FigShare (OLD/OUTDATED)

Selective outcome reporting is a significant
methodological concern in medicine leading to bias in clinical trial results.
Comparisons between the outcomes reported in clinical trial registrations and
those later published allow investigators to understand the extent to which
selective reporting bias occurs among trialists. The purpose of this study is
to examine the possibility of selective outcome reporting in randomized
controlled trials (RCTs) published in high impact factor hematology journals. A
PubMed search was conducted to identify randomized controlled trials (RCT)
published in five hematology journals from 2010-2015. After screening, primary
and secondary outcomes were recorded for each trial and then compared with
outcomes prespecified in the trial’s registration. Comparisons were made to determine
if pre-specified trial registry outcomes were changed in the published reports.
Of 113 RCTs, 56 (49.6%) demonstrated major outcome discrepancies between
registry and publication, totaling 119 major discrepant outcomes, the majority
of which favored a statistically significant result. We found a high rate of discordance
between registered and published outcomes in high impact factor hematology journals.
In addition, a high proportion of outcome discrepancies favored publication of
a statistically significant result, suggesting selective outcome reporting
bias.