Redefining pandemic preparedness: Multidisciplinary insights from the CERP modelling workshop in infectious diseases, workshop report.

In July 2023, the Center of Excellence in Respiratory Pathogens organized a two-day workshop on infectious diseases modelling and the lessons learnt from the Covid-19 pandemic. This report summarizes the rich discussions that occurred during the workshop. The workshop participants discussed multisource data integration and highlighted the benefits of combining traditional surveillance with more novel data sources like mobility data, social media, and wastewater monitoring. Significant advancements were noted in the development of predictive models, with examples from various countries showcasing the use of machine learning and artificial intelligence in detecting and monitoring disease trends. The role of open collaboration between various stakeholders in modelling was stressed, advocating for the continuation of such partnerships beyond the pandemic. A major gap identified was the absence of a common international framework for data sharing, which is crucial for global pandemic preparedness. Overall, the workshop underscored the need for robust, adaptable modelling frameworks and the integration of different data sources and collaboration across sectors, as key elements in enhancing future pandemic response and preparedness.

COVID-19 case rates, spatial mobility, and neighbourhood socioeconomic characteristics in Toronto: a spatial-temporal analysis.

OBJECTIVES: This study has two primary research objectives: (1) to investigate the spatial clustering pattern of mobility reductions and COVID-19 cases in Toronto and their relationships with marginalized populations, and (2) to identify the most relevant socioeconomic characteristics that relate to human mobility and COVID-19 case rates in Toronto's neighbourhoods during five distinct time periods of the pandemic. METHODS: Using a spatial-quantitative approach, we combined hot spot analyses, Pearson correlation analyses, and Wilcoxon two-sample tests to analyze datasets including COVID-19 cases, a mobile device-derived indicator measuring neighbourhood-level time away from home (i.e., mobility), and socioeconomic data from 2016 census and Ontario Marginalization Index. Temporal variations among pandemic phases were examined as well. RESULTS: The paper identified important spatial clustering patterns of mobility reductions and COVID-19 cases in Toronto, as well as their relationships with marginalized populations. COVID-19 hot spots were in more materially deprived neighbourhood clusters that had more essential workers and people who spent more time away from home. While the spatial pattern of clusters of COVID-19 cases and mobility shifted slightly over time, the group socioeconomic characteristics that clusters shared remained similar in all but the first time period. A series of maps and visualizations were created to highlight the dynamic spatiotemporal patterns. CONCLUSION: Toronto's neighbourhoods have experienced the COVID-19 pandemic in significantly different ways, with hot spots of COVID-19 cases occurring in more materially and racially marginalized communities that are less likely to reduce their mobility. The study provides solid evidence in a Canadian context to enhance policy making and provide a deeper understanding of the social determinants of health in Toronto during the COVID-19 pandemic.

RéSUMé: OBJECTIFS: Cette étude a deux grands objectifs de recherche : 1) examiner les schémas d'agrégation spatiale des baisses de mobilité et des cas de COVID-19 à Toronto et leurs liens avec les populations marginalisées; et 2) cerner les caractéristiques socioéconomiques les plus pertinentes liées à la mobilité humaine et aux taux de cas de COVID-19 dans les quartiers de Toronto au cours de cinq périodes distinctes de la pandémie. MéTHODE: À l'aide d'une approche spatio-quantitative, nous avons combiné des analyses de points chauds, des analyses de corrélation de Pearson et des tests de Wilcoxon à deux échantillons pour analyser des ensembles de données incluant : les cas de COVID-19, un indicateur dérivé d'appareils mobiles pour mesurer le temps passé à l'extérieur du domicile au niveau du quartier (c.-à-d. la mobilité), ainsi que les données socioéconomiques du recensement de 2016 et de l'indice de marginalisation ontarien. Nous avons aussi examiné les variations temporelles entre les phases de la pandémie. RéSULTATS: Nous avons repéré d'importants schémas d'agrégation spatiale des baisses de mobilité et des cas de COVID-19 à Toronto, ainsi que leurs liens avec les populations marginalisées. Les points chauds de la COVID-19 se trouvaient dans des grappes de quartiers plus défavorisés sur le plan matériel, où il y avait davantage de travailleurs essentiels et de personnes passant du temps à l'extérieur de leur domicile. La structure spatiale des grappes de cas de COVID-19 et de la mobilité a légèrement changé au fil du temps, mais les caractéristiques des groupes socioéconomiques communes à toutes les grappes sont restées semblables durant toutes les périodes sauf la première. Nous avons créé une série de cartes et de visualisations pour faire ressortir les schémas spatio-temporels dynamiques. CONCLUSION: Les quartiers de Toronto ont vécu la pandémie de COVID-19 de façons très différentes : les points chauds des cas de COVID-19 sont survenus dans des communautés plus marginalisées sur le plan matériel et racial et moins susceptibles de réduire leur mobilité. L'étude fournit des preuves solides dans un contexte canadien pour améliorer l'élaboration des politiques et approfondir la compréhension des déterminants sociaux de la santé à Toronto pendant la pandémie de COVID-19.

