A new study has found that AI can make highly technical rheumatology research easier for patients to understand without compromising scientific accuracy. Using ChatGPT-4o, researchers showed that AI can simplify the language of scientific papers potentially transforming how patients access and interpret medical information about chronic joint and autoimmune diseases.
Health literacy remains a major challenge worldwide. Most peer-reviewed studies are written at a reading level suitable for those with advanced education, far beyond what most patients can easily follow. The study aimed to see whether AI could help bridge that gap by rewriting dense medical literature into clear, patient-friendly summaries.
The American Medical Association (AMA) and the National Institutes of Health (NIH) recommend that health information for patients be written at a reading level suitable for 6th- to 8th-grade students to ensure better understanding.
How the Study Was Conducted?
The research team selected twelve open-access rheumatology papers authored by senior investigators. They measured the readability of the original texts using the Flesch-Kincaid and SMOG indices, then asked ChatGPT-4o to rewrite the articles in simpler terms. Two independent rheumatologists reviewed the AI versions to ensure medical accuracy and completeness. Results showed a sharp improvement. The average reading level dropped from the 15th to the 10th grade (P < 0.0001), and the length of the summaries decreased from about 3,500 words to just 450 (P = 0.047). Experts confirmed that the AI versions were concise, clear, and still scientifically sound.
“AI tools like ChatGPT can help close the gap between scientific communication and patient understanding,” the authors wrote. “But expert supervision remains essential.”
What It Means for Healthcare?
Despite the progress, the simplified texts were still more advanced than the recommended sixth-grade level for public education materials. Researchers say that human review remains crucial to make the outputs suitable for general audiences. Even so, the study highlights how large language models (LLMs) could help doctors, educators, and health organizations produce more readable resources empowering patients to make better-informed decisions about their care and improving communication in complex fields like rheumatology.
