Static Algorithm, Evolving Epidemic: Understanding the Potential of Human-AI Risk Assessment to Support Regional Overdose Prevention

• DOI: 10.1145/3711072
• Publication Date: 2025-02-14
• Authors: Venkatesh Sivaraman; Y. Kwak; Courtney C. Kuza; Qingnan Yang; Kayleigh M. Adamson; Katie J. Suda; Lu Tang; Walid F. Gellad; Adam Perer

Drug overdose deaths, including those due to prescription opioids, represent a critical public health issue in the United States and worldwide. Artificial intelligence (AI) approaches have been developed deployed help prescribers assess patient's risk for overdose-related death, but it is unknown whether experts can leverage similar predictions make local resource allocation decisions more effectively. In this work, we evaluated how AI-based assessment could be used inform using working prototype system. Experts from three departments, of varying locations sizes with respect staff population served, were receptive potential benefits algorithmic prediction AI-augmented visualization connect across data sources. However, they also expressed concerns about model's formulation underlying would match state epidemic as evolved their specific locations. Our findings extend other studies on systems sector, present opportunities future human-AI collaborative tools support decision-making local, time-varying contexts.