Artificial Intelligence in Healthcare: A National Framework for Evaluating Economic, Productivity, Public Health, and Environmental Impacts

Abstract

Artificial intelligence is being adopted across national healthcare systems at an accelerating pace, promising substantial economic savings, productivity gains, and improved public health outcomes through more precise, targeted, and efficient care delivery. This expansion, however, depends on an increasingly energy- and resource-intensive computing infrastructure whose own environmental and public health footprint has received comparatively limited attention within healthcare-specific AI evaluation frameworks. This article proposes and examines a national framework for evaluating the economic, productivity, public health, and environmental impacts of artificial intelligence in healthcare as an integrated system, rather than as separate, independently assessed domains. It synthesizes national frameworks for AI-driven precision public health and chronic disease prevention, economic impact estimation of AI adoption in healthcare, and national-scale predictive analytics for identifying high-risk populations within Medicare and Medicaid, alongside recent evidence quantifying the public health and environmental costs of the data center infrastructure that underlies large-scale AI deployment. It further draws on subsurface reactive transport modeling and reactive heat exchanger engineering for long-term geothermal reservoir performance, and high-resolution petrophysical characterization of geothermal reservoir quality, to examine geothermal energy's emerging role as a low-carbon power source capable of mitigating the environmental footprint of the computing infrastructure that underlies AI-driven healthcare. The article concludes by identifying persistent gaps in cross-domain evaluation methodology, infrastructure transparency, and health equity, and proposes a research agenda for advancing an integrated national evaluation framework for AI in healthcare.