|
Abstract:
|
The updated CONSORT guidelines (2010 BMJ) state that "Overwhelming evidence shows that the quality of reporting of randomized controlled trials is not optimal." This results in "biased results" and can "mislead decision making in health care at all levels," and is one of the causes for the low reproducibility rate of medical research.
The goal of the talk is to build on the CONSORT checklist to improve the credibility of results reported, especially for late stage (Phase 2 and 3) blinded trials, by providing specific recommendations of the changes that should be made. The main theme is: a hierarchical display and discussion of results with emphasis first on the primary endpoint, next on the alpha controlled secondary endpoints, followed by other secondary endpoints, and other pre-specified endpoints, and finally by post hoc results. Post hoc includes endpoints, populations (e.g. subgroups), and analysis methods that were identified after seeing the data even if applied to the primary endpoint.
|
