Pub Date : 2026-03-01Epub Date: 2025-12-13DOI: 10.1016/j.epidem.2025.100882
Jack Ward , Oswaldo Gressani , Sol Kim , Niel Hens , W. John Edmunds
Introduction
In the light of the COVID-19 pandemic many countries are trying to widen their pandemic planning from its traditional focus on influenza. However, it is impossible to draw up detailed plans for every pathogen with epidemic potential. We set out to try to simplify this process by reviewing the epidemiology of a range of pathogens with pandemic potential and seeing whether they fall into groups with shared epidemiological traits.
Methods
We reviewed the epidemiological characteristics of 19 different pathogens with pandemic potential (those on the WHO priority list of pathogens, different strains of influenza and Mpox). We extracted data on key parameters (reproduction number serial interval, proportion of presymptomatic transmission, case fatality risk and transmission route) and applied an unsupervised learning algorithm. This combined Monte Carlo sampling with ensemble clustering to classify pathogens into distinct epidemiological archetypes based on their shared characteristics.
Results
From 154 articles we extracted 302 epidemiological parameter estimates. The clustering algorithms categorise these pathogens into six archetypes (1) highly transmissible Coronaviruses, (2) moderately transmissible Coronaviruses, (3) high-severity contact and zoonotic pathogens, (4) Influenza viruses (5) MERS-CoV-like and (6) MPV-like.
Conclusion
Unsupervised learning on epidemiological data can be used to define distinct pathogen archetypes. This method offers a valuable framework to allocate emerging and novel pathogens into defined groups to evaluate common approaches for their control.
{"title":"The epidemiology of pathogens with pandemic potential: A review of key parameters and clustering analysis","authors":"Jack Ward , Oswaldo Gressani , Sol Kim , Niel Hens , W. John Edmunds","doi":"10.1016/j.epidem.2025.100882","DOIUrl":"10.1016/j.epidem.2025.100882","url":null,"abstract":"<div><h3>Introduction</h3><div>In the light of the COVID-19 pandemic many countries are trying to widen their pandemic planning from its traditional focus on influenza. However, it is impossible to draw up detailed plans for every pathogen with epidemic potential. We set out to try to simplify this process by reviewing the epidemiology of a range of pathogens with pandemic potential and seeing whether they fall into groups with shared epidemiological traits.</div></div><div><h3>Methods</h3><div>We reviewed the epidemiological characteristics of 19 different pathogens with pandemic potential (those on the WHO priority list of pathogens, different strains of influenza and Mpox). We extracted data on key parameters (reproduction number serial interval, proportion of presymptomatic transmission, case fatality risk and transmission route) and applied an unsupervised learning algorithm. This combined Monte Carlo sampling with ensemble clustering to classify pathogens into distinct epidemiological archetypes based on their shared characteristics.</div></div><div><h3>Results</h3><div>From 154 articles we extracted 302 epidemiological parameter estimates. The clustering algorithms categorise these pathogens into six archetypes (1) highly transmissible Coronaviruses, (2) moderately transmissible Coronaviruses, (3) high-severity contact and zoonotic pathogens, (4) Influenza viruses (5) MERS-CoV-like and (6) MPV-like.</div></div><div><h3>Conclusion</h3><div>Unsupervised learning on epidemiological data can be used to define distinct pathogen archetypes. This method offers a valuable framework to allocate emerging and novel pathogens into defined groups to evaluate common approaches for their control.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100882"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145760587","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-12-20DOI: 10.1016/j.epidem.2025.100884
Byul Nim Kim , Minchan Choi , Hyosun Lee, Sunmi Lee
Understanding the spatial and temporal dynamics of infectious disease transmission is critical for effective epidemic preparedness and response. COVID-19 transmission is influenced by mobility patterns, regional connectivity, and evolving public health interventions, making it challenging to quantify region-specific transmission risks. Our study integrates intervention-driven analysis, real-world data, and high-resolution modeling to establish a robust computational framework for assessing interregional transmission dynamics. We employ a multi-patch model to estimate the time-dependent regional effective reproduction number and systematically quantify interregional infection contributions. By integrating high-resolution mobility and COVID-19 incidence data from South Korea, we identify key transmission hubs and assess the impact of mobility-driven transmission across different epidemic phases. Our results highlight Seoul and Gyeonggi as dominant sources of interregional spread, with their influence varying across phases of the pandemic. By distinguishing locally transmitted infections from mobility-induced cases, we introduce a data-driven approach to evaluate the effectiveness of movement restrictions and targeted interventions. Findings from the Pre-Delta phase demonstrate that mobility controls in transmission hubs significantly reduced the spread of infections. Our results underscore that densely connected regions disproportionately drive nationwide transmission, emphasizing the need for adaptive, phase-dependent intervention strategies rather than uniform nationwide policies. This study advances computational epidemiology by providing a scalable framework for integrating real-world mobility data with epidemic modeling to inform targeted, data-driven public health responses.
