LloydChapman/COVID_homeless_modelling

This repository contains Approximate Bayesian Computation (ABC) code and simulation code for the analyses in 'Comparison of infection control strategies to reduce COVID-19 outbreaks in homeless shelters in the United States: a simulation study' [1]. The code implements a discrete-time stochastic SEIR simulation model of COVID-19 transmission in a closed environment (here a homeless shelter) with importation of infection from the local community. The model is fitted to data on numbers of PCR-positive and negative individuals from outbreaks in 5 homeless shelters in San Francisco, Boston and Seattle, and used to predict the impact of different intervention strategies on the probability of averting an outbreak over 30 days in a representative homeless shelter into which a single latently infected individual is introduced.

Keywords

COVID-19; SARS-CoV-2; Homelessness; Shelters; Infection control; Outbreaks; Symptom-based screening; PCR testing; Universal masking