Background: Evidence is growing for the existence of selective reporting of research results. While it has been known for years that positive studies are more likely to be published based on their results, recent evidence has shown that even among published trials, there is selective reporting of outcomes. In a recent study (2002/2003), CIHR found that even its large, government funded trials were susceptible to selective reporting. These biases threaten the reliability of trials and reviews.

Since 1999, in the effort to move to a new level of accountability, CIHR has conducted several evaluations of RCTs, including a workshop and the outcome reporting study. Recommendations and findings of this process led to actions aimed at improving the reliability of the medical literature.

Objectives: To outline efforts by the Canadian Institutes of Health Research (CIHR) to improve the quality of trials and their reporting and reduce bias in the dissemination of results from trials that it funds. Results: CIHR has taken three major steps to address the issue of selective reporting: 1) The creation of the CIHR Randomized Controlled Trial Database to capture information about trials at three major stages: upon approval, interim, and at trial completion. In particular, trial characteristics and methodological details from protocols will be recorded to promote accountability and transparency 2) Compusory trial registration in a public registry with the International Standard Randomized Controlled Trials Number (ISRCTN). 3) Mandatory final reports submitted to CIHR after trial completion are to be written based on the CONSORT statement.

Conclusions: Through these initial steps, CIHR aims to promote reliable trial reporting. However, much remains to be done by funding agencies, journals, and the research community to ensure the public dissemination of protocol details and trial results.