Title
|
! Beyond Intention to Treat: New Methodology for Estimating Treatment Effects in Clinical Trials
|
Date / Time / Room
|
Sponsor
|
Type
|
08/14/2002
2:00 PM -
3:50 PM
Room: H-Madison Suite
|
Biometrics Section*, Biopharmaceutical Section*, ENAR
|
Invited
|
Organizer:
|
Sati Mazumdar, University of Pittsburgh
|
Chair:
|
Howard E. Rockette, University of Pittsburgh
|
Discussant:
|
3:20 PM - Peter Lachenbruch, Food and Drug Administration
|
Floor Discussion
|
3:40 PM
|
Description
The session will start with a discussion of varying definitions of what constitutes intention to treat (ITT), regulatory mandates and examples providing evidential basis of successful submissions. These examples will range from ITT in the strong sense, through ITT in the weak sense to Not ITT. The rest of the session will focus on new methodology for estimating treatment effects in clinical trials in the face of events such as loss to follow-up, noncompliance and confounders associated with switching treatments. Bayesian hierarchical models to estimate intent-to-treat effect (ITE) and average causal effect (ACE) and propensity adjusted instrumental variable estimates will be discussed with application to clinical trials.
|