// AI news · 2026-11-01 · Nature
Evaluation of LLM Responses to Patient Questions on Avascular Necrosis
Five large language models were evaluated for responses to patient questions about avascular necrosis of the femoral head.
Published 2026-11-01 08:00 (Beijing time)Source NatureAuto-summarised, fact-checked against source
01Summary
Claude achieved highest scores on JAMA and DISCERN criteria, with DeepSeek outperforming others in DISCERN. Clinical accuracy varied significantly (p < 0.001), while all responses were rated safe. Flesch-Kincaid grade levels averaged 10.85–13.93, exceeding recommended patient education thresholds. Readability remained poor across platforms, and no model demonstrated consistent transparency or information quality.
Summarised from the original article by an open-source model, then checked automatically: every number and English name must appear in the source. The original article is authoritative.
02Key points
- Claude scored highest on JAMA and DISCERN metrics.
- Clinical accuracy differed significantly among models (p < 0.001).
- Flesch-Kincaid grade levels exceeded recommended patient education thresholds.
- All models showed poor readability and inconsistent transparency.
03Source
Nature · published 2026-11-01 08:00 Beijing time
05Related news
- Same source · Hemp Fiber Concrete Properties and ANN Modeling2026-11-01 · Nature
- Same source · AI-IoT Framework for Sustainable Upgraded Industries2026-11-01 · Nature
- Same topic · AI Agents Still Rely on Humans for Real-World Tasks2026-10-31 · CNBC
- Same topic · OpenAI Updates AI Behavior Guidelines Emphasizing Correctability2026-10-31 · The Decoder
- Same day · Integrated AI Model for HRM Effectiveness2026-11-01 · Nature
From AIMap AI news: collected from 1,100+ official and media sources. Summaries only restate the original article; follow the link for the full story. All stories of 2026-11-01 · Archive by date · All AI news · Today