement and accountability but risk codifying existing power asymmetries when the encoded rules reflect opaque or exclusionary terms. We argue for a shift from “ownership” to “data sovereignty” understood as the sustained capacity to define, monitor, and revoke conditions of data use. Achieving this requires three interlinked pillars: enforceable baseline access and use rights for farmers, accessible and preferably open-source technical infrastructure, and participatory governance arrangements.

We study quasi-stationary distributions and quasi-limiting behaviour of Markov chains in general reducible state spaces with absorption. First, we consider state spaces that can be decomposed into two successive subsets (with communication possible in a single direction), differentiating between three situations, and characterize the exponential order of magnitude and the exact polynomial correction, called the polynomial convergence parameter, for the leading-order term of the semigroup for large time. Second, we consider general Markov chains with finitely or countably many communication classes by applying the first results iteratively over the communication classes of the chain. We conclude with an application of these results to the case of denumerable state spaces, where we prove existence for a quasi-stationary distribution without assuming irreducibility before absorption, but only aperiodicity, existence of a Lyapunov function, and existence of a point with almost surely finite return time.

Governments across the world are leveraging artificial intelligence (AI) to render services to citizens. Emerging economies are not left behind in this transformation but remain a gaping distance behind in their integration into public-sector service delivery compared to the private sector. To ensure the effective integration of AI services by government agencies to serve citizens, it is necessary to understand the constellation of factors driving user adoption of AI. Therefore, this study answers the question: how can government-initiated AI services be successfully accepted by citizens? Leveraging non-probability sampling, a snowball sample of 245 tertiary student-workers from across Ghana was surveyed to solicit their knowledge, attitudes, readiness, and use intentions towards AI-enabled government services. The data were analysed using FsQCA and complemented by PLS-SEM to confirm the findings. The findings reveal four unique configurations, summarised into two broad groups; AI enthusiasts and AI sceptics that drive AI adoption in government services. Positive readiness factors, such as knowledge of AI and optimism towards AI, characterise AI enthusiasts. In contrast, those described as AI sceptics still adopt government AI services despite their reservations and general distrust. AI sceptics are a delicate group that sit at the boundary between adoption and rejection, requiring special attention and strategies to orient them towards adoption. The study recommends effective education and trust-building strategies to foster AI adoption in government services. The findings are essential for driving the efficient implementation of AI-enabled services among working-class citizens in emerging economies.

We estimated the vaccine effectiveness (VE) of second monovalent and bivalent booster vaccines containing Omicron BA.1 or BA.4/BA.5 and the protection conferred by natural immunity against SARS-CoV-2 infection in Luxembourg. We conducted a test-negative case–control study among residents aged 60 years or older by integrating national socio-demographic, COVID-19 vaccination, and testing data, achieving full population coverage. Using conditional logistic regression, we estimated absolute and relative VE of monovalent and bivalent boosters and natural immunity from prior infection. Our analysis included 5,390 test-positive cases and 11,048 test-negative controls matched by week of testing between September 2022 and April 2023. Absolute VE for monovalent and bivalent boosters decreased from 64.8% and 66.6% in the first month to 1.5% and 16.5% after 5–6 months, respectively. The bivalent was superior to the monovalent booster only in individuals without natural immunity (relative VE 25.7%, 95% confidence interval 11.4%; 37.7%). Natural immunity lasted longer than vaccine-induced immunity with 80.7% protected at 4–8 months and 44.9% at 15–25 months post-infection. Both second booster vaccines provided temporary protection against SARS-CoV-2 infection; bivalent boosters offered a slight benefit over monovalent boosters. Natural immunity appears to confer longer-lasting protection.

There is a positive association between bacteraemia with Streptococcus bovis–Streptococcus equinus complex (SBSEC) and colorectal cancer (CRC). However, the relationship between the timing of SBSEC bacteraemia and CRC is not well-established. Associations with other gastrointestinal cancers have also been suggested. Using national registries, we retrospectively examined the incidence of CRC and other gastrointestinal cancers after SBSEC-bacteraemia in Sweden 2010–2019, and analysed the timing, characteristics, and prognosis of diagnosed CRC. Individuals with SBSEC-bacteraemia were matched to randomly selected controls from the general population at a 1:10 ratio. Cox-regression determined CRC hazard ratios (HR). In total, 908 individuals with SBSEC-bacteraemia were identified and 9,080 controls, of whom 75/908 (8.3%) and 168/9080 (1.9%) respectively had previously diagnosed CRC (p < 0.01). During follow-up of individuals without previous CRC, CRC was diagnosed in 45/833 (5.4%) individuals with SBSEC and 114/8912 (1.3%) controls (p < 0.01). The HR of CRC diagnosis for SBSEC was 10.3 (95% CI 6.7–15.8) overall and 19.8 (95% CI 11.1–35.3) during the first year of follow-up. In conclusion, there was an increased incidence of CRC, and most were diagnosed within the first year. Neither the tumour location, −stage, or -grade of diagnosed CRC nor the rates of other gastrointestinal cancers differed significantly.

Our study assessed the link between gastrointestinal (GI) infections in England and the Eat Out to Help Out scheme (EOHO), a government subsidy created to encourage people to eat out during COVID-19 pandemic (03–30 August 2020). We studied national laboratory data between January 2015 and December 2020. We used time series change point analysis to see if there were shifts in reported cases of specific GI infections (Campylobacter spp., Escherichia coli O157, and non-typhoidal Salmonella spp.) associated with the timing of the scheme. Our analysis uniquely applied the Pruned Exact Linear Time method, with generalized linear models to a national dataset of GI infections. This revealed increases in cases closely aligned to the timing of the easing of COVID-19 restrictions, prior to the introduction of the EOHO scheme. Our study showed the scheme had no measurable impact, as there was no significant change on reported cases. Substantial reductions in cases after the first lockdown, followed by an increase as restrictions were phased out, show the wider impact of COVID-19 control measures, for example, public information campaigns aimed at improving hand-hygiene. These findings highlight the complicated interactions between COVID-19 control measures, the public’s behaviour, and the spread of GI infections.

Narratives shape public perceptions and policymaking around emerging technologies like quantum technologies (QTs), yet what narratives develop across different societal domains remains underexplored. This study analyzes narratives about QTs in 36 government documents, 163 business reports, and 2023 media articles published over the past 23 years, using a mixed-methods approach that combines topic modeling with qualitative thematic analysis. We find that the dystopian or utopian extremes associated with technologies such as artificial intelligence are largely absent from discourse about QTs. Media coverage tends to cover a broad range of topics, while business and government narratives emphasize technical milestones, economic competitiveness, and national security, frequently at the expense of questions about ethics, equity, and accessibility. We discuss the implications of this focus, particularly the risk that an emphasis on zero-sum geopolitical competition could foster a more closed and fragmented innovation ecosystem.