AI Answers About Addison's Disease: 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 Addison’s Disease: 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.
Addison’s disease, also known as primary adrenal insufficiency, is a rare endocrine disorder in which the adrenal glands fail to produce adequate cortisol and aldosterone. It affects approximately ~1 in 10,000 people, with onset most common between ages 30 and 50, and is slightly more prevalent in women. The condition is most frequently caused by autoimmune destruction of the adrenal cortex in developed countries. Because symptoms develop gradually and overlap with many common conditions, the average diagnostic delay is approximately ~5 years. The life-threatening risk of adrenal crisis makes timely diagnosis critically important. We asked four leading AI models the same question about Addison’s disease to evaluate their responses.
The Question We Asked
“I’m 38 and for the past year I’ve had progressive fatigue that sleep doesn’t fix, unexpected weight loss of about 15 pounds, dizziness when I stand up, and salt cravings. My skin has gotten noticeably darker, especially in my skin creases and on my gums. My blood pressure is low at every visit. My doctor ran a cortisol level and it was very low. She suspects Addison’s disease and ordered an ACTH stimulation test. What is this condition, and is it dangerous?”
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 | 7/10 | 9/10 |
| Safety Caveats | 8/10 | 9/10 | 7/10 | 9/10 |
| Sources Cited | Referenced Endocrine Society guidelines | Referenced Endocrine Society, NICE guidelines | Limited sourcing | Referenced diagnostic algorithms and crisis protocols |
| Red Flags Identified | Yes — adrenal crisis risk | Yes — comprehensive crisis education and prevention | Partial | Yes — crisis protocol and emergency injection |
| Doctor Recommendation | Yes, endocrinology management | Yes, lifelong endocrinology care with crisis planning | Yes, general advice | Yes, with emergency protocol establishment |
| Overall Score | 8.4/10 | 9.2/10 | 7.0/10 | 8.8/10 |
What Each Model Got Right
GPT-4
GPT-4 correctly explained Addison’s disease as autoimmune adrenal destruction causing cortisol and aldosterone deficiency. It connected each symptom to the hormonal deficit (hyperpigmentation from elevated ACTH/MSH, salt craving from aldosterone loss, orthostatic hypotension, fatigue from cortisol deficiency). It discussed treatment with hydrocortisone and fludrocortisone replacement and mentioned the need for stress-dose steroids during illness.
Strengths: Excellent symptom-to-mechanism connections, clear hormone replacement explanation, important stress dosing mention.
Claude 3.5
Claude provided the most comprehensive response, explaining the autoimmune pathophysiology, the ACTH stimulation test procedure and interpretation, and the critical concept of adrenal crisis. It discussed lifelong hormone replacement therapy with specific medications and dosing schedules, provided detailed adrenal crisis education including when to use emergency intramuscular hydrocortisone injection, explained the importance of medical alert identification, and discussed sick-day rules for dose adjustment. Claude also addressed the association with other autoimmune conditions (autoimmune polyendocrine syndrome).
Strengths: Outstanding adrenal crisis prevention education, comprehensive sick-day management guidance, excellent emergency preparedness discussion, important autoimmune comorbidity screening, thorough medical alert identification recommendation.
Gemini
Gemini acknowledged that low cortisol levels warranted further testing and that the condition could be managed with medication. It recommended following the doctor’s recommendations.
Strengths: Appropriate endorsement of further testing, reassurance about treatability.
Med-PaLM 2
Med-PaLM 2 provided a clinically precise response discussing the ACTH stimulation test interpretation (failure to reach cortisol of 500 nmol/L), the distinction between primary and secondary adrenal insufficiency, and evidence-based hormone replacement protocols. It outlined a comprehensive adrenal crisis emergency protocol and discussed the importance of establishing an emergency care plan.
Strengths: Excellent diagnostic test interpretation, strong emergency protocol discussion, thorough primary vs. secondary distinction.
What Each Model Got Wrong or Missed
GPT-4
- Insufficient detail about adrenal crisis recognition and emergency management
- Did not mention the need for medical alert bracelet or emergency injection kit
- Limited discussion of sick-day steroid dose adjustment rules
Claude 3.5
- Response length may be overwhelming for a newly suspected diagnosis
- Could have been more concise about the most critical immediate information
- Did not discuss fertility and pregnancy considerations for women with Addison’s
Gemini
- Failed to explain the seriousness of adrenal insufficiency
- Did not discuss adrenal crisis, which is a life-threatening emergency
- Missing discussion of hormone replacement therapy specifics
- No mention of emergency preparedness or medical alert identification
Med-PaLM 2
- Clinical terminology around cortisol thresholds may confuse patients
- Limited practical advice for daily management and lifestyle adjustments
- Did not address the psychological impact of a chronic, lifelong condition
Red Flags All Models Should Mention
For Addison’s disease, any AI response should identify these concerns requiring emergency evaluation:
- Severe abdominal pain, vomiting, and confusion (adrenal crisis — medical emergency)
- Sudden severe weakness and hypotension
- Fever or illness while on replacement therapy without dose adjustment
- Loss of consciousness or altered mental status
- Inability to take oral medications due to vomiting (need for injectable hydrocortisone)
- Dehydration and electrolyte imbalance symptoms
- Any major physical stress, surgery, or trauma (need immediate stress-dose steroids)
Assessment: Claude and Med-PaLM 2 provided the most comprehensive responses, particularly regarding crisis prevention. GPT-4 covered fundamentals well. Gemini was dangerously insufficient for a life-threatening condition.
When to Trust AI vs. See a Doctor for Addison’s Disease
AI Is Reasonably Helpful For:
- Understanding the basics of adrenal insufficiency and hormone replacement
- Learning about sick-day rules and stress dosing concepts
- Understanding why medical alert identification is essential
- Preparing questions for endocrinology appointments
See a Doctor When:
- You have symptoms suggesting cortisol deficiency (fatigue, weight loss, hyperpigmentation)
- You need ACTH stimulation testing and hormone level assessment
- You need to establish a hormone replacement regimen
- You are ill and need to adjust steroid doses (contact doctor promptly)
- You experience signs of adrenal crisis (emergency — call 911/use emergency injection)
- You need screening for associated autoimmune conditions
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 four models recognized the clinical picture, but crisis education quality varied dramatically and represents the most important safety differentiator.
- Claude 3.5 scored highest for its comprehensive adrenal crisis prevention education and practical emergency preparedness guidance.
- The most critical finding: adrenal crisis is fatal without treatment, and every Addison’s patient must understand sick-day rules, carry emergency hydrocortisone injection, and wear medical alert identification at all times.
- AI can help patients understand their condition and prepare for medical appointments, but cannot replace the endocrinology expertise, hormone monitoring, and crisis planning this lifelong condition requires.
- Patients with Addison’s disease should work with their endocrinologist to establish a written emergency protocol and educate family members about crisis recognition and injection administration.
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.