Transforming Healthcare Delivery through Technology: Analyzing Hospital-at-Home

Hospital-at-home (HaH) programs provide acute care services in patients’ homes within a technologically enriched environment. These programs depend on robust health information systems (HIS), including telehealth platforms, electronic health records, and digital monitoring infrastructure. Prior studies suggest that HaH care improves patient outcomes and satisfaction while enhancing hospital capacity and alleviating resource strain (Truong & Siu, 2024). However, the geographic distribution and healthcare accessibility of HaH programs remain understudied, particularly from a geospatial perspective. Health conditions and outcomes vary significantly across U.S. counties, with notable differences in chronic disease prevalence and healthcare access. HaH programs may offer a viable solution for vulnerable areas facing high hospital closure rates and housing large populations with chronic diseases. This study examines HaH programs across the United States, focusing on the nuanced attributes and health outcomes of the areas they serve. Our primary research question is: How are demographics, socioeconomic factors, and digital infrastructure and literacy associated with the health outcomes of HaH programs? We analyze data on 372 hospitals that implemented HaH programs as of November 2024, spanning 216 counties across 39 states. Data sources include the Area Health Resources dataset, the Digital Divide Index, and the American Hospital Association database, offering insights into digital infrastructure and literacy, socioeconomic and demographic factors, and health outcomes. T-tests are used to identify statistically significant differences between independent groups, such as counties with and without HaH programs. Multiple regression analysis is applied to investigate the relationships between various factors (e.g., demographics, socioeconomic indicators, and digital divide) and outcome variables (e.g., hospital readmission rates and healthcare costs). Additionally, we use machine learning techniques to identify non-linear relationships between factors and health outcomes. Preliminary results suggest that counties with HaH programs exhibit lower hospital readmission rates and per capita Medicare costs compared to counties without HaH coverage. Furthermore, our analysis indicates that factors such as the elderly population, poverty levels, health insurance enrollment, education levels, and rurality are associated with these health outcomes. Notably, these associations are not linear, suggesting threshold effects wherein socioeconomic and demographic factors influence health outcomes differently at varying levels. We will continue our analysis and expect to present additional findings at the AMCIS conference. Our study aims to inform public health policies to enhance regional health outcomes and well-being. The insights gained will help identify vulnerable areas in need of healthcare support and guide initiatives to improve digital infrastructure and literacy.