Conceptual and Computational Representations of Influenza Vaccine Effectiveness

Influenza epidemic dynamics are impacted by many factors including transmissibility of the circulating strain, immunity of the population, social distancing behaviors, vaccine efficacy, and other factors. Vaccine efficacy can be broken down into three separate parameters to represent the different effects the vaccine can have on protecting an individual (1) against infection, (2) against symptoms, or (3) against transmitting the infection on to others. Using the University of Pittsburgh-developed agent-based modeling platform FRED “Framework for Reconstructing Epidemiological Dynamics” we simulated influenza epidemics based on the interactions of individual persons and their individual properties and behaviors and explored the effects of vaccines with varying efficacy profiles. The model contains multiple interacting conditions including influenza disease and transmission, residual immunity, waning immunity, and staying at home behaviors, as well as variable vaccine efficacies. To ensure realistic results, the transmissibility of influenza and age specific probabilities of being symptomatic were fit to real- world epidemiological data using parameter sweeps. Pilot runs of the model were then produced with varying assumptions about vaccine efficacy profiles. The outcomes from these simulations have public health relevance because they can help increase understanding of the different components of the vaccine, reconcile outcomes of a season, and guide public policy and effective interventions. Being able to simulate an influenza season is vital in helping inform public health policy and practice