P170 - PERCEPTIONS OF REGISTERED DIETITIANS ON THE USE OF AI DURING DIETETIC CONSULTS: A SURVEY

P170

PERCEPTIONS OF REGISTERED DIETITIANS ON THE USE OF AI DURING DIETETIC CONSULTS:

A SURVEY

J. Edakkanambeth Varayil1,*, R. Hurt1, M. Mundi1, O. Mohamed Elfadil1, G. Kolar2

1Home Parenteral and Enteral Nutrition, 2Hospital Internal Medicine, Mayo Clinic, Rochester, United States

 

Rationale:  The role and utilization of artificial intelligence in healthcare has been rapidly increasing. There is a paucity of data exclusively on the perceptions of registered dietitian nutritionists (RDNs).  This study evaluates the perceptions of RDNs working within a large health system. 

Methods: A 26-item electronic survey was created to assess the perceptions of various uses of AI in the dietetics practice.  The survey was sent electronically within a large health care system. Responses were analyzed for descriptive trends.

Results: The survey was distributed to 185 dietitians, yielding a 35% response rate (n = 64). Respondents were diverse in experience, with 37% having over 21 years and 23% having five or fewer years. Work settings were evenly split between inpatient and outpatient care. When asked about artificial intelligence (AI) in dietetic practice, 51% agreed it aligns with the future of healthcare. Views on AI’s impact on care quality were mixed; 44% neutral, and 50% expressed some agreement. Opinions on the accuracy of AI-generated nutrition information also varied, with many respondents remaining neutral or expressing doubt. 75% agreed AI could reduce time spent on routine tasks and about half saw potential productivity gains, 80% of respondents voiced concern over the potential for AI to make errors in clinical recommendations.

Conclusion: These findings reveal a cautious optimism toward AI among experienced dietitians. While many see potential for increased efficiency, concerns remain about AI’s accuracy and reliability, particularly in complex care. Dietitians emphasized the importance of transparency and retaining clinical control. Preference for AI tools that offer clear explanations and the general neutrality regarding AI’s effect on patient engagement and outcomes underscore the need for further education, development, and validation of AI tools in clinical nutrition. 

Disclosure of Interest: None declared