From Algorithm to Policy: A Bibliometric Analysis of Implementation Science and Governance Frameworks in AI Healthcare Research (2016–2026)

This study presents a bibliometric analysis of AI healthcare research related to implementation science and governance frameworks from 2016 to 2026. A systematic search of the Scopus database identified 3780 peer-reviewed articles published across 1500 sources. The dataset was analyzed using Biblioshiny. The findings show an annual growth rate of 47.65% in governance-focused publications, exceeding the growth rate of technical AI research. Four main research themes were identified: regulatory compliance, ethical frameworks with limited operational measures, organizational readiness, and clinical workflow integration. The United States, China, and the United Kingdom are the leading contributors, while the Journal of Medical Internet Research and BMJ Open are among the main publication outlets. International collaboration (34.66%) remains concentrated among high-income countries. Thematic development has progressed from general ethical discussions to pandemic-related applications and more specific regulatory frameworks. Three research gaps contribute to the algorithm-to-policy translation deficit: the principles–practice gap, the regulatory–evidence gap, and the innovation–implementation gap. This study proposes an integrated governance framework based on five evidence-based principles. The framework is a conceptual model derived from bibliometric findings and requires further empirical validation in clinical settings before practical adoption. The study contributes by providing a bibliometric analysis of AI governance research and introducing the concept of the “algorithm-to-policy translation deficit” as an analytical framework. It also offers a structure to support future research and practice toward safe, effective, and equitable clinical implementation of AI. These findings guidance for regulators, healthcare organizations, AI developers, and researchers.