Exploring body donation communication with large language models: Accuracy, readability, and ethical considerations

Area of Science:

• Anatomical Sciences
• Medical Education
• Artificial Intelligence

Background:

• Educational materials for whole-body donation are crucial for anatomy programs but face supply shortages.
• Existing communication methods may not meet the demand for accurate, accessible information on body donation.
• Large Language Models (LLMs) offer a potential solution to enhance donor communication and educational material creation.

Purpose of the Study:

• To investigate the efficacy of LLM-based approaches in generating effective communication materials for whole-body donation.
• To provide a comparative quantitative benchmark and evaluation framework for LLM performance in this domain.
• To assess the accuracy, quality, readability, and vocabulary diversity of LLM-generated responses to body donation FAQs.

Main Methods:

• Five LLMs (ChatGPT-4o, Grok3.0, Claude4Sonnet, Gemini2.5 Flash, DeepSeekR1) generated Turkish responses to six common body donation questions.
• Four anatomists evaluated the generated content based on accuracy, quality, readability, and vocabulary diversity.
• Statistical analysis was performed to compare the performance differences between the LLM models.

Main Results:

• ChatGPT-4o and Grok3.0 significantly outperformed other models in quality scores (mean checklist scores 21.7 ± 2.8 and 21.0 ± 5.1, respectively).
• Top-performing LLMs produced content at a below-secondary-school reading level, indicating high readability.
• LLM-generated materials demonstrated potential for promoting whole-body donation by providing accessible and reliable information.

Conclusions:

• LLM-produced materials can effectively supplement whole-body donation information, potentially streamlining content creation and reducing staff workload.
• Ethical transparency, cultural sensitivity, and continuous human oversight are crucial safeguards for LLM use in sensitive donation contexts.
• Clear governance frameworks, expert audits, and disclosed quality metrics are recommended for responsible LLM implementation in body donation communication.