Using spatial and population mobility models to inform outbreak response approaches in the Ebola affected area, Democratic Republic of the Congo, 2018-2020.

The Democratic Republic of the Congo's (DRC) 10th known Ebola virus disease (EVD) outbreak occurred between August 1, 2018 and June 25, 2020, and was the largest EVD outbreak in the country's history. During this outbreak, the DRC Ministry of Health initiated traveller health screening at points of control (POC, locations not on the border) and points of entry (POE) to minimize disease translocation via ground and air travel. We sought to develop a model-based approach that could be applied in future outbreaks to inform decisions for optimizing POC and POE placement, and allocation of resources more broadly, to mitigate the risk of disease translocation associated with ground-level population mobility. We applied a parameter-free mobility model, the radiation model, to estimate likelihood of ground travel between selected origin locations (including Beni, DRC) and surrounding population centres, based on population size and drive-time. We then performed a road network route analysis and included estimated population movement results to calculate the proportionate volume of travellers who would move along each road segment; this reflects the proportion of travellers that could be screened at a POC or POE. For Beni, the road segments estimated to have the highest proportion of travellers that could be screened were part of routes into Uganda and Rwanda. Conversely, road segments that were part of routes to other population centres within the DRC were estimated to have relatively lower proportions. We observed a posteriori that, in many instances, our results aligned with locations that were selected for actual POC or POE placement through more time-consuming methods. This study has demonstrated that mobility models and simple spatial techniques can help identify potential locations for health screening at newly placed POC or existing POE during public health emergencies based on expected movement patterns. Importantly, we have provided methods to estimate the proportionate volume of travellers that POC or POE screening measures would assess based on their location. This is critical information in outbreak situations when timely decisions must be made to implement public health interventions that reach the most individuals across a network.

Estimating the size of the monkeypox virus outbreak in Nigeria and implications for global control.

BACKGROUND: A multi-country outbreak caused by monkeypox virus (MPXV) has been unfolding across endemic and non-endemic countries since May 2022. Throughout April and May 2022, Nigeria reported 31 MPXV cases, of which 11 were confirmed via testing. In May 2022, three internationally exported cases of MPXV, presumed to have originated in Nigeria, were reported, suggesting that a larger than reported outbreak might be occurring in the country. METHODS: We used previously established methods to estimate the true size of the MPXV outbreak in Nigeria. We estimated the incidence rate of exported MPXV cases among all outbound international air travellers from Nigeria during the time period of April and May 2022, using forecasted air traveller volumes. We then applied this incidence rate to the entire population of Nigeria during April and May 2022 assuming that the rate of infection was the same in Nigeria for both travellers and the resident population. Information on the subset of population that were considered to be travellers was obtained from the United Nations World Tourism Organization (UNWTO). RESULTS: We estimated that there were approximately 4000 (N = 4013; 95% CI: 828-11 728) active cases of MPXV in Nigeria in April and May 2022. This is approximately 360-fold greater than the confirmed number and approximately 130-fold greater than the reported number of cases in Nigeria. CONCLUSION: Our findings suggest that a larger outbreak than is appreciated may be ongoing in Nigeria. The observed international spread of MPXV offers important insights into the scale of the epidemic at its origin, where clinical detection and disease surveillance may be limited. These findings highlight the need to expand and support clinical, laboratory, and public health capacity to enable earlier detection of epidemics of international significance.

Pneumonia of unknown aetiology in Wuhan, China: potential for international spread via commercial air travel.

There is currently an outbreak of pneumonia of unknown aetiology in Wuhan, China. Although there are still several unanswered questions about this infection, we evaluate the potential for international dissemination of this disease via commercial air travel should the outbreak continue.

Development of a global infectious disease activity database using natural language processing, machine learning, and human expertise.

OBJECTIVE: We assessed whether machine learning can be utilized to allow efficient extraction of infectious disease activity information from online media reports. MATERIALS AND METHODS: We curated a data set of labeled media reports (n = 8322) indicating which articles contain updates about disease activity. We trained a classifier on this data set. To validate our system, we used a held out test set and compared our articles to the World Health Organization Disease Outbreak News reports. RESULTS: Our classifier achieved a recall and precision of 88.8% and 86.1%, respectively. The overall surveillance system detected 94% of the outbreaks identified by the WHO covered by online media (89%) and did so 43.4 (IQR: 9.5-61) days earlier on average. DISCUSSION: We constructed a global real-time disease activity database surveilling 114 illnesses and syndromes. We must further assess our system for bias, representativeness, granularity, and accuracy. CONCLUSION: Machine learning, natural language processing, and human expertise can be used to efficiently identify disease activity from digital media reports.

