Can artificial intelligence improve patient educational material readability? A systematic review and narrative...

Area of Science:

• Medical Informatics
• Health Literacy
• Artificial Intelligence Applications

Background:

• Patient comprehension of health information is vital for better health outcomes.
• Artificial intelligence (AI) offers potential for simplifying complex medical data into accessible formats.
• Enhancing health literacy through improved patient education materials (PEMs) is a key healthcare goal.

Purpose of the Study:

• To systematically review the accuracy, reliability, comprehensiveness, and readability of AI-simplified medical patient education materials (PEMs).
• To assess the effectiveness of AI in distilling medical information for improved patient understanding.

Main Methods:

• A systematic review was conducted on studies published between January 2019 and June 2023.
• Inclusion criteria focused on AI use in PEMs, involving physicians and/or patients.
• An inductive thematic analysis was performed on twenty included studies.

Main Results:

• AI effectively simplified PEMs, achieving up to 90.7% reproducibility in certain areas.
• User satisfaction with AI-generated materials exceeded 85%, with notable readability improvements (e.g., ChatGPT achieving 100% readability scores).
• AI performance varied in accuracy and reliability, showing limitations in comprehensiveness and precision for complex medical subjects.

Conclusions:

• AI demonstrates significant potential to enhance patient health literacy by simplifying PEMs.
• Further advancements in AI models and prompt engineering are necessary to improve accuracy, reliability, and personalization.
• Refinement is crucial for AI to effectively handle complex medical information and address limitations in soft skills.