Data from: How have generic large language models have progressed in their ability to write clinic letters and provide accurate management plans in the virtual fracture clinic?

This dataset contains 11 simulated Virtual Fracture Clinic (VFC) case scenarios and the corresponding clinic letters and orthopaedic management plans generated by three large language models: GPT-3, ChatGPT (GPT-3.5), and GPT-4. Each prompt response was collected in November 2024 via the ChatGPT web interface. The dataset includes the input prompts, structured patient scenarios, and unedited AI-generated outputs. This data was used for evaluation of readability and clinical accuracy in the study “How have generic large language models progressed in their ability to write clinic letters and provide accurate management plans in the virtual fracture clinic?” Files are organised by case number and LLM version. This dataset supports the findings of the published manuscript and is made publicly available for reproducibility and further research into AI applications in orthopaedics.