The primary scientific aim of this grant is to uncover how hidden transmission, i.e., unreported or sub-clinical infection, affects the emergence, re-emergence, and persistence of infectious diseases. Specifically, the project will focus on whooping cough (pertussis); however, the computational, statistical, and mathematical methods are readily applicable to a range of diseases. During an ongoing outbreak, accurate forecasts, risk assessments, and prioritization of intervention strategies depend on reliable estimates of both the reporting rate and the prevalence of subclinically or asymptomatically infectious individuals. However, both are often dynamic, potentially varying in time, space, and/or with outbreak size, and are driven by intrinsic properties of the pathogen and complex, extrinsic factors, e.g., human behavior change, the availability of vaccines and pharmaceuticals, evolving diagnostic practices, and available health-care infrastructure. Mechanistic models that capture these factors are critically important for estimating, reducing, and correctly accounting for reporting and subclinical infection, but they do not yet exist. The project proposes the construction of an integrative framework, combining a mechanistic model of asymptomatic transmission and under-reporting with genomic and epidemiological data using a phylodynamic- based inference method, to uncover the integrated?both molecular and epidemiological?variables associated with whooping cough resurgence and persistence.
The specific aims of this project are, 1) to develop a computational toolkit for studying hidden transmission, 2) to construct a phylodynamic framework for studying hidden transmission, and 3) to quantify the prevalence of hidden B. pertussis transmission globally. Methods will include combining an agent-based simulation model of whooping cough transmission with Sequential Monte Carlo (SMC) particle filtering, and combining phylodynamic models of disease outbreaks with statistical methods to estimate relevant epidemiological parameters from sequence data. Publicly available B. pertussis genomes from clinical isolates, and publicly available epidemiological data from several countries will be used. The proposed research will have an additional, practical benefit, by significantly improving our understanding of how disease reporting rate and subclinical carriage alter transmission dynamics. More specifically, an increased understanding refers to a quantification of risk?both in space and time?with a determination regarding how biased reporting rates and/or subclinical burden can hinder public health and/or clinical decision-making. The improvement in risk assessment would advance and support clinical and public health responses, both in terms of person-time as well as the allocation of strategic federal, state, and hospital-level resources.