Stakeholder experience with artificial intelligence in healthcare: a bibliometric study of satisfaction, trust,...

The implementation of artificial intelligence (AI) in healthcare increasingly depends not only on algorithmic performance but also on how patients and healthcare professionals experience, trust, accept, and engage with AI-enabled systems. Although many bibliometric studies have mapped AI in healthcare, the human-centered evidence base on satisfaction-related stakeholder experience remains fragmented. This study therefore analyzed satisfaction as one component of a broader stakeholder-experience framework that also includes trust, attitude, perception, acceptance, willingness, resistance, and usability.

English-language articles, reviews, and conference papers published between January 1, 2010 and December 31, 2025 were retrieved from the Web of Science Core Collection and Scopus. After duplicate removal and eligibility screening, 1,794 publications were included. Bibliometric analyses were conducted using Microsoft Excel, VOSviewer, CiteSpace, and bibliometrix.

Publication output increased rapidly after 2020 and reached 696 publications in 2025. The United States and China were the leading contributors, and the Journal of Medical Internet Research was the most productive journal. Keyword and co-citation analyses showed a shift from early work on robotic surgery and clinical decision support toward human-centered topics, including patient satisfaction, trust, attitude, ChatGPT, large language models, explainable AI, ethics, and nursing workforce adaptation. Three interconnected frontiers were identified: large-language-model-mediated clinical communication, explainability and trust in patient-AI-professional relationships, and the educational and professional needs of nurses and other healthcare workers.

This study maps the knowledge structure and thematic evolution of research on stakeholder experience with healthcare AI. The findings show that satisfaction should not be treated as interchangeable with trust, acceptance, or usability; rather, these constructs jointly define a multidimensional human-centered implementation problem. Future research should integrate explainable AI, ethical governance, regulatory compliance, and active patient engagement into the evaluation of healthcare AI.