From Prediction to Practice: The AI Implementation Traps in Healthcare

Authors

• John J. Hanna, East Carolina University

Document Type

Presentation

Abstract

From Prediction to Practice: The AI Implementation Traps in Healthcare. Dr. John Hanna, Associate CHIO, Physician Informaticist, East Carolina University Health (NC)

Objectives:

• Explain why AI reasons over a filtered, compressed slice of clinical reality--and why the human loop is a structural necessity, not a temporary safeguard 
• Describe how clinical AI evolved--rules, predictive models, generative AI
• Explore the shift from episodic care and autonomous prediction toward continuously co-navigated care, with humans at the center 
• Describe why what AI can see-visibility, context, and equity, especially in rural and underserved settings-determines who benefits from AI in medicine

Publication Date

8-26-2026

Recommended Citation

Hanna, J. J. (2026). From Prediction to Practice: The AI Implementation Traps in Healthcare. Retrieved from https://scholar.rochesterregional.org/grandrounds_rgh/183