GPT-4 Generated Letters Improve Patient Discharge Comprehension

In a prospective randomized crossover study published in 2026, 48 standardized patients compared GPT-4–generated patient-centered discharge letters with standard hospital discharge letters. GPT-4 letters increased fully or partially recalled safety-relevant learning objectives (490/1073, 45.7% vs 413/1073, 38.5%; OR 1.74, 95% CI 1.45–2.08; P<.001) and reduced omissions. Improvements concentrated in medication and organizational content, with smaller effects on higher-order understanding like prevention and lifestyle.
Key Points
- 1Demonstrate GPT-4 patient letters increase fully or partially recalled learning objectives (OR 1.74, P<.001).
- 2Improve medication and organizational comprehension most, addressing high-risk postdischarge information gaps.
- 3Suggest adopting AI-generated patient letters to reduce cognitive load, but supplement for higher-order guidance.
Scoring Rationale
Solid randomized crossover evidence and peer-reviewed publication, but limited to standardized patients and modest higher-order improvements.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems