Data Validity and Regulatory-Grade Real World Evidence

Introduction:

Data validity is the cornerstone of evidence-based medicine. As randomized controlled trials are augmented by real world evidence, data validity has come under question. Are the data sufficiently valid to support a conclusion? In this webinar, our featured speaker will discuss various types of real world evidence and how to think about data validity.

Learning objectives:

• In what ways are registry-based and electronic health record-based studies different?
• What constitutes a valid real world study?
• How has regulatory guidance evolved related to real world evidence?
• What resources are available to pharma to better leverage real world evidence?

Real world evidence types:

Registry-based studies require that a person at or near the point of care enter data related to patient encounters. The strength of this type of study is the high accuracy of the data entered. The weakness is the lack of scale and flexibility. It is labor intensive and questions must be known in advance to assure proper data elements are captured.

EHR-based studies use existing data sets to extract and follow patient cohorts. The strength of this type of study is scale and flexibility. Expanding the sample size or changing the parameters studied can often be done computationally. The weakness is the need for data accuracy assessment, often requiring advanced technologies such as natural language processing and artificial intelligence to achieve adequate accuracy.

Real world evidence uses:

Current uses of real world data often focus on trial recruitment or internal company insight. When real world data are used for trial recruitment, patients are randomized after recruitment, which may wash out effects of biased recruiting methodology. When real world data are used for internal company insight, trends are sought more often than hard evidence.

As real world data are increasingly used for regulatory and reimbursement filings, data validity becomes a larger issue. In a registry-based study, is the patient population applicable to the future treatment population? Is the sample size sufficient to power subgroup analytics or comparative effectiveness? In an EHR-based study, is the underlying data sufficiently accurate to justify the clinical conclusion?

Conclusion:

Data validity assessment is not new. Whether a clinical assertion is made through a randomized controlled trial or real world evidence, the assertion must always be carefully considered and validity of the assertion assessed.

The field of real world evidence is moving forward rapidly. Clinical, technical, and statistical techniques are evolving to meet current demands. Understanding these techniques represents a critical step toward using real world evidence properly and implementing regulatory-grade real world evidence studies.

Speaker

Dan Riskin, CEO, Verantos

Dan Riskin is Founder and Chief Executive Officer of Verantos and previously held positions as Founder and CEO of Health Fidelity and Special Projects Consultant at Apple. He is Adjunct Professor of Surgery and Adjunct Professor of Biomedical Informatics Research at Stanford University.

Dr. Riskin is an expert in healthcare artificial intelligence and a successful serial entrepreneur. Products he has developed and commercialized influence the care of millions of patients annually. His contributions in data-driven healthcare have been featured in Forbes, The Wall Street Journal, and other leading media. He served on the Obama Healthcare Policy Committee for the 2008 Presidential Campaign and testified before Congress on the 21st Century Cures Initiative.

Dr. Riskin is board-certified in four specialties, including surgery, critical care, palliative care, and clinical informatics. He holds degrees and fellowships from University of California, Stanford, and MIT.

Message Presenter