Artificial intelligence (AI) is quietly revolutionizing the way we communicate about health. Imagine being handed a stack of hospital discharge papers or a pamphlet about heart health, only to find the language so complex that it feels like deciphering a foreign code. For many patients, this is a daily reality. But thanks to recent advances in AI, that’s beginning to change.
A team of researchers at NYU Langone Health set out to tackle a longstanding problem: patient education materials (PEMs) are often written at a reading level far above what’s recommended for the general public. The American Heart Association, American Cancer Society, and American Stroke Association all provide valuable online resources, but their materials typically require a reading level of grade 9 or higher—well above the grade 6 level recommended for health information.
The researchers turned to three powerful large language models (LLMs)—ChatGPT, Gemini, and Claude—to see if AI could help. These models, trained on vast amounts of internet data, are designed to predict the next word in a sentence, making them adept at rewriting complex information in simpler terms. The team selected 60 patient education materials at random and asked the AI tools to make them easier to read.
The results were striking. After AI optimization, the readability of these materials improved dramatically. ChatGPT brought the average reading level down to grade 7.6, Gemini to 6.6, and Claude to an impressive 5.6. Not only did the language become more accessible, but the word count also dropped, making the information more concise and less overwhelming.
Dr. Jonah Feldman, the study’s senior author, emphasized the significance of these findings: even materials already designed for patients can benefit from AI-driven improvements. "Our study shows that widely used large language models have the potential to transform patient education materials into more readable content, which is essential for patient empowerment and better health outcomes," he explained.
This isn’t just theory—NYU Langone is already putting these AI tools to work in real-world settings. They’re running a randomized controlled trial to test whether AI-generated, patient-friendly summaries for hospital discharge instructions can improve patient understanding and satisfaction. The hope is that clearer instructions will lead to better care after patients leave the hospital, reducing confusion and improving recovery.
Dr. Paul Testa, another co-author, highlighted the broader impact: "The breadth of possible AI offerings shows how technology can be leveraged to transform the patient experience across health care systems, and not just in the United States." By making health information more accessible, AI is helping to bridge gaps in health literacy and empower patients everywhere.
Actionable Takeaways:
- If you’re a healthcare provider, consider exploring AI tools to simplify your patient education materials.
- Patients should feel empowered to ask for information in plain language—AI is making this more possible than ever.
- Healthcare organizations can look to studies like this as a blueprint for improving communication and patient outcomes.
Summary of Key Points:
- AI tools like ChatGPT, Gemini, and Claude can significantly improve the readability of patient education materials.
- Simplified materials help patients make better health decisions and improve outcomes.
- Real-world trials are underway to validate the effectiveness of AI-generated health information.
- Healthcare providers and organizations can leverage AI to enhance patient communication.
- Accessible health information is a crucial step toward greater patient empowerment and satisfaction.