COVID-19 Risk Estimation using a Time-varying SIR-model

TL;DR

This paper introduces a hybrid data-driven model using a time-varying SIR framework to estimate COVID-19 risk levels, providing real-time, community-specific risk scores with uncertainty quantification.

Contribution

It presents a novel risk scoring method based on a time-varying SIR model that quantifies infection risk and uncertainty, and applies it to real-world data from Los Angeles.

Findings

Risk score $\Gamma_t$ correlates with infection probability within 24 hours.

Method provides confidence intervals for risk estimates.

Open-source code and data support practical application.

Abstract

Policy-makers require data-driven tools to assess the spread of COVID-19 and inform the public of their risk of infection on an ongoing basis. We propose a rigorous hybrid model-and-data-driven approach to risk scoring based on a time-varying SIR epidemic model that ultimately yields a simplified color-coded risk level for each community. The risk score $\Gamma_t$ that we propose is proportional to the probability of someone currently healthy getting infected in the next 24 hours. We show how this risk score can be estimated using another useful metric of infection spread, $R_t$, the time-varying average reproduction number which indicates the average number of individuals an infected person would infect in turn. The proposed approach also allows for quantification of uncertainty in the estimates of $R_t$ and $\Gamma_t$ in the form of confidence intervals. Code and data from our effort have been open-sourced and are being applied to assess and communicate the risk of infection in the City and County of Los Angeles.