{"title":"Evaluating mobility restrictions through spatiotemporal effective reproduction number analysis in a multi-patch model with complex mobility data","authors":"Byul Nim Kim , Minchan Choi , Hyosun Lee, Sunmi Lee","doi":"10.1016/j.epidem.2025.100884","DOIUrl":"10.1016/j.epidem.2025.100884","url":null,"abstract":"<div><div>Understanding the spatial and temporal dynamics of infectious disease transmission is critical for effective epidemic preparedness and response. COVID-19 transmission is influenced by mobility patterns, regional connectivity, and evolving public health interventions, making it challenging to quantify region-specific transmission risks. Our study integrates intervention-driven analysis, real-world data, and high-resolution modeling to establish a robust computational framework for assessing interregional transmission dynamics. We employ a multi-patch model to estimate the time-dependent regional effective reproduction number and systematically quantify interregional infection contributions. By integrating high-resolution mobility and COVID-19 incidence data from South Korea, we identify key transmission hubs and assess the impact of mobility-driven transmission across different epidemic phases. Our results highlight Seoul and Gyeonggi as dominant sources of interregional spread, with their influence varying across phases of the pandemic. By distinguishing locally transmitted infections from mobility-induced cases, we introduce a data-driven approach to evaluate the effectiveness of movement restrictions and targeted interventions. Findings from the Pre-Delta phase demonstrate that mobility controls in transmission hubs significantly reduced the spread of infections. Our results underscore that densely connected regions disproportionately drive nationwide transmission, emphasizing the need for adaptive, phase-dependent intervention strategies rather than uniform nationwide policies. This study advances computational epidemiology by providing a scalable framework for integrating real-world mobility data with epidemic modeling to inform targeted, data-driven public health responses.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100884"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145913652","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-12-15DOI: 10.1016/j.epidem.2025.100874
Peter F.M. Teunis , Jessica C. Seidman , Dipesh Tamrakar , Farah Naz Qamar , Samir K. Saha , Denise O. Garrett , Jason R. Andrews , Richelle C. Charles , Kristen Aiemjoy
Enteric fever, a systematic bacterial infection caused by Salmonella Typhi and Paratyphi, continues to impose a significant public health burden in low and middle-income countries, yet our understanding of the serum antibody dynamics following infection remains incomplete. Although previous work has characterized the longitudinal seroresponses following acute typhoid infection, gaps persist in deciphering how repeated exposures influence antibody decay and protection. In our longitudinal cohort study of blood culture-confirmed enteric fever cases enrolled in Bangladesh, Nepal, and Pakistan, we identified several instances of suspected re-infection defined by an initial decline followed by a subsequent rise in antibody levels. The presence of re-infection events interferes with the estimation of antibody decay dynamics and influences the interpretation of seroepidemiological data at the population level. To study the seroresponses to subsequent infections we employed a synthetic within-host model that accounts for elevated baseline antibody levels at time of infection. Compared to the first seroresponse, second or later responses appear to have similar decay rates. As peak levels depend on the time between infections, a new model-derived metric is proposed that does not depend on time since the most recent infection: the minimum baseline antibody level at infection resulting in a small jump (protective) seroconversion. After infection the time to reach the minimum baseline level increases about tenfold. Finally, we show how ignoring variation in subsequent seroresponses into seroincidence estimates leads to bias in population-level infection rates. These findings underscore the importance of accounting for re-infection in seroepidemiological studies and provide refined metrics for interpreting antibody responses, with critical implications for assessing disease burden and guiding public health strategies in endemic regions.
