Heterogeneity, defined as variation in individual-specific traits such as susceptibility or connectivity that are not subject to short-term change, can strongly influence epidemiological dynamics. In particular, it can reduce the effective reproduction number and the herd immunity threshold, thereby increasing the potential to control infectious disease outbreaks. However, heterogeneities at the epidemic level are extremely challenging to observe directly and difficult to estimate indirectly from epidemiological data. This study develops a methodology to estimate heterogeneities in COVID-19 transmission from under-reported incidence data. We estimate the parameters of a modified Susceptible Infected Recovered model using Bayesian Markov Chain Monte Carlo techniques, first testing the proposed method on simulated data to determine its effectiveness before applying it to incompletely reported time-series data from ten Nigerian states. We estimate the average basic reproduction number to be 1.360 (95% credible intervals (CrI): 1.356-1.366), the heterogeneity parameter representing the coefficient of variation in susceptibility to be 2.561 (95% CrI 2.247-2.884), and the reporting probability to be 0.005 (95% CrI 0.004-0.006). Information criteria indicated that the model with heterogeneity was better supported than a homogenous model. The herd immunity threshold with this estimated level of heterogeneity for the best model drops from 0.26 to 0.04 compared to the homogenous model and the percentage of individuals infected in the first wave from over 20% to about 8%. Underestimating the impact of heterogeneities would lead to overestimating the extent of interventions required to bring the effective reproduction number below 1, important when considering the design of control programs that combine different, partially effective interventions.
BACKGROUND:Gains in malaria transmission reduction achieved through insecticide-treated nets (ITNs) are reversing amid emerging artemisinin partial resistance. While sensitive diagnostics are key to monitoring resistant parasite-driven outbreaks, most malaria outbreak models treat diagnostic sensitivity as a fixed background parameter, obscuring its role both in clinical progression and disease burden. METHODS:We developed an eight-compartment deterministic model in which diagnostic sensitivity acts as a bifurcation parameter partitioning individuals infected with uncomplicated malaria into treated and false-negative pathways, incorporating severity-stratified gametocyte production and asymmetric disease-progression rates. The model was validated through local stability and sensitivity analysis. Seven intervention strategies spanning vector control, diagnostics, and treatment allocation were compared via simulations using two complementary outcomes: time for the parasite reservoir to reach ≤ 10% of baseline, and cumulative severe case-days. RESULTS:Raising sensitivity from 0.95 to 0.98 reduced false-negative cases by 60% and mortality by 14%. The top-ranked strategy (90% diagnostic allocation) reached the modeled reservoir threshold 29% faster and 46% cheaper than current ITN-focused practice. Programmatic experiences from Rwanda and Cabo Verde, which have approached or achieved elimination under diagnostic-prioritized frameworks were qualitatively, though not formally, consistent with model predictions. CONCLUSION:Diagnostic sensitivity is a high-leverage, currently underused control point that should precede and inform targeted vector-control deployment and rational antimalarial treatment allocation, offering a lower-cost path toward WHO 2030 elimination targets. Formal model validation against country-level surveillance data and explicit simulation of WHO-recommended antimalarial drug-diversification approaches such as adaptive rotation are identified as priorities for future work.
BACKGROUND:Socioeconomic disparities in COVID-19 outcomes have been widely documented, but evidence regarding inequities in SARS-CoV-2 infection risk remains mixed. In Canada, infection-induced seroprevalence appeared to converge across socioeconomic strata by late 2022, raising questions about whether inequities in infection risk diminished during the Omicron period. AIM:To assess whether apparent convergence in SARS-CoV-2 seroprevalence reflects true equity in infection risk or masks persistent socioeconomic disparities in force of infection. METHODS:We analysed serial cross-sectional SARS-CoV-2 seroprevalence data (anti-nucleocapsid antibodies) from Canadian Blood Services donors collected between April 2021 and April 2023 and stratified by area-level material deprivation quintile (Q1 = least deprived; Q5 = most deprived). We fitted a dynamic seronegative-seropositive model with sero-reversion to the full seroprevalence time series, estimating quintile-specific forces of infection before and after the emergence of the Omicron variant (January 2022). Models allowing differential Omicron-related amplification by material deprivation were compared using likelihood-based criteria. RESULTS:During the pre-Omicron period, the most deprived quintile (Q5) experienced a 71% higher force of infection than the least deprived (Q1; incidence rate ratio (IRR): 1.71; 95% CI: 1.60-1.83). Following Omicron emergence, force of infection rose markedly in all quintiles. Because the pre-Omicron baseline was lowest in the least-deprived quintile, the relative increase during Omicron was largest there (48.5-fold) and smallest in the most-deprived quintile (31.8-fold), compressing the relative socioeconomic gradient (Q5 vs Q1 IRR: 1.12; 95% CI: 1.11-1.14). Despite this compression in relative terms, the most deprived populations continued to experience higher absolute force of infection throughout the Omicron period. CONCLUSION:Convergence in SARS-CoV-2 seroprevalence across socioeconomic strata masked persistent inequalities in force of infection. Dynamic modelling demonstrates that apparent equity is attributable, in our modelling framework, to differential amplification of force of infection during the Omicron period rather than from elimination of underlying socioeconomic disparities.
