aceSelect large language models outperform hip preservation experts on consensus-based hip preservation questionnaire
Artificial intelligence (AI) is increasingly utilized in medical education and clinical contexts, yet few studies compare the performance of large language models (LLMs) to subspecialized experts in providing guideline-based medical information on hip preservation. The purpose of this study was to evaluate the performance of three LLMs compared to a panel of international hip preservation experts in answering guideline-based questions related to femoroacetabular impingement syndrome, hip dysplasia and microinsta…
Three large language models achieved higher accuracy than a panel of hip preservation experts on a 21-item consensus-based questionnaire covering femoroacetabular impingement syndrome, hip dysplasia and microinstability.
Even when incorrect, ChatGPT and Claude produced thorough justifications, creating risk of convincing but wrong guideline-based information, while Gemini showed formatting deviations.
Findings are based on a small expert sample of 10 respondents and a structured 21-item verifiable question set, limiting generalizability to real-world clinical decision-making, and the study is labeled Level V evidence.
Evidence
- Peer-reviewedKnee Surgery, Sports Traumatology, Arthroscopy2026-09-07
How should this claim be treated?
Truvace Impact Record TRV-2026-1018, v1: “Select large language models outperform hip preservation experts on consensus-based hip preservation questionnaire.” Truvace, 2026-09-08. /record/TRV-2026-1018 (accessed at citation time). sha256 a9faadcee3050f7f…
Calibration history
Every change to this record since certification, in the open. None yet — the reading has held since it entered the record.
Certified into the record
How to verify without trusting this page
Fetch the canonical text of any version from /api/record/TRV-2026-1018 and hash it yourself — for example shasum -a 256 on the saved canonical field. The result must equal content_hash, and each version’s text ends with prev:followed by the prior version’s hash (version 1 chains to 64 zeros). If a single character of any version had been altered since certification, the chain would not reproduce.