What question did this study set out to answer?

The aim is to propose a framework for validating the quality of clinical data extracted by large language models from electronic health records.

Ensuring Reliability of Curated Electronic Health Record–Derived Data: The Validation of Accuracy for Large Language Model–/Machine Learning–Extracted Information and Data (VALID) Framework

View Full Paper

Ask AI

Bookmark

Authors

The VALID framework evaluates the accuracy of LLM-extracted data in oncology, suggesting improvements in data quality and reliability.

The aim is to propose a framework for validating the quality of clinical data extracted by large language models from electronic health records.
Development of a validation framework for LLM-extracted clinical data.
Benchmarking performance against expert human abstraction.
Conducting verification checks for internal consistency and plausibility.
Replication analyses comparing LLM data to human-abstracted and external standards.
Bias assessment by analyzing demographic subgroup data.
Identified critical variables needing improvement in data extraction.
Established systematic detection processes for latent errors.
Confirmed fitness-for-purpose of data sets for real-world research.
Enhanced industry standards for data quality in AI applications in oncology.

Estévez et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark