Verantos-led Study in JAMA Network Open Demonstrates a Defined Approach to Measure and Optimize Data Reliability

Early results of the TRUST Study, a 4-year FDA-funded demonstration project in data reliability, reveal key learnings

PALO ALTO, Calif.--(BUSINESS WIRE)--Verantos, the global leader in high-validity real-world evidence at scale, today announced the publication “Implementing Accuracy, Completeness, and Traceability for Data Reliability” in JAMA Network Open. The study, written in collaboration with a consortium of academic and industry leaders, proposes a reproducible framework for measuring and optimizing the reliability of real-world data (RWD).

In 2024, the FDA published the final guidance document “Real-World Data: Assessing Electronic Health Records and Medical Claims Data To Support Regulatory Decision-Making for Drug and Biological Products.” Included in that guidance is the recommendation that accuracy, completeness, and traceability be measured and sufficient. Verantos led the TRUST study, an FDA-sponsored demonstration project intended to determine the impact of data quality in real-world evidence and evaluate data reliability over three dimensions. The manuscript published today includes early learnings from the TRUST study.

The authors defined objective measurement methods for each of the components of data reliability. To provide context in practice, researchers measured the reliability of a “traditional” approach using only claims data and an “advanced” approach using curation, linkage, and artificial intelligence. The study compared the traditional and advanced approaches in real-world evidence to understand their impact on data reliability metrics.

The manuscript analyzes data from 120,616 individuals from 58 hospitals and more than 1,180 outpatient clinics across the United States. A rigorous method of data reliability measurement was described and applied. The traditional approach yielded a 59.5% F1 score for accuracy, 46.7% completeness, and 11.5% traceability. The advanced approach yielded values of 93.4%, 95.6%, and 77.3%, respectively.

“This research provides concrete steps for understanding and optimizing data quality in real-world evidence,” said Dan Riskin, MD, FACS, founder and CEO of Verantos, Clinical Professor of Surgery at Stanford University School of Medicine, and lead author on the study. “Our study not only demonstrates a path forward for life sciences firms that require measured and high data reliability in regulatory submissions, but also supports credible evidence generation across the industry to enable a learning health system.”

Life sciences firms seeking clarity on measurement of data reliability and how to achieve fit-for-use data may benefit from early learnings from the 4-year FDA-funded demonstration project. Payers and providers seeking credible real-world evidence will also benefit from better understanding the evolving evidentiary standards on data reliability and relevance.

About Verantos

Verantos is the global leader in high-validity real-world evidence for life sciences organizations. By incorporating robust clinical narrative data and artificial intelligence technology, Verantos is the first company to generate high-validity evidence at scale across therapeutic areas with measured accuracy, completeness, and traceability. Credible evidence accelerates clinical development, market access, medical affairs, pharmacovigilance, and regulatory initiatives.