van den Berg L, Engelsma T, Dusseljee-Peute L. Exploring the User Experience of Generative AI in Digital Health: A Rapid Review. Stud Health Technol Inform. 2025;326:12-16. doi:10.3233/SHTI250227
A rapid review was conducted in October 2024 using four databases: PubMed, IEEE, ACM, and Scopus. The review focused on three main topics: artificial intelligence, user experience (UX), and digital health. Studies were included if they were published after 2022 and discussed how Generative AI (GenAI) is used in healthcare, especially in relation to UX and usability. Out of 456 articles initially identified, 91 duplicates were removed, resulting in 365 remaining articles. Only 10 (about 3%) met the criteria. These included five full papers, four conference papers, and one short report. The selected studies were categorized into three groups based on GenAI usage: ChatGPT-specific, ChatGPT-tailored, and custom GPT models. Four studies looked at ChatGPT’s usability, quality, and reliability in healthcare settings. The other six focused on integrating ChatGPT into digital health technologies (DHTs) or creating custom GPT models. Usability findings varied. ChatGPT received mixed scores and was not seen as reliable or trustworthy enough for processing health information. However, ChatGPT-tailored DHTs showed reduced burden and burnout, high usability scores, and highlighted that usability influences user trust in GenAI. A major issue identified was the lack of standardized methods for evaluating usability in AI healthcare tools. Although terms like “usability” and “user experience” were common, their definitions and measurement methods varied, making it hard to compare results across studies. Finally, only one study clearly stated which AI models were used (GPT-3.5 Turbo and GPT-4), while others were vague. While this rapid review provides insights into future directions for utilizing GenAI in digital health, more research and top-down standardizations are needed to move the field forward.