{"title":"Seroresponse to repeated infections with Salmonella enterica Typhi and Paratyphi A","authors":"Peter F.M. Teunis , Jessica C. Seidman , Dipesh Tamrakar , Farah Naz Qamar , Samir K. Saha , Denise O. Garrett , Jason R. Andrews , Richelle C. Charles , Kristen Aiemjoy","doi":"10.1016/j.epidem.2025.100874","DOIUrl":"10.1016/j.epidem.2025.100874","url":null,"abstract":"<div><div>Enteric fever, a systematic bacterial infection caused by <em>Salmonella</em> Typhi and Paratyphi, continues to impose a significant public health burden in low and middle-income countries, yet our understanding of the serum antibody dynamics following infection remains incomplete. Although previous work has characterized the longitudinal seroresponses following acute typhoid infection, gaps persist in deciphering how repeated exposures influence antibody decay and protection. In our longitudinal cohort study of blood culture-confirmed enteric fever cases enrolled in Bangladesh, Nepal, and Pakistan, we identified several instances of suspected re-infection defined by an initial decline followed by a subsequent rise in antibody levels. The presence of re-infection events interferes with the estimation of antibody decay dynamics and influences the interpretation of seroepidemiological data at the population level. To study the seroresponses to subsequent infections we employed a synthetic within-host model that accounts for elevated baseline antibody levels at time of infection. Compared to the first seroresponse, second or later responses appear to have similar decay rates. As peak levels depend on the time between infections, a new model-derived metric is proposed that does not depend on time since the most recent infection: the minimum baseline antibody level at infection resulting in a small jump (protective) seroconversion. After infection the time to reach the minimum baseline level increases about tenfold. Finally, we show how ignoring variation in subsequent seroresponses into seroincidence estimates leads to bias in population-level infection rates. These findings underscore the importance of accounting for re-infection in seroepidemiological studies and provide refined metrics for interpreting antibody responses, with critical implications for assessing disease burden and guiding public health strategies in endemic regions.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100874"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145760588","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2026-01-29DOI: 10.1016/j.epidem.2026.100885
Chad W. Milando , George G. Vega Yon , Kaitlyn Johnson , Alessandra Urbinati , Guillaume St-Onge , Brennan Klein , Anne Cori , Laura F. White , Rt Collabathon participants
The reproductive number, , is a popular metric used for monitoring infectious diseases. describes the expected number of infections that will be generated from a single infection at time , which maps nicely to the likelihood that disease incidence will increase, decrease, or remain constant in the near future. Although this metric has existed for decades, it became more widely used during the COVID-19 pandemic and there was a subsequent proliferation of new estimation methods and software tools. This rapid development of methods and tools presents many opportunities and challenges for users, researchers, and decision makers. In recognition of this growth, we convened a three-day “collabathon” in September 2024 to bring together researchers and public health practitioners to identify challenges and areas for future development in estimation and to begin work in these areas. Here we provide a high-level summary of current methods and report on the findings from the collabathon, including a summary of current challenges and recommendations for future development, evaluation and interpretation of .
