Quantitatively Exploring Perceived risk on use of an Artificial Intelligence System: Blood Utilization Calculator_RCode

Given the opportunities created by artificial intelligence (AI) based decision support systems in healthcare, the vital question is whether clinicians are willing to accept and use this technology as an integral part of routine clinical practice.  It is critical to explore various factors including the perception of AI, risk, and expectancy, and their influence on specific AI tools. In this study, we explored this relationship for  AI-based Blood Utilization Calculator (BUC) first time. BUC delivers data-driven personalized recommendations for the number of packed red blood cells to transfuse for a given patient. A purposeful sampling strategy was utilized to exclusively include BUC users who are clinicians in a university hospital in Wisconsin. We received 119 complete responses. We leveraged structural equation modeling to find out that perception of AI safety (in general) and expectancy of BUC had a negative impact on the perceived risk of BUC. We also observed that the perceived risk of BUC mediated the effect of expectancy and perception of AI on clinicians’ intent to use BUC. In an inpatient setting, clinicians are one of the most important stakeholders of AI technologies. Due to the nature of healthcare services, the implementation of AI should be performed with certain considerations. Our model suggests that a clinician’s perception of risk is a crucial factor. AI developers need to emphasize the benefits of AI technology, ensure ease of use, clarify the system's potential, and minimize the risk perceptions by improving the overall design.