Related Experiment Video
Updated: May 7, 2026
07:30
3D Printing Model of a Patient's Specific Lumbar Vertebra
Published on: April 14, 2023
2.7K
Can Large Language Models Translate Spine Surgery Guidelines for Patients? A Pilot Validation Study Using AO Spine
Rehan R Khan1,2, Rohith Ryali1,2, Daman P Dhunna1,2
1Department of Orthopedic Surgery, Boston Medical Center, Boston, MA.
Clinical Spine Surgery
|May 5, 2026
Summary
ChatGPT-5 accurately reflected AO Spine guidelines for clinicians treating degenerative cervical myelopathy (DCM). Patient-facing responses were generally aligned but lacked some detail, highlighting the need for expert review.
Area of Science:
- Artificial Intelligence in Medicine
- Medical Guideline Dissemination
- Spine Surgery Decision Support
Background:
- Degenerative cervical myelopathy (DCM) management follows AO Spine guidelines.
- Patients increasingly use AI like ChatGPT for medical information.
- Existing guidelines can be challenging for patients to understand.
Purpose of the Study:
- Evaluate ChatGPT-5's accuracy in generating clinician and patient responses.
- Assess AI-generated content against AO Spine degenerative cervical myelopathy (DCM) guidelines.
- Analyze the readability of AI-generated patient education materials.
Main Methods:
- Pilot cross-sectional validation study.
- ChatGPT-5 queried with clinical scenarios from AO Spine DCM guidelines.
- Outputs graded by spine surgeons for accuracy; patient outputs assessed for readability.
Main Results:
- Clinician-facing outputs were highly accurate (mean 2.93/3) with excellent agreement.
- Patient-facing responses showed lower consistency (mean 2.33/3), omitting key details.
- Patient outputs achieved ~7th-grade reading level, slightly above recommendations.
Conclusions:
- ChatGPT-5 reliably mirrors AO Spine guidelines for clinician-facing content.
- Patient-facing AI responses generally align but may miss crucial details.
- Expert spine surgeon review is vital for accurate patient education and guideline adherence.