{"title":"A vision for estimation of the instantaneous reproductive number","authors":"Chad W. Milando , George G. Vega Yon , Kaitlyn Johnson , Alessandra Urbinati , Guillaume St-Onge , Brennan Klein , Anne Cori , Laura F. White , Rt Collabathon participants","doi":"10.1016/j.epidem.2026.100885","DOIUrl":"10.1016/j.epidem.2026.100885","url":null,"abstract":"<div><div>The reproductive number, <span><math><msub><mrow><mi>R</mi></mrow><mrow><mi>t</mi></mrow></msub></math></span>, is a popular metric used for monitoring infectious diseases. <span><math><msub><mrow><mi>R</mi></mrow><mrow><mi>t</mi></mrow></msub></math></span> describes the expected number of infections that will be generated from a single infection at time <span><math><mi>t</mi></math></span>, which maps nicely to the likelihood that disease incidence will increase, decrease, or remain constant in the near future. Although this metric has existed for decades, it became more widely used during the COVID-19 pandemic and there was a subsequent proliferation of new estimation methods and software tools. This rapid development of methods and tools presents many opportunities and challenges for users, researchers, and decision makers. In recognition of this growth, we convened a three-day “collabathon” in September 2024 to bring together researchers and public health practitioners to identify challenges and areas for future development in <span><math><msub><mrow><mi>R</mi></mrow><mrow><mi>t</mi></mrow></msub></math></span> estimation and to begin work in these areas. Here we provide a high-level summary of current methods and report on the findings from the collabathon, including a summary of current challenges and recommendations for future development, evaluation and interpretation of <span><math><msub><mrow><mi>R</mi></mrow><mrow><mi>t</mi></mrow></msub></math></span>.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100885"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147311706","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2026-01-21DOI: 10.1016/j.epidem.2026.100890
Simon Busch-Moreno , Moritz U.G. Kraemer
Early outbreak data analysis is critical for informing about their potential impact and interventions. However, data obtained early in outbreaks are often sensitive and subject to strict privacy restrictions. Thus, federated analysis, which implies decentralised collaborative analysis where no raw data sharing is required, emerged as an attractive paradigm to solve issues around data privacy and confidentiality. In the present study, we propose two approaches which require neither data sharing nor direct communication between devices/servers. The first approach approximates the joint posterior distributions via a multivariate normal distribution and uses this information to update prior distributions sequentially. The second approach uses summaries from parameters’ posteriors obtained locally at different locations (sites) to perform a meta-analysis via a hierarchical model. We test these models on simulated and on real outbreak data to estimate the incubation period of multiple infectious diseases. Results indicate that both approaches can recover incubation period parameters accurately, but they differ in terms of structure and complexity; which makes them suitable for different types of analyses or to be used in combination. We provide a framework for federated analysis of early outbreak data where the public health contexts are complex.
{"title":"Sequential federated analysis of early outbreak data applied to incubation period estimation","authors":"Simon Busch-Moreno , Moritz U.G. Kraemer","doi":"10.1016/j.epidem.2026.100890","DOIUrl":"10.1016/j.epidem.2026.100890","url":null,"abstract":"<div><div>Early outbreak data analysis is critical for informing about their potential impact and interventions. However, data obtained early in outbreaks are often sensitive and subject to strict privacy restrictions. Thus, federated analysis, which implies decentralised collaborative analysis where no raw data sharing is required, emerged as an attractive paradigm to solve issues around data privacy and confidentiality. In the present study, we propose two approaches which require neither data sharing nor direct communication between devices/servers. The first approach approximates the joint posterior distributions via a multivariate normal distribution and uses this information to update prior distributions sequentially. The second approach uses summaries from parameters’ posteriors obtained locally at different locations (sites) to perform a meta-analysis via a hierarchical model. We test these models on simulated and on real outbreak data to estimate the incubation period of multiple infectious diseases. Results indicate that both approaches can recover incubation period parameters accurately, but they differ in terms of structure and complexity; which makes them suitable for different types of analyses or to be used in combination. We provide a framework for federated analysis of early outbreak data where the public health contexts are complex.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100890"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173555","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Reliable hospital admission can aid contingency planning during pandemics. While some studies have developed models for predicting new hospitalizations, most focus on data at the regional or national level. During health crises, a prediction model tailored to a single hospital is essential, as models based on regional data may fail to account for the heterogeneity within the local population in terms of demographics, morbidity, and other relevant characteristics. Our study addresses this gap by presenting an approach for short-term forecasting of daily pandemic admissions to a single hospital. We develop a time series model using COVID-19 admission data during the Alpha wave to one Norwegian hospital with a highly heterogeneous catchment area that includes both urban and rural areas. Previous hospitalizations are included as predictors, along with the local reproduction number (ℜ-number) to capture pandemic trends. To account for demographic differences in admission rates, we employ group-based modelling to divide the catchment area into sub-areas. Forecasts generated from sub-area models are then merged and compared with the forecasts from a model for the entire catchment area. The model’s forecasting ability is tested on the Delta wave. The merged model outperforms the total model on the Alpha wave, and both surpass the ARIMA benchmark. On the out of sample Delta wave, the total model performs better overall. While the model overpredicts admissions at the beginning of the Delta wave and the prediction intervals are somewhat conservative, it demonstrates potential for reliably forecasting new daily pandemic admissions. Continuous model adaption will however be necessary as the pandemic evolves.
