Clinical accuracy and applications of large language models in pediatric orthopedics: a systematic review.
J Pediatr Orthop BJun 23, 2026 (epub)
Ibraheem Qureshi, Nithya Thomas, Matthew Heffelfinger et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review evaluated the accuracy and clinical applicability of large language models (LLMs) in pediatric orthopedics, analyzing 15 studies. The pooled accuracy of LLMs was found to be 74.1%, with significant variability in expert consensus scores across regions. The study highlights the potential of LLMs for patient education and decision support, but also their limitations in complex scenarios.
Key findings
- LLM accuracy ranged from 44.3% to 93%, with a pooled accuracy of 74.1%.
- Up to 33% of responses to guideline-based questions were rated neutral or inaccurate.
- Parent surveys indicated 82% trust in AI as supplementary tools with professional oversight.
- Regional expert consensus scores varied significantly, with Europe at 80 and North America at 65.
Why it matters
Understanding the capabilities and limitations of LLMs in pediatric orthopedics is crucial for clinicians considering their use for patient education and clinical decision support. This knowledge can guide appropriate integration into practice.