AUTHOR=Wang Liuping , Zhang Zhan , Wang Dakuo , Cao Weidan , Zhou Xiaomu , Zhang Ping , Liu Jianxing , Fan Xiangmin , Tian Feng 

TITLE=Human-centered design and evaluation of AI-empowered clinical decision support systems: a systematic review

JOURNAL=Frontiers in Computer Science

VOLUME=Volume 5 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/computer-science/articles/10.3389/fcomp.2023.1187299

DOI=10.3389/fcomp.2023.1187299

ISSN=2624-9898

ABSTRACT=Artificial intelligence (AI) technologies are increasingly applied to empower clinical decision support systems (CDSS), providing patient-specific recommendations to improve clinical work. Equally important to technical advancement is human, social, and contextual factors that impact the successful implementation and user adoption of AI-empowered CDSS. With the growing interest in human-centered design and evaluation of such tools, it is critical to synthesize the knowledge and experiences reported in prior work and shed light on future work. To that end, we conducted a systematic review to gain an in-depth understanding of how AI-empowered CDSS was used, designed, and evaluated, and how clinician users perceived such systems. Our findings highlighted six major challenges, including technical limitations, workflow misalignment, attitudinal barriers, informational barriers, usability issues, and environmental barriers. These sociotechnical challenges prevent the effective use of AI-based CDSS interventions in clinical settings. We further discuss design implications and future research directions.