{"title":"Short term forecast of new daily pandemic hospitalizations: A time series model for a single hospital","authors":"Lieke Fleur Heupink , Espen Rostrup Nakstad , Hilde Lurås , Pål Wiik , Kristine Lippestad , Fred Espen Benth , Jūratė Šaltytė Benth","doi":"10.1016/j.epidem.2026.100894","DOIUrl":"10.1016/j.epidem.2026.100894","url":null,"abstract":"<div><div>Reliable hospital admission can aid contingency planning during pandemics. While some studies have developed models for predicting new hospitalizations, most focus on data at the regional or national level. During health crises, a prediction model tailored to a single hospital is essential, as models based on regional data may fail to account for the heterogeneity within the local population in terms of demographics, morbidity, and other relevant characteristics. Our study addresses this gap by presenting an approach for short-term forecasting of daily pandemic admissions to a single hospital. We develop a time series model using COVID-19 admission data during the Alpha wave to one Norwegian hospital with a highly heterogeneous catchment area that includes both urban and rural areas. Previous hospitalizations are included as predictors, along with the local reproduction number (ℜ-number) to capture pandemic trends. To account for demographic differences in admission rates, we employ group-based modelling to divide the catchment area into sub-areas. Forecasts generated from sub-area models are then merged and compared with the forecasts from a model for the entire catchment area. The model’s forecasting ability is tested on the Delta wave. The merged model outperforms the total model on the Alpha wave, and both surpass the ARIMA benchmark. On the out of sample Delta wave, the total model performs better overall. While the model overpredicts admissions at the beginning of the Delta wave and the prediction intervals are somewhat conservative, it demonstrates potential for reliably forecasting new daily pandemic admissions. Continuous model adaption will however be necessary as the pandemic evolves.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100894"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173556","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-12-12DOI: 10.1016/j.epidem.2025.100880
Silas Koemen , Nuno R. Faria , Leonardo S. Bastos , Oliver Ratmann , André Victor Ribeiro Amaral , on behalf of the Machine Learning & Global Health Network
Nowcasting methods are crucial in infectious disease surveillance, as reporting delays often lead to underestimation of recent incidence and can impair timely public health decision-making. Accurate real-time estimates of case counts are essential for resource allocation, policy responses, and communication with the public. In this paper, we propose a novel probabilistic neural network (PNN) architecture, named NowcastPNN, to estimate occurred-but-not-yet-reported cases of infectious diseases, demonstrated here using dengue fever incidence in São Paulo, Brazil. The proposed model combines statistical modelling of the true number of cases, assuming a Negative Binomial (NB) distribution, with recent advances in machine learning and deep learning, such as the attention mechanism. Uncertainty intervals are obtained by sampling from the predicted NB distribution and using Monte Carlo (MC) Dropout. Using proper scoring rules for the prediction intervals, NowcastPNN achieves nearly a 30% reduction in losses compared to the second-best model among other state-of-the-art approaches. While our model requires a large training dataset (equivalent to two to four years of incidence counts) to outperform benchmarks, it is computationally cheap and outperforms alternative methods even with significantly fewer observations as input. These features make the NowcastPNN model a promising tool for nowcasting in epidemiological surveillance of arboviral threats and other domains involving right-truncated data.
