Core outcome sets (COS) improve the consistency and comparability of outcome selection across clinical studies. However, their development, implementation, and updating still require considerable time and effort, including information identification, classification, and terminology standardization. Artificial intelligence (AI) and other intelligent technologies may support these processes, particularly through text recognition, information extraction, and automated classification. Yet evidence on their use in COS research remains fragmented, and the views of researchers and other stakeholders have not been systematically examined. This study will map current applications, identify suitable uses and practical limitations, and explore future directions. The findings will inform the development of a methodological framework for AI-assisted COS development.
Shiguang Chai, Zhao Chen, Zhiyue Guan, Siyi Li, Shuangqiu Wang, Lulu Shi, Naipisa Wumaierjiang, Yadan Tan, Hongcai Shang, Ruijin Qiu
Disease Category: Other
Disease Name: N/A
Age Range: Unknown
Sex:
Nature of Intervention:
- Clinical experts
- Methodologists
- Researchers
- COS methods research
- Interview
- Survey
- Systematic review
This study will use a mixed-methods design. In Phase 1, we will conduct a systematic review. We will search PubMed, Embase, Web of Science, Scopus, CNKI, WanFang Data, SinoMed and the COMET database. We will include studies on COS research that report the use of AI or other intelligent technologies. We will select studies and extract data according to predefined criteria, and use descriptive analysis to summarise COS research stages, technology types, supported tasks and evidence gaps. In Phase 2, we will conduct an online survey among researchers in COS and AI-related fields to collect their views on the roles, suitable stages, practical limitations and future directions of AI in COS research. We will use descriptive analysis for closed-ended questions and thematic analysis for open-ended responses. In Phase 3, we will conduct semistructured interviews with clinicians, clinical researchers, methodologists and AI experts in China to clarify key survey findings and explore their views and experiences. Interview data will be analysed using thematic analysis.