More guidance needed to prevent outcome reporting bias

A review of reporting guidelines for randomized controlled trials (RCTs), conducted by Dwan et al. and published in Trials has highlighted the need for more detailed guidance from RCT funders to prevent outcome reporting bias (ORB).

As part of the Outcome Reporting Bias in Trials (ORBIT) project, set up to investigate the prevalence of ORB and its impact on the outcome of systematic reviews, this study assessed reporting guidelines from national and international organizations, as well as UK based charities, which fund RCTs. Assessment was carried out with regard to trial registration, protocol adherence, trial publication and monitoring.

Research
   

Reporting of clinical trials: a review of research funders’ guidelines
Kerry Dwan, Carrol Gamble, Paula R Williamson, Douglas G Altman
Trials 2008, 9:66 (25 November 2008)
[Abstract] [Provisional PDF]

Publication bias, where statistically significant results are more likely to be published than those that are statistically insignificant, is a well-recognized problem. However, ORB, where publications report only a subset of the original variables according to the nature of the results, is an equally important but less well-documented issue.

Dwan et al. found that although trial registration, protocol adherence, trial publication and monitoring were frequently discussed, only 11 of the 73 organizations and charities contacted mentioned the publication of both negative and positive outcomes. The authors conclude that the majority of funders have yet to incorporate ORB into their recommendations, and that more detailed guidance, stating the importance of reporting all primary and secondary outcomes specified in the protocol, regardless of the nature of results, is required to prevent the selective reporting of outcomes.

For more information about Trials, please contact the editorial office.

Victoria Thompson
Assistant Editor – Trials

View the latest posts on the On Medicine homepage

Comments