Enhancing patients’ understanding and risk perception of pathology reports using artificial intelligence
Patients increasingly access pathology reports through electronic health records, yet often struggle with medical jargon. This project evaluated whether ChatGPT, an AI language model, could improve patient comprehension and risk perception. A sample colonoscopy report was simplified using ChatGPT and assessed via surveys of 208 patients and 33 pathologists. While 83% of patients found the summary helpful, pathologists noted omissions and inflated risk scores. These findings suggest AI can support patient understanding, but clinical oversight is essential to ensure accuracy, prevent misinformation, and responsibly integrate AI tools into medical communication.
Publication Details
- Journal
- ScholarWorks -A service of University of Vermont Libraries (University of Vermont)
- Published
- 2025-06-02
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00