Back to Agenda
Session 7: Methodological Insights on Aspects of Non-Interventional Studies
Session Chair(s)
John Concato, MD, MPH, MS
Associate Director for Real-World Evidence Analytics, OMP, CDER
FDA, United States
This session will highlight specific methodological issues relevant to using real-world data to generate real-world evidence. The first presentation will discuss the history and evolving landscape of causal methods including propensity scores, g-computation, and methods based on machine learning. The second presentation will discuss the test-negative design derived from the case-control study design. The third presentation will discuss negative control outcome studies designed to assess comparability of potential treatments groups. The third presentation will discuss a structured framework for sensitivity analyses to assess unmeasured confounding. A panel discussion will follow.
Learning Objective : At the conclusion of this session, participants should be able to:- Summarize basic concepts, applications, and methods of causal methods
- Describe advantages and pitfalls of studies using negative control outcomes
- Identify major methodological challenges when planning or evaluating a test-negative study design
Speaker(s)
Strengths and Challenges of Test-Negative Designs to Assess Post-Marketing Vaccine Effectiveness
Delphine Saragoussi, MD, MSc
PPD, part of Thermo Fisher Scientific, France
Executive Director, Epidemiology and Scientific Affairs
Using Negative Outcome Control Studies to Assess Study Validity in the Real-World
David Pritchard, PhD
Target RWE, United States
Director, Data Management & Statistics
Traditions and Frontiers in Causal Methods - From the Rise and Fall of Propensity Scores to Causal AI for RWE
Andrew Wilson, PhD, MS
Parexel, United States
Head of Innovative RWD Analytics
Have an account?