Do people prefer AI-generated patient educational materials over traditional ones?

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Patient Education

Background:

• Patient Education Materials (PEMs) are crucial for cancer staging comprehension.
• Traditional PEMs are resource-intensive to develop.
• Artificial Intelligence (AI) offers potential for efficient PEM generation.

Purpose of the Study:

• To compare patient and companion preference between traditional and AI-generated colon cancer staging PEMs.
• To evaluate the potential of AI tools like ChatGPT and Google Bard for creating patient education content.
• To inform future strategies for developing accessible and cost-effective cancer information.

Main Methods:

• A study was conducted at UT Health Austin Colon and Rectal Surgery Clinic.
• Participants (patients and companions) were blinded to the PEM generation method.
• Preference was assessed by ranking de-identified traditional (National Cancer Institute, American Cancer Society) and AI-generated (ChatGPT, Google Bard) PEMs.

Main Results:

• 35 participants (ages 27-79) were recruited.
• Preference was nearly evenly split: 51% favored traditional PEMs, 49% favored AI-generated PEMs.
• No statistically significant difference in preference was observed (p=NS).

Conclusions:

• Patients do not exhibit a dominant preference for either traditional or AI-generated colon cancer PEMs.
• AI holds promise for generating PEMs that match traditional quality while reducing resource demands.
• Further research into AI-driven PEM development is warranted to optimize patient education resources.