Epidemiological parameters characterise the natural history, transmission and severity of a pathogen and are necessary to understand the spread of infectious diseases. These parameters underpin our ability to quantify and respond to disease outbreaks. Parameters can be estimated from observations using a range of methods and are often reported in varied ways throughout the literature. These parameter estimates constitute essential inputs to infectious disease models used to quantify and project disease spread and burden, and assess intervention impact. Hence, any incompleteness or ambiguity in reported parameter estimates can have downstream consequences on the inferences drawn from the models that use these estimates. We summarise common issues with incomplete or ambiguous reporting of epidemiological parameter estimates and illustrate the impact through five case studies. Specifically, we show that in many instances, misinterpreting parameters reported in the literature can lead to biased conclusions that mislead subsequent public health responses. Additionally, we provide recommendations on how to clearly communicate common epidemiological parameter estimates consistently and reproducibly, to maximise their secondary use and comparison, in turn minimising erroneous extraction from the literature and application in epidemiological analysis.
Social contact patterns are key drivers of infectious disease transmission. During the COVID-19 pandemic, differences between pre-COVID and COVID-era contact rates were widely attributed to non-pharmaceutical interventions such as lockdowns. However, the factors that drive changes in the distribution of contacts between different subpopulations remain poorly understood. Here, we present a clustering analysis of 45 contact matrices generated from surveys conducted before and during the COVID-19 pandemic, and analyse key structural features that distinguish contact matrices generated from POLYMOD and CoMix, two of the largest contact studies. Our analysis suggests that, while contextual features such as lockdowns could account for some of these distinguishing features, others can be explained by differences in the design of the two studies and long-term demographic trends. Our results caution against using survey data from different studies in counterfactual analysis of epidemic mitigation strategies. Doing so risks attributing differences stemming from survey design choices or long-term changes to the short-term effects of interventions.
Real-time near-term forecasts of seasonal respiratory infections are currently available from the European Centre for Disease Prevention and Control through the Respicast forecast hub. In this study, we explored the generation of probabilistic forecasts of short-term change in influenza-like illness (ILI) and acute respiratory illness (ARI) disease activity for over twenty countries in Europe.Building on publicly available datasets of respiratory disease activity in Europe, we describe a forecasting pipeline that includes: a) a flexible method to define categories of interest, in relation to disease activity seen in past seasons; and b) an approach based on smoothing splines to translate real-time probabilistic forecasts of continuous targets (available in Respicast) to forecasts of categorical targets.We applied the methods to forecast ILI/ARI consultation rates in Europe, and influenza-associated hospitalization in the United States during the 2024/25 season. For forecasts in the US, results suggest that the simple approximation method can match the skill of at least a third of the models with which a direct comparison was possible. Skill of the approximated forecasts of nearly half of the models in Respicast were better than the benchmark’s when predicting ILI rate changes. Approximation of the hub-ensemble’s continuous forecasts yielded an improvement of 24% over benchmark; further improvement through combining individual model forecasts was noted. ARI rate changes were harder to predict than ILI, and the value of model combination was less clear. Besides forecast skill, heterogeneity in models’ calibration was also observed.We believe these categorical forecasts can provide a plausibly more intuitive indicator of predicted change in respiratory disease burden, and a useful addition to routine surveillance reports. The approaches are flexible, computationally inexpensive and can help realize an addon value from extant Respicast infrastructure.
