Researchers ‘very surprised’ by scale of errors in clinical trial registries
Researchers recently examined 10,492 clinical trials registered on both ClinicalTrials.gov and the EU Clinical Trials Register (EUCTR) for completion and reporting status discrepancies.
According to the study, 16.2% of dual-registered trials had discrepant completion statuses and 33.9% of trials listed as “ongoing” on EUCTR were listed as “completed” on ClinicalTrials.gov.
The study was inspired during the creation of OpenTrials, a collaboration between Open Knowledge International and Dr Ben Goldacre from the University of Oxford DataLab. The project aims to provide a centralized home for all clinical trials data.
During the build of the database, Goldacre and the study’s co-author Jessica Fleminger found several errors and discrepancies between different registries.
“We work a lot with trial registry data, and often find it to be inaccurate,” said Goldacre, explaining that he and Fleminger wanted to describe the scale of the problem.
“We found that the completion status is very commonly wrong, and that the errors seem to be on the EU trial register," he said. “I expected errors, but I was very surprised by the scale of the errors."
Researchers, doctors, and patients use completion status as an indicator to assess if a trial should have reported results, Goldacre explained.
“As with all issues around trials transparency, whenever there are problems like those documented here, it undermines trust in industry, and in medicine more broadly,” he said.
As Goldacre explained, “Trialists should ensure their registry data is correct. Registry owners should run basic checks to assess whether the data they hold is discrepant.”
“There is no reason for trials to be unreported, nor for trial registry data to be inaccurate.”
Authors: Fleminger J, Goldacre B
Title: Prevalence of clinical trial status discrepancies: A cross-sectional study of 10,492 trials registered on both ClinicalTrials.gov and the European Union Clinical Trials Register
Source: PLoS ONE 13(3): e0193088
DOI: 10.1371/journal.pone.0193088
