A0299
Title: An integrated population model to estimate the population size of persons contending with homelessness
Authors: Gracia Dong - University of Toronto (Canada) [presenting]
Laura Cowen - ()
Abstract: Open population capture-recapture methods have been used with electronic healthcare records to estimate the population size of hidden vulnerable human populations, such as persons contending with homelessness. These methods rely on various assumptions about the data, which are often violated when data is not complete. Data privacy concerns can also limit access to computational resources, necessitating the need for more efficient model fitting or alternative modelling approaches. Integrated population models, which jointly analyse survey data from point-in-time counts and capture-recapture data from electronic healthcare records, can be used to provide more effective estimates. Electronic healthcare records spanning 2013-2023 are used from the Vancouver Island Health Authority, British Columbia, to identify adults in Greater Victoria contending with homelessness based on their self-reported housing status and the locations of the healthcare services used. Additionally, point-in-time counts (2018, 2020) are incorporated, which offer a snapshot of the homeless population on specific days. These counts include a breakdown by demographic data, proportions of individuals with underlying health concerns, and information on healthcare services accessed within the past year, including the proportion of individuals with emergency room visits.
