Researchllmpatient educationddhrandomized trial
LLMs Improve DDH Caregiver Education Outcomes
7.2
A 2026 two-phase study evaluated four large language models (ChatGPT-4, DeepSeek-V3, Gemini 2.0 Flash, Copilot) for caregiver education on developmental dysplasia of the hip (DDH) using expert ratings and a pilot randomized controlled trial. In a 127-caregiver RCT, LLM-generated materials varied in quality and were associated with modest improvements in eHealth literacy and DDH knowledge, with small-to-moderate effect sizes.
Key Points
- 1Evaluated four LLMs (ChatGPT-4, DeepSeek-V3, Gemini 2.0 Flash, Copilot) on accuracy, fluency, and readability.
- 2Found significant inter-model differences in content quality, with ChatGPT-4 and DeepSeek-V3 generally outperforming Copilot.
- 3Demonstrated modest gains in caregivers' eHealth literacy and DDH knowledge in a randomized pilot intervention.
Scoring Rationale
Strong peer-reviewed RCT evidence and expert evaluation support findings, but effects are modest and generalizability is limited.
Sources
Public references used for this report.
Practice with real Health & Insurance data
90 SQL & Python problems · 15 industry datasets
Used by DS/ML engineers at top companies
250 free problems · No credit card
See all Health & Insurance problems
