PALO ALTO, Calif., Sept. 18, 2019 /PRNewswire/ — Verantos, the market leader in regulatory-grade real-world evidence (RWE), today announced the publication of a study on cardiovascular real world evidence jointly authored with Amgen and researchers from the Stanford University School of Medicine. The study was published in the Journal of American Medical Informatics Association. The study proposes a definition of and approaches to achieve the high standard needed for advanced regulatory grade evidence, specifically real world evidence studies that make clinical assertions for regulatory and reimbursement purposes.
The study focused on identifying the best approaches to create a highly valid observational study in cardiovascular medicine. It assessed several sources of real-world data, including structured and unstructured electronic health records (EHR) data. It also tested several technology approaches, including traditional queries, natural language processing, and artificial intelligence-based inference. The goal was to first define regulatory grade and then assess which real world evidence approaches in cardiovascular medicine produced sufficient data quality to meet the protocol definition.
The Journal of American Medical Informatics Association published the study online on August 12, 2019 and will publish it within an issue this winter. This study was developed through an industry-academic collaboration between Amgen Inc., Verantos, Inc., and the Stanford University School of Medicine. Project support was provided by Amgen, the National Center For Advancing Translational Sciences of the National Institutes of Health and the National Cancer Institute of the National Institutes of Health.
While study authors agree with regulatory agencies that data must be fit for purpose, the study goes one step further to provide a framework to assess data quality. The study discussion incorporates aspects of data accuracy and generalizability assessed in the context of effect size when considering fit for purpose. Wide variation in results between data sources and technologies suggests that data accuracy determination, including sensitivity and specificity, should be a part of regulatory-grade cardiovascular medicine observational study protocols going forward.
One notable finding from the study was the great variation between structured and unstructured EHR data. As an example, sensitivity for myocardial infarction was 29.8% in structured data versus 90.4% in unstructured data using artificial intelligence. This suggests that for a study that uses myocardial infarction as an exclusion criterion, using the wrong data source and not measuring sensitivity could result in 70% of patients with previous heart attacks inadvertently being included in the study without any reviewer being aware.
“This highlights the importance of routinely measuring accuracy, including sensitivity and specificity, in EHR-based studies intended for regulatory use,” says Dan Riskin, M.D., FACS, founder and CEO of Verantos, and senior author on the study. “These are the types of findings that should generate important discussion in the healthcare community as we determine what regulatory-grade RWE means.”
“Amgen believes this manuscript provides an important contribution to understanding the role of RWE in cardiovascular medicine. It reveals and helps to pave a path for what will be required for regulatory-grade studies,” says Keri Monda, Ph.D., Director, Center for Observational Research at Amgen. “Bringing our collective expertise to bear, this study examined one of the harder and more important challenges in real world evidence — evaluating data quality in advanced real world evidence.”
“We believe the study demonstrates new science on data quality to support valid clinical assertions in real world evidence,” says Tina Hernandez-Boussard, Ph.D., Associate Professor of Medicine, Biomedical Informatics, of Biomedical Data Sciences, and of Surgery at the Stanford University School of Medicine, and first author on the study. “We look forward to continuing our efforts to elucidate the more challenging aspects of regulatory-grade RWE studies.”
Verantos (https://verantos.com) is the market leader in high accuracy real world evidence (RWE) generation. The Verantos RWE platform integrates heterogenous real world data sources and generates evidence with the accuracy necessary for regulatory and reimbursement use. The Verantos RWE platform leverages data science and artificial intelligence along with advanced data sources such as electronic health records (EHR) to generate RWE capable of supporting clinical assertions.
CONTACT: Heather Kerr, 1-203-504-8230, [email protected]
View original content to download multimedia:http://www.prnewswire.com/news-releases/verantos-announces-publication-of-a-real-world-evidence-collaborative-study-with-amgen-300920621.html