RIAT is an international effort to tackle bias in the way research is reported with the goal of providing more accurate information to patients and other healthcare decision makers.

Randomized controlled trials are known as medicine’s “gold standard” for reliable evidence. However, they are falling short of that standard, in large part due to two fundamental problems:

  1. MISREPORTING: many trials that are published are inaccurately or incompletely reported (misreported trials)
  2. INVISIBILITY: not all trials conducted are published (unpublished trials)

When the original investigators or sponsors do not correct misreporting, or even leave the entire trial unpublished, they can be considered to have abandoned their trial. And the downstream effects can be substantial, drawing to false conclusions about the effectiveness and safety of medical interventions.

The RIAT initiative aims to address these problems by offering a methodology that allows other people to responsibly correct the record.  Anyone who can access trial data and document trial abandonment can use RIAT to:

  1. CORRECT MISREPORTING by republishing a trial (or just those sections of a trial where misreporting has occurred)
  2. CORRECT INVISIBILITY by publishing an unpublished trial

Some RIAT projects will be large, while others will be far more focused.  It all depends on the nature of the trial abandonment. To get a sense of the effort involved, see How to RIAT.

What makes RIAT different than other forms of trial reanalysis?

Reanalysis of trial data is not a new concept.  Here’s what makes RIAT distinct:

  • Reanalysis with independence from the original investigators/sponsor
  • Reanalysis that sticks to the original trial protocol (RIAT is not about data mining, it’s about ensuring the literature contains trustworthy publications of trials that are reported according to the trial’s original protocol)
  • Reanalysis with underlying data available (RIAT authors publish trials with underlying data publicly available, for maximum transparency.)
  • Reanalysis with an audit trail (RIAT authors make clear how they selected data with an auditable record of decisions)
  • Reanalysis with impact through discoverability by systematic reviews and clinical practice guidelines (RIAT publications are expected to be credible reports of trials that will be used by systematic reviewers and others conducting evidence synthesis)

The Case for RIAT

Although publication bias (invisibility) is perhaps the most notorious form of reporting bias there are other types of reporting bias which distort contemporary clinical trial literature. The evidence for this has been accumulating in the last decade or so. Below is a non systematic collection of the review papers which have looked at how different types of reporting bias affect clinical trials literature across a whole range of interventions. The reviews did this by comparing different sources of information (usually regulatory) with publications. Grouping is by study main finding, but many of the conclusions overlap.

The RIAT Support Center

“RIATing” a trial (i.e., fixing misreporting/non-publication) is a large and complex undertaking. The RIAT Support Center aims to provide support to researchers working towards the complete and accurate dissemination of misreported or unpublished clinical trials. We aim to do this by:

  • Raising awareness of misreporting or invisibility of clinical trials, which are experiments on human beings
  • Providing instruments for the identification of misreported or invisible trials
  • Making sponsors and authors take responsibility for the misreporting or invisibility of their trials
  • Publishing a restoration of misreported or invisible trials if sponsors refuse to take responsibility (abandonment)
  • Supporting the restoration of the written record with advice and resources