The Influence of Sociodemographic Factors on the Engagement of Citizens in the Detection of Dead Corvids During the Emergence of West Nile Virus in Ontario, Canada.

West Nile virus (WNv) was introduced into North America in 1999, and by 2002 was identified in most regions of Ontario, Canada. Surveillance of WNv included testing of corvids found dead and reported by citizens across Ontario, which at the time was a novel citizen science application for disease surveillance. While this surveillance program was successful for timely identification of WNv as it emerged and spread across the province, it is important to consider the influence of non-disease factors on surveillance data collected by the public. The objective of this study was to examine associations between rates of citizen phone reports of dead corvids and sociodemographic factors within the geographic areas where the reports were obtained. The data were grouped by forward sortation area (FSA), a geographical area based upon postal codes, which was linked with census data. Associations between the weekly rate of citizen reports and FSA-level sociodemographic factors were measured using multilevel negative binomial models. There were 12,295 phone call reports of dead corvids made by citizens in 83.3% of Ontario FSAs. Factors associated with the weekly rate of phone reports included the proportion of high-rise housing, the proportion of households with children, the proportion of seniors in the population, the proportion of citizens with no knowledge of either official language and the latitude of the FSA. There were higher rates of citizen phone reports in FSAs with <80% high-rise housing and greater proportions of households with children. A positive and negative association in the rate of calls with the proportion of seniors and latitude of the FSA, respectively, were moderated by the proportion of the population with knowledge of official language(s). Understanding the sociodemographic characteristics associated with citizen reporting rates of sentinels for disease surveillance can be used to inform advanced cluster detection methods such as applying the spatial scan test with normal distribution on residuals from a regression model to reduce confounding. In citizen-derived data collected for disease surveillance, this type of approach can be helpful to improve the interpretation of cluster detection results beyond what is expected.

Infectious disease implications of large-scale migration of Venezuelan nationals.

Background: The ongoing economic and political crisis in Venezuela has resulted in a collapse of the healthcare system and the re-emergence of previously controlled or eliminated infectious diseases. There has also been an exodus of Venezuelan international migrants in response to the crisis. We sought to describe the infectious disease risks faced by Venezuelan nationals and assess the international mobility patterns of the migrant population. Methods: We synthesized data on recent infectious disease events in Venezuela and among international migrants from Venezuela, as well as on current country of residence among the migrant population. We used passenger-level itinerary data from the International Air Transport Association to evaluate trends in outbound air travel from Venezuela over time. We used two parameter-free mobility models, the radiation and impedance models, to estimate the expected population flows from Venezuelan cities to other major Latin American and Caribbean cities. Results: Outbreaks of measles, diphtheria and malaria have been reported across Venezuela and other diseases, such as HIV and tuberculosis, are resurgent. Changes in migration in response to the crisis are apparent, with an increase in Venezuelan nationals living abroad, despite an overall decline in the number of outbound air passengers. The two models predicted different mobility patterns, but both highlighted the importance of Colombian cities as destinations for migrants and also showed that some migrants are expected to travel large distances. Despite the large distances that migrants may travel internationally, outbreaks associated with Venezuelan migrants have occurred primarily in countries proximate to Venezuela. Conclusions: Understanding where international migrants are relocating is critical, given the association between human mobility and the spread of infectious diseases. In data-limited situations, simple models can be useful for providing insights into population mobility and may help identify areas likely to receive a large number of migrants.

A comparison of West Nile virus surveillance using survival analyses of dead corvid and mosquito pool data in Ontario, 2002-2008.

The aim of this study was to improve understanding of the relative performance of the use of dead wild corvids and mosquito pools infected with West Nile virus (WNv) in surveillance for WNv activity in the environment. To this end, all records on dead corvid submissions and mosquito pools tested in Public Health Units (PHUs) in Ontario, from 2002 to 2008, were explored. Survival analyses were employed using the first-WNv-positive cases detected each year for each PHU, and censored observations for PHUs which did not detect WNv during a given year using each data source (504 observations). Survival analyses were employed to compare the number of surveillance weeks before WNv was detected by either data source, and the influence of temporal, geographic and sociodemographic factors on these data. The outcome measurement for the final accelerated failure time (AFT) model with log-logistic distribution was a time ratio, which represents the ratio of the survival time of one group relative to another. Dead corvid surveillance was faster at detecting WNv than testing mosquito pools during the early years of WNv incursion into Ontario, while mosquito testing found WNv more quickly later in the study period. There was also regional variation in time-to-detection of WNv, by modality, as well as for various types of urban/rural settings. In comparison to mosquito surveillance, West Nile virus was detected more quickly using dead corvid surveillance in sparsely populated regions. These areas may benefit from collection of dead corvids to optimize detection and direct early surveillance efforts. When we compared the time-to-detection of WNv using dead corvids and the onset of human cases in PHUs, we found that dead corvid surveillance was predictive of West Nile activity in health units that reported human cases during the first 3 years of the incursion into Ontario.

