Abstract: FR-PO946
Enhancing Patient Comprehension of Glomerular Disease Terminology: Role of Artificial Intelligence (AI) in Simplifying Medical Communication
Session Information
- Glomerular Diseases: Potpourri
October 25, 2024 | Location: Exhibit Hall, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Glomerular Diseases
- 1402 Glomerular Diseases: Clinical, Outcomes, and Therapeutics
Authors
- Abdelgadir, Yasir, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Thongprayoon, Charat, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Craici, Iasmina, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Cheungpasitporn, Wisit, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Miao, Jing, Mayo Clinic Minnesota, Rochester, Minnesota, United States
Background
Glomerular disease is complex and difficult for patients to understand due to its involvement in various areas such as renal anatomy, pathophysiology, immunology, genetics, and diverse treatment strategies. This study employed the AI language model ChatGPT to simplify glomerular disease terms, proposing a new way to improve patient comprehension of this condition.
Methods
Seventy-three terms related to glomerular disease were analyzed using GPT-4 through two inquiries: one for easy-to-understanding and another for patients with an education level of 8th grade or lower. GPT-4’s accuracy was scored from 1 (incorrect) to 5 (correct and comprehensive), and readability was assessed using the Consensus Reading Grade (CRG) Level, incorporating the Flesch-Kincaid Grade (FKG) and SMOG indices. Paired t-test compared the accuracy and readability of both inquiries.
Results
GPT-4’s easy-to-understanding explanations of glomerular disease terms averaged a 12th-grade readability level (12.5 by CRG and 12 by FKG), reflecting the topic’s complexity (Fig.). When inquired for patients with 8th grade or lower, readability improved, averaging 9.7 by CRG and 8.7 by FKG, with the SMOG index indicating a further reduction from 10.2 to 7.2. However, the accuracy in GPT-4’s explanations for patients at or below the 8th-grade level was significantly lower than that for easy-to-understanding explanations (p=0.01).
Conclusion
While GPT-4 effectively simplified information about glomerular diseases, it compromised its accuracy in the process. Although ChatGPT has the potential to contribute to enhancing patient understanding of glomerular disease, it is crucial to closely examine and work on improving its accuracy when generating simplified explanations with human nephrologist to ensure that the information provided is not only easy to understand but also medically accurate and appropriate for patient education.