MPOPHC: Quantitative design of effective testing-based policies through infection trajectory modeling

Diagnostic tests play a crucial role in the management of infectious disease transmission. Testing is the fastest most reliable way to inform a person whether they are infected, and thus whether they should adjust their behavior to prevent onward spread. Testing policies have long contributed to public health, including in the control of HIV, tuberculosis, and malaria. During the COVID-19 pandemic, various test-based policies were successful, including pre-event screening (e.g., testing before entering a sporting event), traveler screening (e.g., testing before boarding a flight), and regular screening (e.g., weekly testing at universities). Such policies could also help control the spread of other existing and novel respiratory pathogens. However, we currently lack a robust, data-driven framework to estimate the potential impact of testing-based infection control strategies in general. To fill this gap, this project will develop a flexible modeling framework to simulate how different testing policies might perform for various pathogens, tests, and human behavioral scenarios. This project will also develop the statistical tools needed to infer how diagnostic test results, infectiousness, and behavior relate to one another, informed by data on SARS-CoV-2 and other respiratory pathogens. To maximize the impact of these findings, this project will build mature, open-source software products to compare testing-based policies, accompanied by tutorials for policymakers and a new open-source data hub to consolidate information relevant to testing-based policies. The successful completion of this project will improve our ability to control existing respiratory pathogens and enhance our preparedness for future pandemics. Fundamental to this project is the characterization of how infectiousness, detectability, symptoms, and behaviors change over the course of a respiratory infection - a collection of features called an infection trajectory. While the details of an infection trajectory can be omitted for some types of policy assessments, testing-based policies depend critically on an accurate and statistical understanding of infection trajectories. Infection trajectory-based models allow for the separation of individual-level features of disease transmission from the between-host dynamics, permitting a "plug-and-play" approach to policy design, without compromising the ability to tailor solutions to local needs and populations. This project's policy modeling framework will develop a stochastic description of infection trajectories, represented by a joint distribution of an infection's measurable variables. This will allow the researchers to assess variability in policy outcomes and to identify cross-policy interactions. This project will develop a framework to infer infection trajectory distributions from multimodal data and will deploy that framework to guide the design of studies for collecting new infection trajectory data. Finally, this project will create a suite of software, educational, and data tools for informing infection trajectories and associated policies. For the public health policy community, successful completion of this project will produce new, high-quality policy design models and assessment tools, complemented by educational and interactive exploration webpages. For the scientific community, this project will provide statistical tools and data sharing standards for infection trajectory data, supporting advances in virology and modeling. This award is co-funded by the NSF Division of Mathematical Sciences (DMS) and the CDC Coronavirus and Other Respiratory Viruses Division (CORVD). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.