Manual review of clinical data analysis, though essential for accuracy, is often time-consuming and costly. While automated tools like Pinnacle21 validate CDISC standards and ensure data consistency, the need for expert manual inspection remains a bottleneck in the process. Recognizing this challenge, SGS and SAS have joined forces to create an innovative solution that leverages Large Language Model (LLM)-based AI agents to streamline code validation.
Working with the Statistical Analysis Plan (SAP), the solution will use AI to conduct initial code reviews, identify programming mistakes, and pre-emptively correct issues. The solution is under further development to automate tasks such as generating define.xml files - a critical component in regulatory submissions.
SGS contributed its expertise in biostatistics, data standards and CDISC validation to guide the AI agent’s design and testing. Acting as a CRO partner, SGS ensures that automation aligns with real-world clinical workflows and meets the stringent expectations of regulators. This collaboration bridges advanced analytics with the operational and compliance realities of clinical research.
“This collaboration with SAS exemplifies how combining deep domain expertise with cutting-edge technology can lead to meaningful innovation. By contributing our field knowledge to the development of AI-driven validation tools, we’re helping shape solutions that not only accelerate clinical workflows but also uphold the highest standards of data quality,” said José Miguel Fernández Pueyo, Pharma Clinical Research Biostatistics Manager at SGS.
Brittany Shriver, global leader of life sciences industry consulting for Global Health and Life Sciences at SAS, said, “Our vision is to develop a suite of intelligent agents that support every stage of the drug submission process, from raw data transformation to final report generation. The AI agent’s ability to learn from context and examples, and its integration with secure environments, ensures both speed and data confidentiality.”
“The collaboration between SGS and SAS marks a significant step forward in the use of generative AI for enterprise knowledge management in clinical research, promising to minimize manual review, reduce costs and improve data integrity for regulatory submissions,” Shriver said.
Learn more about how SAS delivers the most productive AI platform to transform data into life-changing insights and accelerate breakthroughs in life sciences here.
