An expedited chart review process for large database studies using natural language processing and multi-wave adaptive sampling

Key Points

• The study aims to improve the efficiency of chart reviews and validate algorithms in large database studies.
• Developed a natural language processing system for chart review.
• Implemented a multi-wave adaptive sampling approach.
• Analyzed the reliability of code-based algorithms used in database studies.
• Improved validation processes for algorithmic definitions of study parameters.
• Enhanced understanding of result reliability in database environments.

Abstract