Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 14, 2026npj Digital MedicineOpen Access

Performance of a large language model in the informed consent process for participation in a clinical trial

View Full Paper
Ask AI
Bookmark
Share

Authors

RMRebeca MoscatelSinai Health SystemKAKomal AryalSinai Health SystemDCDavid ChenSinai Health System

Discussion

Loading...

Member takes

Implication

Randomized trial assesses large language models for informed consent accuracy, indicating potential for efficient evaluations.

Key Points

  • This study aims to evaluate the accuracy and reproducibility of large language models in the informed consent process for clinical trials.
  • Conducted a proof-of-concept study using a randomized clinical trial framework.
  • Evaluated responses generated by a large language model against human ratings for accuracy and readability.
  • Developed tools to improve efficiency in responding to informed consent questions and validating models.
  • Mean accuracy ratings were high, with human ratings averaging 4.8 (95% CI: 4.7-4.9) and LLM ratings at 4.7 (4.6-4.8).
  • Readability scores showed appropriateness for both human (grade 7.5) and LLM (grade 6.4) responses.
  • Human raters displayed significant variability in accuracy ratings compared to the moderator LLM's more consistent scores.

Cite This Study

Moscatel et al. (2026) studied this question.

synapsesocial.com/papers/6a2e44e4b1cc60ccdea8a3b1https://doi.org/10.1038/s41746-026-02745-9
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Evaluating the Use of Large Language Models to Answer Patient-Facing Clinical Trial Questions2025
  2. 2Design and feasibility of lay clinical trial summaries using large language models.2025
  3. 3Large language models for automated consent form generation: A design and feasibility pilot.2025
  4. 4Large Language Models in Randomized Controlled Trials Design2024 · 3 citations
  5. 5Evaluating large language models for simplifying non-English medical consent with clinician involvement2026