Synthetic Content Validation of Pediatric Trust Instruments Using Persona-Driven Large Language Models: Development and Evaluation Study

CONCLUSIONS: Persona-prompted LLMs effectively assisted preliminary content validation and may help reduce timelines from months to weeks, but showed limitations in quantitative computations.