INTRODUCTION:Diarrheal disease causes 1·7 billion cases and 340,000 childhood deaths annually, with rotavirus still bearing the highest death burden despite vaccine availability. A combination vaccine targeting multiple enteropathogens could expand protection while reducing injections, visits, and costs. However, optimizing administration schedules is complex due to differences in pathogen transmissibility, immunity, and age-specific burden. METHODS:We developed an age-structured, compartmental model with pathogen-specific parameters from eight LMIC sites. Two-dose viral (rotavirus, norovirus, adenovirus 40/41) and bacterial (Shigella, ETEC) vaccines were modeled across five age schedules and variable vaccine coverage. We examined two scenarios where (a) VE was the same at all ages of administration (age-uniform) and (b) VE was lower in early childhood (age-attenuated). Outcomes included annual cases and deaths in children under five. RESULTS:With age-uniform VE, the earliest schedules produced the greatest reductions in burden, with birth/6 weeks yielding the largest mean reduction in deaths (56%, 95% UI: 41-74%) and 6/10 weeks producing similar reductions (55%, 95% UI: 39-74%). Case reductions were also comparable between schedules, with birth/6 weeks reducing cases by 34% (95% UI: 15-69%) and 6/10 weeks by 35% (95% UI: 15-70%). With age-attenuated VE, 6/10 weeks was optimal, reducing cases by 21% (95% UI: 7-49%) and deaths by 42% (95% UI: 29-58%). CONCLUSION:All schedules reduced diarrheal disease burden. Under age-uniform VE, birth/6 weeks conferred the greatest reduction in deaths, although impacts were similar to 6/10 weeks. Under age-attenuated VE, the 6/10 weeks schedule performed best.
Zoonotic cutaneous leishmaniasis (ZCL) caused by Leishmania major remains a significant public health concern in Algeria. The disease is maintained through a zoonotic cycle involving wild rodents as reservoirs and sand flies as vectors, whose population dynamics are influenced by climatic conditions. In the context of global climate change, this study investigated the relationship between climatic factors and ZCL incidence, and mapped high-risk areas in the Algerian central steppe. Epidemiological and climatic data from Djelfa, Laghouat, and Tiaret Wilayas collected between 2010 and 2022 were analyzed using generalized additive models (GAMs) to assess spatiotemporal patterns and climatic associations. A total of 8488 ZCL cases were reported over the study period, with an incidence rate of 3.10/10,000 inhabitants. The highest burden was observed in Laghouat, with a declining gradient northward, although incidence in northern areas increased over time. Seasonal peaks of reported cases occurred between November and February, primarily affecting adults aged between 45-65 and children under 10 years old. Incidence was slightly higher in men, though sex and age differences were not significant. Modeling results revealed that Palmer Drought Severity Index at a two-month lag showed a strong nonlinear association with ZCL incidence, with higher risk under extreme drought and wet conditions, and reduced transmission under moderate drought. Precipitation exhibited a marginal, inverse nonlinear effect, with higher rainfall associated with lower case counts. These findings underscore the role of climate variability in ZCL dynamics and highlight the need for climate-informed early warning systems and targeted public health interventions.
Accurate modelling of epidemic dynamics often requires accounting for how individuals modify their behaviour in response to perceived infection risk. While existing behavioural change epidemic models attempt to capture this feedback, they usually make simple, ad hoc assumptions about how memory affects responses. For example, they mostly focus on recent cases only, thereby overlooking how earlier experiences continue to shape current perceptions of risk. This study introduces a unified framework of Memory Mechanism Enhanced Behavioural Change (MEBC) models within a Bayesian SIR epidemic modelling setting. Five alternative memory mechanisms are examined - memoryless, sliding window, power-law, exponential, and reciprocal - each characterizing a different way in which past epidemic information influences current behaviour. A fully Bayesian data-augmented MCMC scheme is used to jointly estimate transmission and behavioural parameters, while accounting for uncertainty in infectious periods. Simulation experiments demonstrate that the MEBC models recover parameters accurately and remain robust under misspecified memory structures. Applications to early-stage COVID-19 outbreak in Miami-Dade County and to the 2023-2024 influenza season in Manitoba show that incorporating an easy-to-interpret memory mechanism substantially improves model fit, highlighting the critical role of collective memory in shaping behavioural adaptation and transmission dynamics.
Individual-based models (IBMs) provide a mechanistic framework in which population-level outcomes emerge from interactions between individuals. We conducted a systematic review on IBMs for respiratory pathogens published in 2020-2024. We identified 855 eligible studies. Publications peaked in 2021, with a geographical distribution positively correlated with national GDP, leaving regions understudied. Most studies focused on SARS-CoV-2 and assessed public health interventions. Research priorities evolved over time, shifting from social distancing to vaccination. Age was included in 72.4% of studies; other sociodemographic factors (e.g., race/ethnicity) were rarely considered. This review maps the IBM landscape, offering a framework to guide future modeling efforts.