{"title":"Fast and trustworthy nowcasting of dengue fever: A case study using attention-based probabilistic neural networks in São Paulo, Brazil","authors":"Silas Koemen , Nuno R. Faria , Leonardo S. Bastos , Oliver Ratmann , André Victor Ribeiro Amaral , on behalf of the Machine Learning & Global Health Network","doi":"10.1016/j.epidem.2025.100880","DOIUrl":"10.1016/j.epidem.2025.100880","url":null,"abstract":"<div><div>Nowcasting methods are crucial in infectious disease surveillance, as reporting delays often lead to underestimation of recent incidence and can impair timely public health decision-making. Accurate real-time estimates of case counts are essential for resource allocation, policy responses, and communication with the public. In this paper, we propose a novel probabilistic neural network (PNN) architecture, named NowcastPNN, to estimate occurred-but-not-yet-reported cases of infectious diseases, demonstrated here using dengue fever incidence in São Paulo, Brazil. The proposed model combines statistical modelling of the true number of cases, assuming a Negative Binomial (NB) distribution, with recent advances in machine learning and deep learning, such as the attention mechanism. Uncertainty intervals are obtained by sampling from the predicted NB distribution and using Monte Carlo (MC) Dropout. Using proper scoring rules for the prediction intervals, NowcastPNN achieves nearly a 30% reduction in losses compared to the second-best model among other state-of-the-art approaches. While our model requires a large training dataset (equivalent to two to four years of incidence counts) to outperform benchmarks, it is computationally cheap and outperforms alternative methods even with significantly fewer observations as input. These features make the NowcastPNN model a promising tool for nowcasting in epidemiological surveillance of arboviral threats and other domains involving right-truncated data.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100880"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145760662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2026-01-12DOI: 10.1016/j.epidem.2026.100888
Jessica E. Biddle , Stacey House , Jennie H. Kwon , Rachel M. Presti , Stephanie A. Fritz , Tara Curley , Son H. McLaren , Melissa S. Stockwell , Jonathan Schmitz , H. Keipp Talbot , Carlos G. Grijalva , Elie A. Saade , Zainab Albar , Vel Murugan , Rick A. Cruz , Emily T. Martin , Ivana A. Vaughn , Karen J. Wernli , Brianna M. Wickersham , Richard K. Zimmerman , Olivia L. Williams
Household studies play a critical role in estimating influenza transmission parameters, which are essential for real-time modeling of epidemic and pandemic dynamics to inform influenza control strategies. We compared two approaches for estimating household influenza transmission parameters from multisite studies conducted in the United States during the 2023–2024 influenza season: interviewing index cases about illnesses among household contacts (n = 1537 contacts) and prospective enrollment of index cases and their household contacts with systematic, daily symptom assessment and testing (n = 857 contacts). We compared the detection of symptomatic illness, influenza-like illness (ILI; fever and either cough or sore throat), influenza virus infection, and estimated serial illness onset intervals among household contacts across studies. Symptomatic illness episodes among household contacts were identified in 40 % of contacts by index case interview compared to 59 % of contacts from individual daily follow-up. Reports of ILI were more comparable between platforms (20 % vs. 26 % respectively). Index case interviews identified 12 % of household contacts with positive influenza tests while systematic, daily testing in the individual daily follow-up platform identified influenza infection among 44 % of household contacts. Both platforms yielded a median serial interval of 4 days. While index case interviews offer rapid, resource-efficient data collection and can inform epidemiological outcomes such as age-related risks and serial intervals, they substantially underestimate laboratory-confirmed influenza cases compared to systematic daily follow-up. These findings highlight the importance of study design in accurately capturing transmission dynamics and underscore the need for systematic laboratory testing to inform public health responses.
