2020/04/10 by Claire Donnat, Susan Holmes, Donnat, Claire +1
Mathematics · Medicine · #Applications (stat.AP) #COVID-19 epidemiological studies #FOS: Biological sciences #FOS: Computer and information sciences #Mathematical and Theoretical Epidemiology and Ecology Models #Populations and Evolution (q-bio.PE)
paper · pdf · doi:10.48550/arxiv.2004.05272
openalex publication_date 2020/04/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The correct evaluation of the reproductive number R for COVID-19 -- which\ncharacterizes the average number of secondary cases generated by each typical\nprimary case -- is central in the quantification of the potential scope of the\npandemic and the selection of an appropriate course of action. In most models,\nR is modeled as a universal constant for the virus across outbreak clusters\nand individuals -- effectively averaging out the inherent variability of the\ntransmission process due to varying individual contact rates, population\ndensities, demographics, or temporal factors amongst many. Yet, due to the\nexponential nature of epidemic growth, the error due to this simplification can\nbe rapidly amplified and lead to inaccurate predictions and/or risk evaluation.\nFrom the statistical modeling perspective, the magnitude of the impact of this\naveraging remains an open question: how can this intrinsic variability be\npercolated into epidemic models, and how can its impact on uncertainty\nquantification and predictive scenarios be better quantified? In this paper, we\npropose to study this question through a Bayesian perspective, creating a\nbridge between the agent-based and compartmental approaches commonly used in\nthe literature. After deriving a Bayesian model that captures at scale the\nheterogeneity of a population and environmental conditions, we simulate the\nspread of the epidemic as well as the impact of different social distancing\nstrategies, and highlight the strong impact of this added variability on the\nreported results. We base our discussion on both synthetic experiments --\nthereby quantifying of the reliability and the magnitude of the effects -- and\nreal COVID-19 data.\n