High fidelity Location-based Risk Estimates for COVID-19 to Guide Reopening of Los Angeles

A novel coronavirus (or COVID-19) has rapidly spread across the USA and social distancing and stay-at-home measures have been adopted in many states—significantly disrupting our daily lives. Researchers and policymakers are struggling to fully understand the nature of COVID-19 transmission and how to respond to the outbreak. As U.S. policymakers are planning to lift restrictions and reopen businesses, many what if scenarios need to be investigated considering the spatio-temporal mobility dynamics of a region. Epidemiological models cannot capture the underlying spatio-temporal dynamics of human movements at a high resolution that can support a nuanced approach of reopening. Existing agent-based models on COVID-19 have considered limited data on population movement. On the other hand, data-driven approaches (e.g., statistical or machine learning models) project the spread of the disease based on historical data only, while the underlying mobility dynamics and policies may change in future. In this study, we develop an agent-based model (ABM) that combines individual-based data (spatial distribution of households, mobility behavior, and social vulnerability) and infection transmission process to estimate the risk for COVID-19 cases. We simulate multiple scenarios using SafeGraph mobility data and epidemiological factors to understand the effect of mobility and policy changes on COVID19 risks. From the simulated results we estimated a location-based risk score on what to expect if reopened or mobility pattern changes based on different parameters. The proposed method and risk score are easily scalable and can support policy makers in reopening decision.

Type: Health Care
Release Date: Jun 09, 2020
Last Updated: Jun 12, 2020

Average rating is 5.0 with 2 vote(s)

Please sign in or create an account to give a rating or comment.

Please sign in or create an account to view the download file