A Model Predicting Artificial Intelligence Use by Gastroenterology Nurses in Clinical Practice: A Cross-Sectional Multicenter Survey

Background and Aims: Nurses' participation during colonoscopy has been demonstrated to significantly improve the detection rate of polyps and adenomas. Nonetheless, the adoption of AI in clinical practice still poses challenges. There is limited understanding of the factors influencing gastroenterology nurses' intentions to use AI in clinical practice. We aimed to examine how gastroenterology nurses' intentions to use AI are affected by perceived usefulness, acceptance of this technology, and perceived risk via a moderated mediation model controlling for nurses' characteristics. Methods: A cross-sectional multicenter survey study was conducted among gastroenterology nurses from 54 hospitals in Taiwan, Hong Kong, and mainland China. A total of 337 nurses (mean age 37.40 ± 8.29 years, 81.6% females) completed the survey. Results: After controlling for previous experience with AI, number of working years, and work role, a statistically significant direct effect of perceived usefulness on use intention was found. The indirect effect of perceived usefulness on use intention through AI technology acceptance was the most robust when perceived risk was at the lowest level. Conclusions: Findings suggest that perceived usefulness facilitated the intentional use of AI in clinical practice through acceptance of AI, especially when perceived risk was low.