DIA/FDA Biostatistics Industry and Regulator Forum

There is currently a drive and energy behind the use of “big data” and “real world data”. There are three main motivations: Use of RWE data are viewed as 1) a quicker and less expensive approach to clinical development, 2) a way to answer questions of how treatments are or will be used in the real world, and 3) where randomized trials are not ethically possible. We will restrict the discussion to studies done post approval of the first indication as “RWE” is a non-sequitur for an unapproved drug. One of the current issues is the blurring in the use if RWE to answer scientific and medically important questions and if RWE is somehow distinct from the conduct randomized trials and only used to do observational studies. In this sense, the debate is often posed as “should I do a RWE study OR a clinical trial” or “how can I show that an Obs study will yield the same answer as a RCT”. The truth for drug development is the this should be an “and” discussion not an “or” discussion as both approaches can be valuable. In addition, should more care should be given not to avoid the concept of randomization too quickly in the use of RWE as there many techniques allow us to bridge that gap?