Ebola virus disease (EVD) remains a constant international public health threat. Developing models that integrate the complex transmission dynamics of EVD is essential for informing evidence-based strategies for outbreak preparedness and response. Here, we introduce a stochastic, meta-population, compartmental model of EVD epidemics which accounts for key stages of the disease transmission including ecologically-driven zoonotic introductions, person-to-person transmission, spatial spread, and potentially complex interventions. Our model can distinguish between different transmission modes (direct transmission from contact with infectious cases, funeral exposures, or sexual transmission from contact with convalescent individuals) as well as different intervention mechanisms (overall reduction of contacts, safe and dignified burials, and vaccination). We illustrate our approach by simulating EVD epidemics in an area at high risk of zoonotic introduction in the Democratic Republic of the Congo, and show how it can be used to identify potential future transmission hotspots and help assess the scaling of future responses. Our model is implemented in a computer-efficient, free, open-source software, and can be used for informing public health policies.
Zoos may serve as sentinel sites for zoonotic vector-borne diseases. West Nile virus (WNV) and Usutu virus (USUV) are closely related orthoflaviviruses transmitted between Culex mosquitoes and a bird reservoir. Both viruses can also infect mammals, including humans, where they may cause symptoms and, more rarely, hospitalization and death. However, serological cross-reactivity between WNV and USUV complicates their differential diagnosis. Here, we aimed to reconstruct the dynamics of emergence of WNV in a zoo located in a newly affected area in Europe, using ELISA and Virus Neutralization Test (VNT) serological analysis of 1707 animal sera collected between 2015 and 2024. Combining this data in a model accounting for cross-reactivity with USUV, we estimated yearly forces of infection (FOI) by both viruses, and thus found that WNV likely circulated in the area one year prior to the first cases reported to the passive surveillance system. Our results also showed that, in the zoo, mammals and reptiles had a lower risk of infection than birds (relative risk of 0.14 [0.05; 0.28]), and that the exposure of birds to water (aquatic lifestyle or proximity to stagnant water) affected the risk. Finally, we estimated diagnosis parameters, including the sensitivity of the VNT (80.4% [76.5%; 84.3%]), the expected VNT titer value, and the level of serological cross-reactivity between viruses during the VNT. To conclude, our modelling framework allowed to disentangle the co-circulation of two closely related viruses, a crucial point in ensuring the reliable sentinel surveillance of these vector-borne zoonotic pathogens.
The SARS-CoV-2 pandemic in Singapore revealed pronounced age-specific differences in disease transmission and severity. To investigate these dynamics, we developed a semi-mechanistic Bayesian hierarchical model that jointly estimated SARS-CoV-2 transmission across seven age groups from 2021 to 2023. Our model integrates both case and hospitalization data, while accounting for vaccination rollouts and time-varying contact patterns between age groups. Using a non-parametric approach, we estimated changes in contact behaviour directly from the epidemiological data, avoiding assumptions about their functional form. The effective reproduction number (Rt,1) for children aged 0–14 fluctuated around one before October 2021 followed by a fluctuation with a mean smaller than one thereafter. Rt,a values for 2≤a≤6 fluctuated around one while Rt,7 fluctuated around one before September 2021 followed by a fluctuation with a mean of 1.5 thereafter. The model closely matched empirical trends, with posterior estimates and 95% credible intervals aligning well with observed case and hospitalization data across all age groups. These findings underscore the critical role of age-specific contact behaviour and highlight the importance of targeted interventions during evolving pandemic phases.
Objectives The COVID-19 pandemic highlighted the need for non-pharmaceutical interventions (NPI) to mitigate hospitalization and death prior to vaccine availability. However, the precise impact of those mitigations have remained controversial. In particular, the Cochrane Review’s recent update on the effectiveness of NPI for respiratory infections was inconclusive, especially for facemasks and COVID-19, versus a number of similarly respected studies that showed efficacy and effectiveness. Here we show that the inconclusive results could have resulted from statistical analysis not suited to the non-linear mode of action of NPI. Methods The Cochrane Review uses robust linear regression models to combine randomized controlled trial (RCT) data in their meta-analyses. Even basic models, however, show that NPI impact transmission in a non-linear way. We use a simplified, mechanistic, dynamic differential equation-based model to evaluate an NPI, facemasks, in a virtual RCT. The simulation is intended to illustrate the non-linear nature of the NPI impacts only and not intended to be a detailed replication of an actual RCT. Results NPI like facemasks have a variable impact on infection risk as a function of time, percentage of use, and basic reproductive number of the pathogen. Importantly, if the design, duration, and statistical power of the RCT are either resource limited or chosen poorly, the study will show no statistically significant effect, even when the NPI are benefitting the population exactly as predicted. Moreover, if the NPI are widely deployed and significantly reduce the effective reproductive number of the pathogen, a basic assumption of the RCT is violated, because the NPI are affecting the control group as well as the intervention group. Conclusions Meta-analyses need to account for the facts that NPI effects are non-linear and not constant, and could be benefitting the control group through source control, all of which can produce non-significant risk ratios in an RCT, even though the NPI are performing exactly as designed.
