Epidemics最新文献

UVA-EpiHiper is a national scale agent-based model to support the US COVID-19 Scenario Modeling Hub (SMH). UVA-EpiHiper uses a detailed representation of the underlying social contact network along with data measured during the course of the pandemic to initialize and calibrate the model. In this paper, we study the role of heterogeneity on model complexity and resulting epidemic dynamics using UVA-EpiHiper. We discuss various sources of heterogeneity that we encounter in the use of UVA-EpiHiper to support modeling and analysis of epidemic dynamics under various scenarios. We also discuss how this affects model complexity and computational complexity of the corresponding simulations. Using round 13 of the SMH as an example, we discuss how UVA-EpiHiper was initialized and calibrated. We then discuss how the detailed output produced by UVA-EpiHiper can be analyzed to obtain interesting insights. We find that despite the complexity in the model, the software, and the computation incurred to an agent-based model in scenario modeling, it is capable of capturing various heterogeneities of real-world systems, especially those in networks and behaviors, and enables analyzing heterogeneities in epidemiological outcomes between different demographic, geographic, and social cohorts. In applying UVA-EpiHiper to round 13 scenario modeling, we find that disease outcomes are different between and within states, and between demographic groups, which can be attributed to heterogeneities in population demographics, network structures, and initial immunity.

While the benefits of early antiretroviral therapy (ART) initiation in perinatally infected infants are well documented, early initiation is not always possible in postnatal pediatric HIV infections. The timing of ART initiation is likely to affect the size of the latent viral reservoir established, as well as the development of adaptive immune responses, such as the generation of neutralizing antibody responses against the virus. How these parameters impact the ability of infants to control viremia and the time to viral rebound after ART interruption is unclear and has never been modeled in infants. To investigate this question we used an infant nonhuman primate Simian/Human Immunodeficiency Virus (SHIV) infection model. Infant Rhesus macaques (RMs) were orally challenged with SHIV.C.CH505 375H dCT and either given ART at 4-7 days post-infection (early ART condition), at 2 weeks post-infection (intermediate ART condition), or at 8 weeks post-infection (late ART condition). These infants were then monitored for up to 60 months post-infection with serial viral load and immune measurements. To gain insight into early after analytic treatment interruption (ATI), we constructed mathematical models to investigate the effect of time of ART initiation in delaying viral rebound when treatment is interrupted, focusing on the relative contributions of latent reservoir size and autologous virus neutralizing antibody responses. We developed a stochastic mathematical model to investigate the joint effect of latent reservoir size, the autologous neutralizing antibody potency, and CD4+ T cell levels on the time to viral rebound for RMs rebounding up to 60 days post-ATI. We find that the latent reservoir size is an important determinant in explaining time to viral rebound in infant macaques by affecting the growth rate of the virus. The presence of neutralizing antibodies can also delay rebound, but we find this effect for high potency antibody responses only. Finally, we discuss the therapeutic implications of our findings.

Influenza A has two hemagglutinin groups, with stronger cross-immunity to reinfection within than between groups. Here, we explore the implications of this heterogeneity for proposed cross-protective influenza vaccines that may offer broad, but not universal, protection. While the development goal for the breadth of human influenza A vaccine is to provide cross-group protection, vaccines in current development stages may provide better protection against target groups than non-target groups. To evaluate vaccine formulation and strategies, we propose a novel perspective: a vaccine population-level target product profile (PTPP). Under this perspective, we use dynamical models to quantify the epidemiological impacts of future influenza A vaccines as a function of their properties. Our results show that the interplay of natural and vaccine-induced immunity could strongly affect seasonal subtype dynamics. A broadly protective bivalent vaccine could lower the incidence of both groups and achieve elimination with sufficient vaccination coverage. However, a univalent vaccine at low vaccination rates could permit a resurgence of the non-target group when the vaccine provides weaker immunity than natural infection. Moreover, as a proxy for pandemic simulation, we analyze the invasion of a variant that evades natural immunity. We find that a future vaccine providing sufficiently broad and long-lived cross-group protection at a sufficiently high vaccination rate, could prevent pandemic emergence and lower the pandemic burden. This study highlights that as well as effectiveness, breadth and duration should be considered in epidemiologically informed TPPs for future human influenza A vaccines.

Mathematical modeling of epidemic dynamics is crucial to understand its underlying mechanisms, quantify important parameters, and make predictions that facilitate more informed decision-making. There are three major types of models: mechanistic models including the SEIR-type paradigm, alternative data-driven (DD) approaches, and hybrid models that combine mechanistic models with DD approaches. In this paper, we summarize our work in the COVID-19 Scenario Modeling Hub (SMH) for more than 12 rounds since early 2021 for informed decision support. We emphasize the importance of deep learning techniques for epidemic modeling via a flexible DD framework that substantially complements the mechanistic paradigm to evaluate various future epidemic scenarios. We start with a traditional curve-fitting approach to model cumulative COVID-19 based on the underlying SEIR-type mechanisms. Hospitalizations and deaths are modeled as binomial processes of cases and hospitalization, respectively. We further formulate two types of deep learning models based on multivariate long short term memory (LSTM) to address the challenges of more traditional DD models. The first LSTM is structurally similar to the curve fitting approach and assumes that hospitalizations and deaths are binomial processes of cases. Instead of using a predefined exponential curve, LSTM relies on the underlying data to identify the most appropriate functions, and is capable of capturing both long-term and short-term epidemic behaviors. We then relax the assumption of dependent inputs among cases, hospitalizations, and death. Another type of LSTM that handles all input time series as parallel signals, the independent multivariate LSTM, is developed. Independent multivariate LSTM can incorporate a wide range of data sources beyond traditional case-based epidemiological surveillance. The DD framework unleashes its potential in big data era with previously neglected heterogeneous surveillance data sources, such as syndromic, environment, genomic, serologic, infoveillance, and mobility data. DD approaches, especially LSTM, complement and integrate with the mechanistic modeling paradigm, provide a feasible alternative approach to model today’s complex socio-epidemiological systems, and further leverage our ability to explore different scenarios for more informed decision-making during health emergencies.