{"title":"Estimating influenza transmission parameters: Comparing two study designs, 2023–2024","authors":"Jessica E. Biddle , Stacey House , Jennie H. Kwon , Rachel M. Presti , Stephanie A. Fritz , Tara Curley , Son H. McLaren , Melissa S. Stockwell , Jonathan Schmitz , H. Keipp Talbot , Carlos G. Grijalva , Elie A. Saade , Zainab Albar , Vel Murugan , Rick A. Cruz , Emily T. Martin , Ivana A. Vaughn , Karen J. Wernli , Brianna M. Wickersham , Richard K. Zimmerman , Olivia L. Williams","doi":"10.1016/j.epidem.2026.100888","DOIUrl":"10.1016/j.epidem.2026.100888","url":null,"abstract":"<div><div>Household studies play a critical role in estimating influenza transmission parameters, which are essential for real-time modeling of epidemic and pandemic dynamics to inform influenza control strategies. We compared two approaches for estimating household influenza transmission parameters from multisite studies conducted in the United States during the 2023–2024 influenza season: interviewing index cases about illnesses among household contacts (n = 1537 contacts) and prospective enrollment of index cases and their household contacts with systematic, daily symptom assessment and testing (n = 857 contacts). We compared the detection of symptomatic illness, influenza-like illness (ILI; fever and either cough or sore throat), influenza virus infection, and estimated serial illness onset intervals among household contacts across studies. Symptomatic illness episodes among household contacts were identified in 40 % of contacts by index case interview compared to 59 % of contacts from individual daily follow-up. Reports of ILI were more comparable between platforms (20 % vs. 26 % respectively). Index case interviews identified 12 % of household contacts with positive influenza tests while systematic, daily testing in the individual daily follow-up platform identified influenza infection among 44 % of household contacts. Both platforms yielded a median serial interval of 4 days. While index case interviews offer rapid, resource-efficient data collection and can inform epidemiological outcomes such as age-related risks and serial intervals, they substantially underestimate laboratory-confirmed influenza cases compared to systematic daily follow-up. These findings highlight the importance of study design in accurately capturing transmission dynamics and underscore the need for systematic laboratory testing to inform public health responses.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100888"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2026-02-20DOI: 10.1016/j.epidem.2026.100899
Xuemei Yan, Nicholas C. Grassly , Margarita Pons-Salort
A notable decline in incidence of hand, foot, and mouth disease (HFMD) was observed globally since the beginning of the COVID-19 pandemic, indicating a change in the transmission dynamics of enteroviruses (EVs) causing HFMD. This study estimates the impact of COVID-19 non-pharmaceutical interventions (NPIs) on the transmission of EVs in Japan. Japan has a well-established sentinel surveillance system for enteroviruses and HFMD. A susceptible-infected-recovered stochastic transmission model was fitted to the time-series of weekly reported number of HFMD enterovirus A71 (EV-A71), coxsackievirus A16 (CVA16), and CVA6 cases from January 2005 to December 2024, to estimate the impact of government containment interventions during the COVID-19 pandemic. Compared to pre-pandemic levels, the transmission rate decreased by 7.86 % (95 % Credible Interval (CrI): 6.31 %, 10.3 %) for EV-A71, 12.0 % (10.3 %, 14.1 %) for CVA16, and 20.1 % (16.3 %, 24.0 %) for CVA6 on average from January 2020 to December 2022. The reduction in transmission rate during this period correlated with the COVID-19 Stringency Index, such that 49.3 %, 19.9 % and 25.5 % of the reduction for each serotype respectively was explained by the index, with the biggest reductions occurring during “State of Emergency” periods. Enterovirus serotypes with higher transmissibility () and therefore, a younger mean age at infection, were estimated to experience a bigger reduction in transmission that was more strongly associated with COVID-19 non-pharmaceutical interventions captured by the Stringency Index. This has implications for the impact of NPIs on other viral pathogens.
{"title":"Impact of COVID-19 on the transmission dynamics of HFMD associated enterovirus serotypes in Japan: A modelling study of surveillance data","authors":"Xuemei Yan, Nicholas C. Grassly , Margarita Pons-Salort","doi":"10.1016/j.epidem.2026.100899","DOIUrl":"10.1016/j.epidem.2026.100899","url":null,"abstract":"<div><div>A notable decline in incidence of hand, foot, and mouth disease (HFMD) was observed globally since the beginning of the COVID-19 pandemic, indicating a change in the transmission dynamics of enteroviruses (EVs) causing HFMD. This study estimates the impact of COVID-19 non-pharmaceutical interventions (NPIs) on the transmission of EVs in Japan. Japan has a well-established sentinel surveillance system for enteroviruses and HFMD. A susceptible-infected-recovered stochastic transmission model was fitted to the time-series of weekly reported number of HFMD enterovirus A71 (EV-A71), coxsackievirus A16 (CVA16), and CVA6 cases from January 2005 to December 2024, to estimate the impact of government containment interventions during the COVID-19 pandemic. Compared to pre-pandemic levels, the transmission rate decreased by 7.86 % (95 % Credible Interval (CrI): 6.31 %, 10.3 %) for EV-A71, 12.0 % (10.3 %, 14.1 %) for CVA16, and 20.1 % (16.3 %, 24.0 %) for CVA6 on average from January 2020 to December 2022. The reduction in transmission rate during this period correlated with the COVID-19 Stringency Index, such that 49.3 %, 19.9 % and 25.5 % of the reduction for each serotype respectively was explained by the index, with the biggest reductions occurring during “State of Emergency” periods. Enterovirus serotypes with higher transmissibility (<span><math><msub><mrow><mi>R</mi></mrow><mrow><mn>0</mn></mrow></msub></math></span>) and therefore, a younger mean age at infection, were estimated to experience a bigger reduction in transmission that was more strongly associated with COVID-19 non-pharmaceutical interventions captured by the Stringency Index. This has implications for the impact of NPIs on other viral pathogens.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100899"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147285874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Timely and accurate short-term forecasting of Influenza-Like Illness (ILI) is crucial for guiding outbreak response, optimizing healthcare resource allocation, and informing public health interventions. The COVID-19 pandemic, which disrupted seasonal ILI dynamics due to widespread nonpharmaceutical interventions (NPI), underscored the urgent need for adaptive and reliable forecasting frameworks.
