Improving Health and Housing Outcomes through a Simulation and Economic Model (iHOUSE Model): An evidence-based protocol of a group model building approach to develop an agent-based model

Danielle Marie Kline^1*Pranav Padmanabhan^1,2Sarah E Brewer^3,4Magdalena Cerdá⁵Elysia Versen⁶Katherine M Keyes⁷Margot Kushel^10,8,9Erin C Wilson¹¹Paul Wesson^12,13Ayaz Hyder¹⁴Alaina Boyer¹⁵Alia Al-Tayyib¹⁶Joshua A Barocas^1,17*

• ¹Department of Medicine, Division of General Internal Medicine, University of Colorado, School of Medicine, Aurora, United States
• ²Colorado School of Public Health, Aurora, United States
• ³Adult & Child Center for Outcomes Research & Delivery Science, University of Colorado, Aurora, United States
• ⁴Department of Family Medicine, University of Colorado, School of Medicine, Aurora, United States
• ⁵Department of Population Health, New York University Grossman School of Medicine, New York, United States
• ⁶Colorado Evaluation & Action Lab, University of Denver, Denver, United States
• ⁷School of Public Health, Columbia University Medical Center, New York, United States
• ⁸Division of Health Equity and Society, University of California, San Francisco, San Francisco, United States
• ⁹Benioff Homeless and Housing Initiative, University of California, San Franscisco, San Franscisco, United States
• ¹⁰Center for Vulnerable Populations, University of California, San Francisco San, San Francisco, United States
• ¹¹Public Health Foundation Enterprises, Inc., City of Industry, United States
• ¹²Benioff Homeless and Housing Initiative, University of California, San Francisco, San Francisco, United States
• ¹³Department of Epidemiology and Biostatistics, University of California, San Francisco San Francisco, San Francisco, United States
• ¹⁴Independent consultant, DataToEngage (Futuwwa LLC), Columbus, United States
• ¹⁵National Health Care for the Homeless Council, Nashville, United States
• ¹⁶Public Health Institute, Denver Health, Denver, United States
• ¹⁷Department of Medicine, Division of Infectious Diseases, University of Colorado, SOM, Aurora, United States

Introduction: Homelessness in the United States increased every year since 2016, with a 38% increase from 2023 to 2024. Much of the increase is attributable to rising home and rent costs, economic hardship caused by the recent pandemic, and the ending of protective legislation. Notably, people who experience homelessness have an increased risk of substance use disorders, HIV infection and poorer HIV outcomes than people who are stably housed. The iHouse model aims to develop feasible, effective, and cost-effective tailored approaches to improve health outcomes in this population including life expectancy, overdose, and HIV.The study will employ Group Model Building methods and use insights from that process to develop an agent-based model simulating the dynamic processes contributing to HIV incidence and treatment, overdose, and life expectancy among people along the housing and homelessness continuum in Denver, CO and San Francisco, CA. The model will evaluate multiple outcomes from 4 conceptual dimensions: 1) movement along the housing continuum, 2) population health (overdose and HIV incidence and life expectancy), 3) budgetary impact, 4) economic value.This study has been approved by the Colorado Institutional Review Board at the University of Colorado under protocol 24-0878. The data generated by this protocol, the methodologies used, and the findings will be made available in a timely manner to other researchers. iHOUSE code and parameter values will be published in Git Hub, such that all model analyses can be reproduced by independent investigators. Documentation of all parameter estimates and model results will be published for independent review and confirmation. In addition, supplemental materials and appendices for the model will be shared on a publicly available website.