Comparison of registered and published outcomes in randomised trials in dermatology journals: a cross-sectional analysis.

Comparison of registered and published outcomes in randomised trials in dermatology journals: a cross-sectional analysis.
Br J Dermatol. 2020 Jul 11;:
Authors: Steele L, Livesey A, Hong A, Thomson J, Flohr C
Abstract
A potential source of bias in randomised controlled trial (RCTs) is selective outcome reporting bias, where outcomes for reporting are chosen based on the significance of their results.1 Significance can arise by chance when multiple tests are performed ("data dredging"). To avoid this problem, a main outcome (a "primary outcome") should be pre-specified prior to data collection in a time-stamped, publicly-available trial registry. Prospective registration has been a prerequisite for publication amongst International Committee of Medical Journal Editors (ICMJE) member journals since 2005. However, even when trials are prospectively registered, selective outcome bias reporting can occur if the primary outcome reported in the manuscript does not match the pre-specified primary outcome in the trial registry.2.
PMID: 32652527 [PubMed - as supplied by publisher]

Source: The British Journal of Dermatology - July 10, 2020 Category: Dermatology Authors: Steele L, Livesey A, Hong A, Thomson J, Flohr C Tags: Br J Dermatol Source Type: research