Method:
In this study, we present a novel ensemble modeling approach that combines a mechanistic n-subepidemic model with a Monte Carlo Dropout Long Short-Term Memory (LSTM) neural network to improve age-specific ILI forecasting performance in South Korea. By capturing both the structured dynamics of disease spread and nonlinear temporal dependencies, our ensemble method adapts to pandemic-altered transmission patterns while offering robust uncertainty quantification. Age-stratified forecasting allows the framework to capture heterogeneity in vulnerability and transmission across demographic groups, providing more targeted insights for policy and planning.
Result:
We evaluated forecasting performance across four epidemic waves using the Weighted Interval Score (WIS), Mean Absolute Error (MAE), consistently finding that the ensemble models outperformed individual approaches.
Conclusion:
These findings underscore the power of hybrid forecasting approaches to improve epidemic preparedness and response, providing a flexible data-driven framework that can evolve with changing transmission dynamics and extend to other emerging infectious threats.
{"title":"Enhancing Influenza-Like Illness forecasting: An ensemble approach combining mathematical and deep learning models amidst the COVID-19 pandemic","authors":"Ganghyun Yoon , Amanda Bleichrodt , Gerardo Chowell , Sunmi Lee","doi":"10.1016/j.epidem.2026.100901","DOIUrl":"10.1016/j.epidem.2026.100901","url":null,"abstract":"<div><h3>Background:</h3><div>Timely and accurate short-term forecasting of Influenza-Like Illness (ILI) is crucial for guiding outbreak response, optimizing healthcare resource allocation, and informing public health interventions. The COVID-19 pandemic, which disrupted seasonal ILI dynamics due to widespread nonpharmaceutical interventions (NPI), underscored the urgent need for adaptive and reliable forecasting frameworks.</div></div><div><h3>Method:</h3><div>In this study, we present a novel ensemble modeling approach that combines a mechanistic n-subepidemic model with a Monte Carlo Dropout Long Short-Term Memory (LSTM) neural network to improve age-specific ILI forecasting performance in South Korea. By capturing both the structured dynamics of disease spread and nonlinear temporal dependencies, our ensemble method adapts to pandemic-altered transmission patterns while offering robust uncertainty quantification. Age-stratified forecasting allows the framework to capture heterogeneity in vulnerability and transmission across demographic groups, providing more targeted insights for policy and planning.</div></div><div><h3>Result:</h3><div>We evaluated forecasting performance across four epidemic waves using the Weighted Interval Score (WIS), Mean Absolute Error (MAE), consistently finding that the ensemble models outperformed individual approaches.</div></div><div><h3>Conclusion:</h3><div>These findings underscore the power of hybrid forecasting approaches to improve epidemic preparedness and response, providing a flexible data-driven framework that can evolve with changing transmission dynamics and extend to other emerging infectious threats.</div></div>","PeriodicalId":49206,"journal":{"name":"Epidemics","volume":"54 ","pages":"Article 100901"},"PeriodicalIF":2.4,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147348617","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号