Non-pharmaceutical interventions (NPIs) have been important for controlling SARS-CoV-2 transmission, particularly before and during initial vaccine rollout. During the pandemic, the US Centers for Disease Control and Prevention issued isolation and masking guidance in case of COVID-19-like illness, a positive SARS-CoV-2 test, or known exposure to SARS-CoV-2. However, the impact of this guidance on mitigating transmission in office workplaces is unclear. We used a network-based mathematical model to estimate the impact of this guidance on SARS-CoV-2 transmission among office workers and their communities. The model represented social contacts in the home, office, and community. We used data from the CorporateMix study to parametrize social contacts among office workers and calibrated the model to represent the COVID-19 epidemic in Georgia, USA from January 2021 through August 2022. In the reference scenario (58% adherence to guidance among office workers and the broader population), workplace transmission accounted for a small fraction of total infections. Reducing adherence among office workers to 0% increased workplace transmissions by 27.1% and increasing adherence to 75% reduced workplace transmission by 7.0%. Increasing adherence to 75% among office workers had minimal impact on symptomatic cases and deaths; increasing it among the broader population was more effective in reducing office worker cases and deaths. In our model, moderate adherence to recommended NPIs in workplaces was effective in reducing transmission, but increasing adherence had limited benefit given workplaces that have low contact intensity and hybrid work arrangements. These results underscore the public health benefits of community-wide adoption of recommended NPIs.
As a company, Merck & Co., Inc., Rahway, NJ, USA (hereinafter “MSD”), strives to create an environment of mutual respect, inclusion, and accountability in all areas. However, an internal survey revealed that inclusion scores were notably low within the research unit called Health Economic and Decision Sciences (HEDS) Vaccines1. Acknowledging the moral imperative, comprehensive value, and proven efficacy gained from diversity and inclusion, the HEDS Vaccines team took immediate action to address the issue. An Inclusion Task Force was established, and the team set out to obtain accurate data to assess the current state of diversity and formulate effective, sustainable solutions to promote cultural diversity and inclusion and institutionalize these mechanisms within the department to ensure diversity in the future. The team developed sustainable recommendations to enhance diversity within the HEDS Vaccines team and created solutions to enhance inclusion within MSD.
Persistent inequities in infectious disease modelling and epidemiology hinder the effectiveness and impact of global health research. These disparities, ranging from limited data infrastructure in low- and middle-income countries (LMICs) to power imbalances in research partnerships, undermine the relevance and accessibility of research outputs for the communities most affected by infectious diseases. This perspective piece, co-authored by teams at Wellcome and the Coalition for Epidemic Preparedness Innovations (CEPI), explores how funders can help reshape research culture to foster equity, inclusivity, and sustainability. It outlines three pillars for progress: enabling thriving research environments, promoting collaborative and engaged research, and embedding equitable and open access to research outputs. Drawing on case studies such as the Global Research on Antimicrobial Resistance (GRAM) study and CEPI’s Global South Leaders in Epidemic Analytics and Response Network (GS LEARN), the authors highlight how inclusive funding strategies, community engagement, and co-funding models can drive systemic change, recognising that structural and political barriers can impede progress. Funders must go beyond grant-making to support capacity building, inclusive governance, and equitable dissemination of research. Funders are in a unique position to help cultivate an engaged and broadly representative research ecosystem. This perspective piece calls on funders to continuously learn, adapt, and collaborate to ensure that the benefits of infectious disease modelling and epidemiology research benefit affected communities to address current and future global health challenges.
SARS-CoV-2, the virus responsible for COVID-19, emerged in late 2019 and rapidly spread worldwide. Inferring transmission direction between epidemiologically linked cases is an important component of outbreak investigation, yet symptom-onset-based heuristics can become unreliable when onset differences are small due to incubation variability, reporting noise, and asymptomatic infections. Using a detailed contact-tracing dataset, we develop Pseudo-Time Reconstruction for Epidemic (PTRE), a network-informed analytical framework that integrates symptom timing, individual-level information, and contact network structure to induce a relative ordering of cases. We apply this framework to regional COVID-19 contact-tracing data and evaluate directional discrimination across onset-gap regimes. We further validate PTRE through simulation studies with known ground-truth transmission directions on diverse network topologies. Our results show a regime-dependent pattern: PTRE aligns with onset-based ordering when temporal separation is large, while structural connectivity provides a complementary directional signal when onset differences are minimal. These findings highlight both the potential and the limitations of integrating network information to enhance directional discrimination under temporal ambiguity.