Exploring relationships between whole carcass condemnation abattoir data, non-disease factors and disease outbreaks in swine herds in Ontario (2001-2007).

BACKGROUND: Improving upon traditional animal disease surveillance systems may allow more rapid detection of disease outbreaks in animal populations. In Ontario, between the years 2001 - 2007, widespread outbreaks of several diseases caused major impacts to the swine industry. This study was undertaken to investigate whether whole carcass condemnation data of market pigs from provincial abattoirs from 2001 - 2007 could have provided useful information for disease surveillance of Ontario swine. The objective was to examine the suitability of these data for detection of disease outbreaks using multi-level models and spatial scan statistics. We investigated the ability of these data to provide spatially-relevant surveillance information by determining the approximate distance pigs are shipped from farm to provincial abattoirs in the province, and explored potentially biasing non-disease factors within these data. RESULTS: Provincially-inspected abattoirs in Ontario were found to be located in close proximity to the hog farms of origin. The fall season and increasing abattoir capacity were associated with a decrease in condemnation rates. Condemnation rates varied across agricultural regions by year, and some regions showed yearly trends consistent with the timing of emergence of new disease strains that affected the Ontario swine population. Scan statistics identified stable clusters of condemnations in space that may have represented stable underlying factors influencing condemnations. The temporal scans detected the most likely cluster of high condemnations during the timeframe in which widespread disease events were documented. One space-time cluster took place during the beginning of the historical disease outbreaks and may have provided an early warning signal within a syndromic surveillance system. CONCLUSIONS: Spatial disease surveillance methods may be applicable to whole carcass condemnation data collected at provincially-inspected abattoirs in Ontario for disease detection on a local scale. These data could provide useful information within a syndromic disease surveillance system for protecting swine herd health within the province. However, non-disease factors including region, season and abattoir size need to be considered when applying quantitative methods to abattoir data for disease surveillance.

Suitability and limitations of portion-specific abattoir data as part of an early warning system for emerging diseases of swine in Ontario.

BACKGROUND: Abattoir data have the potential to provide information for geospatial disease surveillance applications, but the quality of the data and utility for detecting disease outbreaks is not well understood. The objectives of this study were to 1) identify non-disease factors that may bias these data for disease surveillance and 2) determine if major disease events that took place during the study period would be captured using multi-level modelling and scan statistics. We analyzed data collected at all provincially-inspected abattoirs in Ontario, Canada during 2001-2007. During these years there were outbreaks of porcine circovirus-associated disease (PCVAD), porcine reproductive and respiratory syndrome (PRRS) and swine influenza that produced widespread disease within the province. Negative binomial models with random intercepts for abattoir, to account for repeated measurements within abattoirs, were created. The relationships between partial carcass condemnation rates for pneumonia and nephritis with year, season, agricultural region, stock price, and abattoir processing capacity were explored. The utility of the spatial scan statistic for detecting clusters of high partial carcass condemnation rates in space, time, and space-time was investigated. RESULTS: Non-disease factors that were found to be associated with lung and kidney condemnation rates included abattoir processing capacity, agricultural region and season. Yearly trends in predicted condemnation rates varied by agricultural region, and temporal patterns were different for both types of condemnations. Some clusters of high condemnation rates of kidneys with nephritis in time and space-time preceded the timeframe during which case clusters were detected using traditional laboratory data. Yearly kidney condemnation rates related to nephritis lesions in eastern Ontario were most consistent with the trends that were expected in relation to the documented disease outbreaks. Yearly lung condemnation rates did not correspond with the timeframes during which major respiratory disease outbreaks took place. CONCLUSIONS: This study demonstrated that a number of abattoir-related factors require consideration when using abattoir data for quantitative disease surveillance. Data pertaining to lungs condemned for pneumonia did not provide useful information for predicting disease events, while partial carcass condemnations of nephritis were most consistent with expected trends. Techniques that adjust for non-disease factors should be considered when applying cluster detection methods to abattoir data.