The Effect of Artificial Intelligence-Generated Audiovisual Education on Osteoporosis Knowledge and Digital Health Literacy

Osteoporosis is a major public health problem in aging societies, leading to weakened and fragile bones and resulting in increased morbidity and mortality. Because the disease is largely asymptomatic, many individuals remain unaware of having osteoporosis until a fracture occurs. Osteoporotic fractures may cause chronic pain, functional limitations, loss of independence, and significant deterioration in quality of life. In this context, educational programs aimed at improving patients' knowledge have emerged as an important preventive strategy.

Previous studies have emphasized that osteoporosis education leads to positive changes in patients' knowledge and health-related behaviors. The International Osteoporosis Foundation (IOF) develops educational materials for patient education, provides digital resources through its official website, and publishes informative brochures. In addition, IOF supports early diagnosis, risk assessment, and preventive strategies by developing community-based awareness campaigns, online educational modules, and guideline documents for healthcare professionals. Within this framework, the easy accessibility of digital educational materials facilitates access to information and contributes to improving osteoporosis-related health literacy, particularly in aging populations.

The use of artificial intelligence-based tools has recently attracted attention as a novel approach to delivering these digital contents effectively and reaching individuals more efficiently. NotebookLM is an artificial intelligence-powered research assistant developed by Google that generates accurate and evidence-based content by learning from user-uploaded multimodal sources, such as documents, images, and videos, and by explicitly citing these sources. By enabling the transformation of uploaded content into both audiovisual formats, NotebookLM offers important opportunities for enhancing patient education. One of the key features of NotebookLM is its "source-grounded" approach. By generating information strictly based on researcher-provided sources, this approach enhances transparency, verifiability, and reliability of information in patient education processes.

In line with these features, the NotebookLM platform was utilized for the development of educational materials in this study. Patient-oriented informational pages published by the International Osteoporosis Foundation at https://www.osteoporosis.foundation/patients/about-osteoporosis were uploaded to the NotebookLM platform as source materials. Educational materials in PDF format covering the definition of osteoporosis, risk factors, diagnosis, treatment, and prevention strategies were uploaded to the platform in segments. Based on these materials, audiovisual digital educational contents were generated, including an audio-based digital content series (podcasts) and a visual digital content series (videos). The prepared podcast and video contents were reviewed by two specialist physicians, and their scientific validity and consistency with the current literature were verified.

Following the preparation and validation of these contents, the protocol for participant implementation was designed. Participants will be randomly assigned to two groups. One group will receive audio-based digital educational content (podcast series) generated via the Google NotebookLM platform, while the other group will receive visual digital educational content (video series). In both groups, the Osteoporosis Knowledge Test, the Awareness and Behavioral Intention Questionnaire, and the Health Literacy Questionnaire will be administered before and after the educational intervention to evaluate changes in knowledge level, awareness, and health-related behaviors.