Corrections
How to tell us about a paper this search missed, and how corrections are handled.
If you know the paper for one of the 10 trials, the entry is wrong
A trial marked as having no results publication located is a search outcome, and a search can miss things.
Open an issue in the repository at github.com/joooord/et-trials-tracker/issues.
The most useful report includes three things.
- The NCT number of the trial.
- A link to the paper, ideally a DOI or a PMID.
- A line on how the paper corresponds to the registry record, for example the matching intervention, design and sample size.
Errors of any other kind are just as welcome: a wrong completion date, a sponsor name that does not match the registry, a broken link, a miscounted figure, or a sentence that reads as an accusation rather than a count.
Corrections are applied in public
Nothing is altered quietly.
Each accepted change is recorded in CHANGELOG.md in the repository, with the date it was applied, the trials affected and what changed. The full history of the dataset is in the commit log, so any figure here can be traced back to the data it was computed from.
Every figure is recomputed from the data files at build time. Correct the data and the site follows.
Who maintains this
ET Trials Tracker is an independent open data project maintained by Jordan Pitts.
The data and the code that builds this site are public at github.com/joooord/et-trials-tracker. The project has no funding and no affiliation with any trial sponsor, journal or foundation.
Site data as of .