Generative artificial intelligence implementation in REDCap

Area of Science:

• Health Informatics
• Artificial Intelligence in Research
• Data Management

Background:

• Research Electronic Data Capture (REDCap) is a widely used platform for clinical research data management.
• Integrating advanced technologies like generative AI is crucial for enhancing research efficiency and data analysis capabilities.

Purpose of the Study:

• To detail the initial implementation of generative artificial intelligence (AI) within the REDCap platform.
• To evaluate the feasibility and researcher interest in AI-assisted features using a minimum viable product (MVP) strategy.

Main Methods:

• Developed and deployed three AI-assisted features: writing helper, qualitative data summarization, and language translation.
• Emphasized security, optional adoption, and human-in-the-loop oversight in the implementation.
• Released features as part of REDCap version 15.0.

Main Results:

• 18 institutions globally activated the REDCap generative AI module within seven months of release.
• Eight institutions reported sustained use across 1171 projects.
• Over 5700 generative AI API calls were made at Vanderbilt University Medical Center.

Conclusions:

• The MVP strategy demonstrated the feasibility and researcher interest in generative AI within REDCap.
• Successful adoption is contingent upon institutional AI infrastructure and governance.
• The project offers generalizable insights for secure and responsible AI deployment in research data capture systems.