Validation and integration in spread models of influenza: scientific insights and policy implications during influenza epidemics/pandemics

Influenza presents many challenges to society, leading to severe impacts in terms of social, economic and health-care costs. To minimize these impacts, models for the spatial spread of influenza help us prepare and plan for epidemic/pandemic events. These models also increase our scientific understanding about the epidemic process and identify optimal mitigation strategies during such events. Given the human experience with past pandemics and severe seasonal epidemics, modeling studies will continue to be a useful tool for policy-makers in reducing the burden of influenza on society. I highlight two avenues of research which may enhance our understanding of the epidemic process and improve the use of models for setting and implementing policy.Validation remains limited and predictive validation is almost non-existent in complex simulation models of influenza spread. This is a serious concern because policy-makers use predictions from such models as inputs for making important decisions. Current models of influenza spread are coming under increased scrutiny for their lack of predictive ability, but it seems that no one has actually evaluated their predictive ability in the first place. To fill this gap in knowledge, I demonstrate the process of predictive validation by generalizing an individual-based model for the spread of influenza to the urban area of Montreal, Canada. Using this model and extensive data on several past epidemics, I show that the reliability and timing of several epidemic metrics depends on two important factors: the method of forecasting and the type of the epidemic metric which we want to forecast.Predictors of health disparities are not included in current models of influenza spread. This is despite an extensive literature showing that these predictors are related to burden of influenza in vulnerable subpopulations of society. Through formulating two different integrated models, I illustrate novel approaches to address this limitation. In the first model, I integrate social deprivation within an individual-based model for the spread of influenza. Using this model, I examine hypotheses about the relationship between social deprivation and influenza burden. In the second model, I integrate socioeconomic information in a metapopulation model. I develop a novel social-attributes gravity model to describe local-scale contact processes. I perform a theoretical analysis of this model to show the consequences of local-scale heterogeneity, in contact and susceptibility, on large-scale epidemic patterns. For both models, I show their practical application through evaluating vaccination strategies which make use of never-before-available data within complex and dynamic models of influenza spread.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.