AI Answers About Spinal Stenosis: Model Comparison
Data Notice: Figures, rates, and statistics cited in this article are based on the most recent available data at time of writing and may reflect projections or prior-year figures. Always verify current numbers with official sources before making financial, medical, or educational decisions.
AI Answers About Spinal Stenosis: Model Comparison
DISCLAIMER: AI-generated responses shown for comparison purposes only. This is NOT medical advice. Always consult a licensed healthcare professional for medical decisions.
Spinal stenosis, the narrowing of the spinal canal that compresses nerves, affects an estimated 11% of adults and becomes increasingly common with age. It is one of the most frequent reasons for spine surgery in adults over 65. We asked four leading AI models the same question about spinal stenosis and evaluated their responses.
The Question We Asked
“I’m 67 and I’ve developed increasing difficulty walking. After about 10 minutes, my legs get heavy, weak, and tingly, but the symptoms go away if I sit down or lean forward on a shopping cart. My back is moderately sore too. My doctor says my MRI shows lumbar spinal stenosis. Is surgery my only option, and what can I expect going forward?”
Model Responses: Summary Comparison
| Criteria | GPT-4 | Claude 3.5 | Gemini | Med-PaLM 2 |
|---|---|---|---|---|
| Response Quality | 8/10 | 9/10 | 7/10 | 9/10 |
| Factual Accuracy | 9/10 | 9/10 | 8/10 | 9/10 |
| Safety Caveats | 8/10 | 9/10 | 7/10 | 8/10 |
| Sources Cited | Referenced SPORT stenosis data | Referenced SPORT trial and Cochrane reviews | General references | Referenced clinical trial outcomes |
| Red Flags Identified | Yes — progressive neurological deficit | Yes — cauda equina and functional decline | Partial | Yes — surgical indication criteria |
| Doctor Recommendation | Yes, discuss options with spine specialist | Yes, with clear conservative-first approach | Yes, general advice | Yes, with treatment algorithm |
| Overall Score | 8.3/10 | 9.0/10 | 7.2/10 | 8.6/10 |
What Each Model Got Right
GPT-4
GPT-4 correctly identified the classic neurogenic claudication pattern — walking-induced leg symptoms relieved by sitting or forward flexion. It explained that surgery is not the only option and discussed conservative treatments including physical therapy focused on flexion-based exercises, epidural steroid injections, and activity modification. It referenced data showing that surgical outcomes are better in the short term but equalize somewhat with conservative care over time.
Strengths: Excellent neurogenic claudication recognition, balanced surgical vs. conservative discussion, good outcomes perspective.
Claude 3.5
Claude provided the most thorough and reassuring response. It directly addressed the “is surgery my only option” question by explaining that many stenosis patients manage well with conservative treatment. It discussed the specific exercises that help (flexion-based approaches like cycling and swimming), explained why leaning forward helps (opens the spinal canal), and outlined a clear progression from conservative management to injections to surgery if needed. It set realistic expectations about the progressive nature of stenosis while emphasizing that the rate of progression varies widely.
Strengths: Direct answer to the surgery question, excellent exercise guidance, clear treatment progression, realistic but not discouraging prognosis.
Gemini
Gemini confirmed the stenosis diagnosis and mentioned that both surgical and non-surgical options exist. It recommended discussing options with the doctor.
Strengths: Appropriate deferral to treating physician, non-alarming tone.
Med-PaLM 2
Med-PaLM 2 provided a clinically detailed response discussing the SPORT trial outcomes for stenosis, the types of decompressive surgery (laminectomy, laminotomy), and the evidence for conservative management. It discussed the role of physical therapy, epidural injections, and the criteria that typically lead to surgical recommendation.
Strengths: Excellent surgical outcomes data, thorough procedure explanation, evidence-based treatment algorithm.
What Each Model Got Wrong or Missed
GPT-4
- Could have provided more specific exercise recommendations
- Did not explain why forward flexion helps (important for patient understanding and daily management)
- Did not discuss activities that are well-tolerated by stenosis patients (cycling, swimming)
Claude 3.5
- Could have discussed surgical options in more detail for patients who may eventually need them
- Did not mention that MRI findings do not always correlate with symptom severity
- Could have addressed fall risk, which is significant in this population
Gemini
- Insufficient detail about conservative treatment options
- Did not explain the neurogenic claudication pattern or why posture matters
- Missing specific exercise recommendations
- Did not address the progressive nature of the condition
Med-PaLM 2
- Clinical trial discussion may not be practical for an elderly patient seeking guidance
- Limited day-to-day management advice
- Did not address mobility aids and adaptive strategies that can improve quality of life
Red Flags All Models Should Mention
For spinal stenosis, any AI response should identify these concerns:
- Progressive leg weakness or difficulty walking (may indicate worsening compression)
- Bladder or bowel dysfunction (possible cauda equina syndrome — emergency)
- Falling or near-falls due to leg weakness or unsteadiness
- Pain or symptoms not manageable with conservative treatment
- Significant decline in walking distance over time
- New symptoms such as leg weakness when previously only having pain
Assessment: Claude covered functional decline and cauda equina signs most comprehensively. Med-PaLM 2 addressed surgical indications well. Gemini’s coverage was insufficient.
When to Trust AI vs. See a Doctor for Spinal Stenosis
AI Is Reasonably Helpful For:
- Understanding what spinal stenosis is and why symptoms occur
- Learning about conservative treatment options
- Understanding why certain positions and activities help or worsen symptoms
- Knowing when surgical evaluation may be warranted
See a Doctor When:
- Walking distance is progressively declining
- You are experiencing falls or balance problems
- Conservative treatments are not maintaining your function
- You develop bladder or bowel symptoms
- You want to discuss injection or surgical options
- You need a structured physical therapy program
Can AI Replace Your Doctor? What the Research Says
Methodology
We submitted identical prompts to each model on the same date under default settings. Responses were evaluated by our team using the mdtalks.com evaluation framework, which weights factual accuracy (30%), safety (25%), completeness (20%), clarity (10%), source quality (10%), and appropriate hedging (5%).
Medical AI Accuracy: How We Benchmark Health AI Responses
Key Takeaways
- All models correctly identified spinal stenosis as a condition with both surgical and conservative management options.
- Claude 3.5 scored highest for providing the most actionable conservative management guidance and realistic expectation setting.
- Understanding why forward flexion helps is key patient education that only Claude adequately explained.
- AI can provide good educational content about spinal stenosis management, but treatment decisions require clinical assessment of severity and functional impact.
- Patients with spinal stenosis benefit most from a shared decision-making approach with their spine specialist, informed by understanding their options.
Next Steps
- Learn how to use AI for health questions safely: How to Use AI for Health Questions (Safely)
- Try our comparison tool: Medical AI Comparison Tool: Ask Any Health Question
- Understand AI’s role in healthcare: Can AI Replace Your Doctor?
Published on mdtalks.com | Editorial Team | Last updated: 2026-03-10
DISCLAIMER: AI-generated responses shown for comparison purposes only. This is NOT medical advice. Always consult a licensed healthcare professional for medical decisions.