|
Causal and noncausal associations between public health
or clinical interventions and health outcomes are often confused by the medical
science community, media, public, and policymakers. This can have important
consequences: large randomized trials can be launched based on insufficient
information or delayed (or not launched) based on unjustified inferences, and
as a result, ineffective or harmful health policy or practice can be instituted
or continue.
In light of recent large clinical trials that yielded
unexpected outcomes, including harms, this workshop was convened to explore the
central question of how mistakes can be avoided when deciding whether or not to
launch large "definitive" clinical or community trials. The event was
envisioned as a "morbidity and mortality" or "M&M" conference on medical
evidence, akin to the process used in teaching hospitals to review especially
difficult or challenging cases.
Specifically, the workshop's agenda was developed to address these key questions:
- What are the traditional tools to help us judge an
intervention or a body of evidence?
- What new methodologies are available (or needed) to
judge evidence in today's research environment?
- What are research directions for validating methods
to distinguish between causal and noncausal chains of evidence?
The result was 2 days of engaging presentations and
lively discussion among clinical investigators, biostatisticians, and NIH
leaders.
As a service to the workshop participants and the
clinical research community at large, video of the workshop sessions and
speakers' presentations are available on this Web site. Two reports of the
workshop are available. (See links at left.)
|
