A Computational Model Assessing Population Impact of a New Tobacco Product

From the paper

Abstract

OBJECTIVES We developed and validated a computational model to assess the potential health impact of a new tobacco product in the U.S. market. METHODS An Agent-Based Model (ABM) framework was used to estimate changes in tobacco use prevalence and premature deaths based on the difference between modified (counterfactual) and base case (status quo) scenarios. The hypothetical population transitions between different tobacco-use states based on their attributes and transition probabilities over the simulation period. A transition sub-model coupled with mortality sub-models and excess relative risk (ERR) ratio estimates determine survival probability over time. To demonstrate the functionality and capability of our ABM, we modeled a scenario to simulate the population health impact a new tobacco product on the U.S. market. We also demonstrated sensitivity analyses by adjusting key input parameters. RESULTS Our simulation, based on modified- and base-case hypothetical populations using reliable and publicly available input sources, predicts a net benefit to the population with a decrease in premature deaths and cigarette smoking prevalence. CONCLUSION Our computational model, leveraging ABM to assess population impact, is a fit-for-purpose tool for predicting public health outcomes.