Across many fields, scenario modeling has become an important tool for exploring long-term projections and how they might depend on potential interventions and critical uncertainties, with relevance to both decision makers and scientists. In the past decade, and especially during the COVID-19 pandemic, the field of epidemiology has seen substantial growth in the use of scenario projections. Multiple scenarios are often projected at the same time, allowing important comparisons that can guide the choice of intervention, the prioritization of research topics, or public communication. The design of the scenarios is central to their ability to inform important questions. In this paper, we draw on the fields of decision analysis and statistical design of experiments to propose a framework for scenario design in epidemiology, with relevance also to other fields. We identify six different fundamental purposes for scenario designs (decision making, sensitivity analysis, situational awareness, horizon scanning, forecasting, and value of information) and discuss how those purposes guide the structure of scenarios. We discuss other aspects of the content and process of scenario design, broadly for all settings and specifically for multi-model ensemble projections. As an illustrative case study, we examine the first 17 rounds of scenarios from the U.S. COVID-19 Scenario Modeling Hub, then reflect on future advancements that could improve the design of scenarios in epidemiological settings.

To mitigate the spread of SARS-CoV-2, the Swiss government enacted restrictions on social contacts from 2020 to 2022. In addition, individuals changed their social contact behavior to limit the risk of COVID-19. In this study, we aimed to investigate the changes in social contact patterns of the Swiss population. As part of the CoMix study, we conducted a survey consisting of 24 survey waves from January 2021 to May 2022. We collected data on social contacts and constructed contact matrices for the age groups 0–4, 5–14, 15–29, 30–64, and 65 years and older. We estimated the change in contact numbers during the COVID-19 pandemic to a synthetic pre-pandemic contact matrix. We also investigated the association of the largest eigenvalue of the social contact and transmission matrices with the stringency of pandemic measures, the effective reproduction number (R_e), and vaccination uptake. During the pandemic period, 7084 responders reported an average number of 4.5 contacts (95% confidence interval, CI: 4.5–4.6) per day overall, which varied by age and survey wave. Children aged 5–14 years had the highest number of contacts with 8.5 (95% CI: 8.1–8.9) contacts on average per day and participants that were 65 years and older reported the fewest (3.4, 95% CI: 3.2–3.5) per day. Compared with the pre-pandemic baseline, we found that the 15–29 and 30–64 year olds had the largest reduction in contacts. We did not find statistically significant associations between the largest eigenvalue of the social contact and transmission matrices and the stringency of measures, R_e, or vaccination uptake. The number of social contacts in Switzerland fell during the COVID-19 pandemic and remained below pre-pandemic levels after contact restrictions were lifted. The collected social contact data will be critical in informing modeling studies on the transmission of respiratory infections in Switzerland and to guide pandemic preparedness efforts.

The onset of the COVID-19 pandemic drove a widespread, often uncoordinated effort by research groups to develop mathematical models of SARS-CoV-2 to study its spread and inform control efforts. The urgent demand for insight at the outset of the pandemic meant early models were typically either simple or repurposed from existing research agendas. Our group predominantly uses agent-based models (ABMs) to study fine-scale intervention scenarios. These high-resolution models are large, complex, require extensive empirical data, and are often more detailed than strictly necessary for answering qualitative questions like “Should we lockdown?” During the early stages of an extraordinary infectious disease crisis, particularly before clear empirical evidence is available, simpler models are more appropriate. As more detailed empirical evidence becomes available, however, and policy decisions become more nuanced and complex, fine-scale approaches like ours become more useful. In this manuscript, we discuss how our group navigated this transition as we modeled the pandemic. The role of modelers often included nearly real-time analysis, and the massive undertaking of adapting our tools quickly. We were often playing catch up with a firehose of evidence, while simultaneously struggling to do both academic research and real-time decision support, under conditions conducive to neither. By reflecting on our experiences of responding to the pandemic and what we learned from these challenges, we can better prepare for future demands.

As we emerge from what may be the largest global public health crises of our lives, our community of epidemic modellers is naturally reflecting. What role can modelling play in supporting decision making during epidemics? How could we more effectively interact with policy makers? How should we design future disease surveillance systems? All crucial questions. But who is going to be addressing them in 10 years’ time? With high burnout and poor attrition rates in academia, both magnified in our field by our unprecedented efforts during the pandemic, and with low wages coinciding with inflation at its highest for decades, how do we retain talent? This is a multifaceted challenge, that I argue is underpinned by privilege. In this perspective, I introduce the notion of privilege and highlight how various aspects of privilege (namely gender, ethnicity, sexual orientation, language and caring responsibilities) may affect the ability of individuals to access to and progress within academic modelling careers. I propose actions that members of the epidemic modelling research community may take to mitigate these issues and ensure we have a more diverse and equitable workforce going forward.