The effect of Behavioral Factors and Intervention Strategies on Pathogen Transmission: Insights from a Two-Week Epidemic Game at Wenzhou-Kean University in China

Abstract

BackgroundEffective control of infectious diseases relies heavily on understanding transmission dynamics and implementing interventions that reduce the spread. Non-pharmaceutical interventions (NPIs), such as mask-wearing, social distancing, and quarantining, are vital tools in managing outbreaks where vaccines or treatments are limited. However, the success of NPIs is influenced by human behavior, including compliance with guidelines, and attitudes such as beliefs about the effectiveness of interventions. In this study, we applied an innovative proximitybased experimentation platform to generate empirical data on behaviors and attitudes and their effect on disease transmission. Our platform uses a smartphone application that enables the spread of a digital pathogen among participants via Bluetooth during open-world “experimental epidemic games”. This creates an environment for epidemiology field experimentation where researchers can control transmission mechanics and collect full ground-truth datasets.MethodsOur study employed the “epidemic” app to investigate the impact of risk perception and compliance to NPIs on pathogen transmission. Involving nearly 1,000 participants in a two-weeks long epidemic game at Wenzhou-Kean University (WKU) in China, the app generated a multimodal dataset, which allowed us to develop and parameterize Susceptible-Exposed-Infected-Recovered (SEIR) models. We quantified the extent by which behavioral factors, such as risk perception and compliance with quarantine, and strength of intervention strategies influence disease transmission. The model incorporates time-varying transmission rates that reflect changes in attitudes and behavior, and we calibrated it using the empirical data from the epidemic game to provide critical insights into how variations in NPI compliance levels affect outbreak control.FindingsThe findings reveal that adherence to NPIs alone, which is influenced by changes in behavior and attitudes, may not result in the expected reduction in transmission, illustrating the complex interplay between behavioral factors and epidemic control. Moreover, the model further shows that changes in risk perception coupled with NPI adherence could significantly reduce infection levels as well as susceptibility.InterpretationOur study highlights the usefulness of experimental epidemic games to generate realistic datasets, and the importance of integrating behavioral dynamics into epidemiological models to enhance the accuracy of predictions and the effectiveness of public health interventions during infectious disease outbreaks.Research in ContextEvidence before this studyWe conducted a comprehensive review of the existing literature to evaluate the current state of knowledge regarding empirically-informed infectious disease modeling, with a particular focus on the role of human behavior and non-pharmaceutical interventions (NPIs) in mitigating disease transmission. Our search spanned databases such as PubMed, MEDLINE, and Web of Science, targeting publications up to March 1, 2024, using keywords including “infectious disease modeling,” “simulation,” “experimental game,” “human behavior,” “non-pharmaceutical interventions,” and “epidemiology.” While a substantial body of research explores the influence of human behavior on disease dynamics, there is a notable gap in studies that integrate large-scale mobility and behavioral data collected with smartphone apps within open-world environments, such as a university campus. Most existing studies fail to incorporate the complexity of real-time human behavioral responses and NPIs, which are crucial for accurately modeling the dynamics of disease transmission in such contexts.Added value of this studyThis study is the first to use our proximity-based experimentation platform to conduct an epidemic game in a large-scale university setting while integrating human behavioral factors and NPIs into a mechanistic modeling framework. By employing a flexible, time-varying transmission rate model, our research highlights the impact of human behavior and NPIs on pathogen spread dynamics. This novel approach provides a more accurate and nuanced depiction of real-world transmission scenarios, as observed during the proximity-based experiment. Through the integration of empirical data from nearly 1,000 participants, combined with detailed model simulations and rigorous sensitivity analyses, we offer insights into how timely and coordinated interventions, alongside public compliance, can significantly influence the trajectory of an outbreak. This study underscores the necessity of adaptive strategies in outbreak management and presents a robust framework that can inform and enhance future public health planning and response efforts.Implications of all the available evidenceOur findings underscore the pivotal role of experimental and computational approaches for generating realistic outbreak datasets and integrating behavioral dynamics and NPIs into epidemiological models. This results in significantly more accurate models that then can become valuable tools for public health planning. The study provides a solid foundation for refining models with additional complexities, such as age-based behaviors, and offers a framework for optimizing outbreak management and future pandemic preparedness.